ENCYCLOPÆDIA BRITANNICA.

COMPARATIVE ANATOMY.

PART I.

ANATOMY OF THE ORGANS OF RELATION.

CHAP. I.—COMPARATIVE OSTEOLOGY.

Comparative Anatomy. RED-BLOODED ANIMALS only can be said to possess that assemblage of bones denominated skeleton; and as in these the most constant part is the vertebral column, it furnishes the general character of Vertebrated. The shells of the MOLLUSCA and the Crustacea have been generally regarded as a species of internal skeleton; and in the circumstance of affording mechanical support and external protection, they indeed resemble the skeleton of the VERTEBRATA. But neither in mode of development nor in chemical constitution can they be regarded as of the same nature. Hence it is only in the vertebrated classes that it is requisite to study the peculiarities of the skeleton.

Bones generally. In general characters the bones of the Mammalia resemble those of the human subject. Like them, they are white, firm, elastic, and incompressible. They consist also of compact and reticular or cancellated tissue. In the lower animals the latter kind of structure is in general coarser and looser than in man; and in the CETACEA especially the cavities are large. In the carnivorous animals the compact structure is exceedingly dense, and gives the bone much greater weight than in other animals. In the CETACEA also the acoustic or lithoid portion of the temporal bone is of a marble hardness.

The bones of the Mammalia may, like those of man, be distinguished, according to their mechanical form, into long, flat, and short bones. Though the first class in general possess a medullary canal, this cavity is imperfect or wholly wanting in the bones of the CETACEA and AMPHIBIA.

The cavities denominated sinuses are much more completely developed in several of the MAMMALIA than in the human skeleton. In the pig these cavities extend into the occipital bone; in the elephant they not only give the frontal bone extraordinary protuberance, but they extend into the parietal, temporal, and even the occipital bones, and contribute much to augment the volume of the head. In the ox, deer, and sheep, they communicate with the cavity of the horns.

The bones of BIRDS are in general whiter, firmer, and smoother than those of the MAMMALIA; and they consist of a firm, compact substance, which is elastic and hard

in the bones of the trunk, and extremely brittle in those of the extremities. With the exception also of some of the thin, flat bones, as the sternum, they consist of thin, compact walls, inclosing large capacious cavities, which contain not marrow, but air, and which communicate by one or more minute holes with the windpipe and lungs. While these cavities, which may be regarded as the most perfect and advanced form of sinuses, diminish considerably the weight of the whole skeleton, by the density and completely cylindrical shape, they rather augment the strength. In the chick, and at birth, the bones are homogeneous and without cavities; afterwards they contain marrow; and eventually this disappears and gives place to air.

The bones of the REPTILES are not remarkable in any respect, unless in being void in general of medullary cavity. The absence of this canal was originally observed by Caldesi, and afterwards by Cuvier, in the tortoise; by Troja in the bones of the frog and toad; and by Carus in those of the turtle. In the crocodile, however, and in several of the lizard family, they are large and distinct. The bones of Reptiles also are less firm than those of Birds and Mammals.

The bones of FISHES are remarkable for great softness, flexibility, and elasticity. Those of the lamprey, shark, ray or skate, and sturgeon family, are soft, flexible, sectile, of a bluish white colour, translucent, and so elastic that a cutting instrument forced into them is speedily retracted by the resilient nature of the bony matter. From these characters, the bones of these families have been regarded as cartilaginous, and the fishes themselves have been distinguished by this character. (PISCES CARTILAGINEI, PISCES CHONDROPTERYGII.) In the other fishes, the bones, though softer than those of Mammals, Birds, and Reptiles, present a greater degree of firmness and solidity, are whiter and more opaque, and are much less sectile, than those of the cartilaginous division. As in this respect, therefore, they approach the genuine bone of the Mammals, these are distinguished as fishes with osseous skeletons. (PISCES OSSEI.)

The bones of both classes of fishes consist of a large quantity of gelatine, with a small proportion of phosphate

Comparative Anatomy. of lime. In the osseous fishes, however, the latter substance is most abundant. The colour of the bones of fishes, though in general whitish gray, is observed to vary in certain genera. In the gar-pike (esox belone), for instance, they are green, and in the viviparous blenny (blenius viviparus), the sand lance (ammodytes tobanus), and two species of labrus (the l. lapina and aruginosa), they assume a green colour after boiling. The causes of these varieties in colour are unknown.

SECT. I.—OSTEOLOGY OF THE MAMMIFEROUS ANIMALS.

Plate XXXIV. fig. 1. The skeleton of the MAMMALIA bears a general resemblance to that of the human subject, in the construction, shape, and disposition of its component pieces. Distinguished, like that, into head, trunk, and extremities, we recognise the importance of the trunk, and especially of the spine, in the different classes of mammiferous animals.

The spine. The spine consists of separate vertebrae, which are conveniently distinguished, as in man, into cervical, dorsal or costal, lumbar, sacral, and coccygeal or caudal.

Cervical vertebrae. The number of cervical vertebrae is the same in animals with the longest and shortest necks,—in the horse, camel, and giraffe, and in the mole and ant-eater. They are always seven. The only exception is observed in the Al or three-toed sloth (bradypus tridactyla), which has 9 cervical vertebrae (Cuvier); and an apparent exception is presented in the dolphin and porpoise, in which the first two are consolidated into one; and in the cachalot or large-headed whale, in which the last six, sometimes the whole seven, are united or ankylosed. The last six are also united in the ant-eater and manis (Cuvier). Even in this state, however, the traces of the original separation are distinct.

In the ape the cervical vertebrae are distinguished from those of man chiefly by the spinous processes being stronger and not bifid, and in their bodies being projected more over each other before, so as to support the head more perfectly. In the ZOOPHYGA the transverse processes of the cervical vertebrae are flattened from behind backwards, and those of the atlas are very large, both for supporting the head and giving attachment to the strong muscles employed in defence, attacking prey, or bearing it off. For the same purpose the spinous process of the axis is very prominent, while the others are short and directed towards the head. In the mole and shrew the cervical vertebrae, which are void of spinous processes, are simple osseous rings, which move easily on each other, probably to facilitate the frequent motions requisite in these animals in burrowing. In the hog the cervical transverse processes are compressed and broad before, so as to appear double. In the elephant the cervical vertebrae have short single spinous processes, and the bodies projecting over each other as in the ape. In the Ruminants the length of the spinous processes diminishes as the neck is elongated. Thus they are almost wanting in the camel and giraffe, in which the arched neck is much retro-flected; and the same peculiarity is recognised in those of the horse.

From these facts it results that the length of the neck depends not on the number, but on the longitudinal extent, of the cervical vertebrae.

Dorsal vertebrae. The dorsal, thoracic, or costal vertebrae are distinguished by forming the central fixed basis of the ribs; and their number depends on that of the latter class of bones, which is very variable. The number of costal vertebrae varies from 11, which is that of the Chinese monkey, common bat, armadillo, helmet-headed dolphin (delphinus globiceps), and Gangetic dolphin, to 23, which is that of the

Unau (Bradypus didactylus). The most common number is 12, which is that not only of man, but of the ourang-outang, silky monkey (simia marikina), patas (s. patas), maimon (simia maimon), macaca (simia cynomolgus), baboon (s. sphynx), magot (s. inuus), mandrill (s. maimon), pongo (s. pongo), macauco (lemur catta), vampyre, great and horse-shoe bat, colugo (galeopithecus), shrew, hare and rabbit, agouti, flying squirrel, mouse and field rat, and camel and dromedary. The next most frequent number is 13, which is that of the mole, white bear, civet, the cat tribe (felis), the dog, wolf and fox, the didelphis tribe, the cavy, guinea pig and paca, the mouse tribe, excluding the two exceptions already mentioned, the long-tailed manis, the stag, the antelope genus, the goat, sheep, and ox, and the dolphin and porpoise. The number is 14 in the gibbon, coaita, and weeping monkey, in the howling ape (simia belzebuth), the tarsius, the brown bear, raccoon and coati, the weasel genus, the porcupine, hog, and giraffe. It is 15 in the lori, hedgehog and tenrec, in the badger, pangolin, and seal. The number is 16 in the glutton, hyena, ant-eater, American lamantin, and megalotherium. In the horse, quagga, and dugong, they are 18; in the rhinoceros 19; in the Indian elephant and tapir 20; and in the Unau or two-toed sloth 23, which, as already stated, is the greatest number yet known.

In the ape family the dorsal vertebrae resemble those of the human subject, but their spinous processes are long, and erect in the macaca and magot. In the bats, instead of spinous processes, which are wanting, there are minute tubercles. The want of these, however, in sundry species, leaves the column comparatively smooth behind. In the proper quadrupeds these processes are larger, straighter, and stronger, as the head is weighty or supported on a long neck, in order to give attachment to the strong yellow cervical ligament. This peculiarity is very distinct in the giraffe, camel, ox, rhinoceros, and elephant. In the dolphin they are straight, and smaller than those of the loins.

The lumbar vertebrae vary in number still more, perhaps, than the cervical and dorsal; and this variety may occasionally be traced to the greater or less distinctness with which the sacral and coccygeal are distinguished. The smallest number is 2, which is that of the two-toed ant-eater, ornithorhynchus, and American lamantin; and the greatest 9, which is that of the lori. The most frequent number is 7, which is that of the greater part of the monkeys, the macauco, the great bat (noctula), the hedgehog, shrew, raccoon; the tiger, panther, puma, and cat, in the feline genus; the wolf and fox in the dog; the hare and rabbit; the whole murine genus except the hamster; and in the camel and dromedary. The next number in frequency is 6, which is that of the horse-shoe bat, the colugo (galeopithecus), the white and brown bear, the coati, the weasel genus, the civet, the lion, among the feline, and the dog among the canine genus, the didelphis and cavy genera, the hamster, the stag, antelope, goat, sheep, ox, horse, and quagga. The gibbon, coaita, Al, Echidna or Ornithorhynchus hystrix of English zoologists, six-banded armadillo, and dugong, have only 3 lumbar vertebrae; the ourang-outang, pongo, and howling ape, 4; the vampyre bat 4; the hyena, armadillo, Unau, and tapir, 4; the jocko, tarsier, and common bat, 5; the badger and glutton, the porcupine, beaver, pangolin, long-tailed manis, hog, giraffe, gazelle, chamois, and seal, all 5; and the agouti and flying squirrel have 8.

In the QUADRUMANA and ZOOPHYGA generally the outer side of each posterior articular process presents an apex turned backward, so that the anterior articular process of the next vertebra is locked between two eminences, which confine its movement much. Though this apex is found

Comparative Anatomy. in the RODENTIA, it is there shorter; and the arrangement is wanting in the other tribes. The size of the transverse processes indicates the strength of the loins,—a fact which is evinced especially in the instance of the horse, porpoise, &c.

Sacral vertebrae. The number of sacral vertebrae is still more various, even in the species of the same genus. Thus, while in several of the ape genus, in the lori, in the vampyre bat, the colugo (galeopithecus), the coati, and two of the didelphis, there is one sacral vertebra only, most of the apes have sacra consisting of 2, 3, 4, 5, or 6 pieces; the majority of other animals have 3 sacral vertebrae; the hedgehog, porcupine, guinea pig, paca, hare, tiger, several of the murine genus, the ant-eater, rhinoceros, camel, dromedary, chamois, goat, sheep, and ox, have 4; the elephant has 5; the Ai 6; the Unau 7; and in the mole, white bear, and quagga, they also amount to 7. The frequency of the three sacral vertebrae in the lower animals shows that Galen, who ascribes only 3 to the human subject, must have derived this inference from the former.

These vertebrae are in the mammalia narrower than in man, and their direction forms with the spine, instead of receding backwards, a straight line; an arrangement evidently connected with the horizontal position of the former. The shape of the sacrum in the lower mammals is that of an elongated triangle; and it is further remarkable, that in those species which occasionally assume the erect attitude on the hind leg, as apes, bears, and sloths, the width of the sacrum is proportionally greater. The sacral spines, which are short in man and the ape, become longer in the ZOOPHAGA, and form a continuous ridge in the rhinoceros, most ruminants, and especially in the mole. In the vampyre bat the sacrum forms a long compressed cone, the extremity of which is united to the ischial tuberosities, without sustaining a coccyx. The seal has two sacral bones; but the CETACEA, e. g. the dolphin and porpoise, are void both of sacrum and coccyx.

Coccygeal or caudal vertebrae. The coccygeal bones constitute the tail of the lower animals, and in many instances they are extremely numerous. The smallest number is 3, which is that of the magot (simia sylvanus, pithecus, et inuus) or Barbary ape; and the greatest yet known is that of the ant-eater, in which they amount to 40, and the long-tailed manis, in which they amount to 45. Next to these may be placed that of the coaita 32, the baboon 31, the phalanger (didelphis orientalis) 30, the marmoset (didelphis murina) 29, the pangolin 28, the silky monkey (simia rosalia) and black rat 26, the weeping monkey and howling ape 25; the panther, mouse, dormouse, and elephant, 24; the lion, beaver, water-rat, Norway rat, and field-rat, 23; the flying-cat, puma, cat, dog, marmot, and rhinoceros, 22; the otter, 21; the Chinese monkey, raccoon, civet, hare, and rabbit, 20; the tiger and wolf, 19; the macauco, glutton, marten, fat dormouse, dromedary, giraffe, and quagga, 18; the tarsier, shrew, camel, and horse, 17; and other genera and species, without any determinate order, descending so low as to 9, 8, 7, 6, and 4. The quilled duckbill (echidna, ornithorhynchus hystrix) has only 12 caudal vertebrae, while the common one (ornithorhynchus paradoxus) has at least 20. The gibbon and vampyre bat are the only mammiferous animals, excepting the CETACEA, in which there are no coccygeal bones. It sometimes happens that a monkey or opossum loses a portion of its tail, when the truncated end is converted into a knotty excrescence, sometimes carious, always different from the taper point of the last coccygeal vertebra; and in this case it is difficult to determine the exact species.

In the CETACEA, in which the absence of pelvis affords no mark to distinguish the lower vertebrae into lumbar, sacral, and coccygeal, those below the dorsal may be re-

garded as lumbo-coccygeal; and their number is estimated by deducting that of the cervical and costal from the total number. The following table, which shows the number of the costal, the lumbo-coccygeal, and the total number of vertebrae, indicates that their number varies much in various genera of this family.

In general, however, if we reckon the first 2, 3, or 5 vertebrae after the costal as lumbar, it may be said that the caudal vertebrae of the CETACEA vary from 22 or 25, which are the numbers respectively in the lamantin and dugong, to 34, 38, and 42, at which they may be estimated in the dolphin. We shall see that, in the dugong at least, we are guided in this estimate by the rudimentary bones of the pelvis.

The coccygeal or caudal vertebrae of the MAMMALIA may be distinguished into two kinds; those which contain a canal in continuity with that of the vertebral column and sacrum, and those in which the pieces are solid. The first, which are next the sacrum, have articular, transverse, and spinous processes, distinct in proportion as the animals move their tails. The latter are generally prismatic in shape, diminish in size towards the extremity, and have only slight tubercles for muscular attachments. Animals with prehensile tails, as the American ape (sapajous), have above, at the base of the body of each vertebra, two small tubercles, between which pass the tendons of the flexor muscles. By means of this mechanism these animals can twine the tail round the branch of a tree with sufficient force to support the weight of the body.

The MAMMALIA with long mobile tails have often two or three small supernumerary bones placed on the lower surface of the junctions of several of the coccygeal vertebrae, from the 3d or 4th to the 7th or 8th. These sesamoid bones give attachment to muscles. In the beaver, which employs its tail as a trowel, the transverse processes are remarkable for size, while the lower spinous processes are larger than the upper ones,—an arrangement which enables it to depress the tail forcibly when it beats the ground.

The shape of the chest in the MAMMALIA varies in The chest. general as the clavicles are present or wanting. In animals provided with clavicles, as the QUADRUMANA, bats, the squirrel, beaver, mole, ant-eater, hedgehog, and sloth, the shape of the chest approaches to the human, or is conoidal, and flattened before and behind. In those void of clavicles it is compressed laterally, from the smaller incurvation of the ribs; and the sternum makes a remarkable prominence, so that the transverse or intercostal diameter is less proportionally, and the sterno-vertebral is greater proportionally, than in man and the clavicular animals. In the long-legged animals, as the giraffe and those of the stag kind, this prominence of the sternum is sufficient to give it a keel-like appearance (thorax carinatus). In the carnivorous animals the chest presents its greatest longitudinal extent.

The number and shape of the ribs varies in the different tribes. In number, indeed, the ribs always correspond with that of the costal vertebrae. Thus, in the QUADRUMANA, ZOOPHAGA, RODENTIA, EDENTATA, and RUMINANTIA, they vary from 12 to 15 pair, with only three exceptions, the glutton, hyena, and ant-eater. In the Chinese monkey, common bat, and armadillo, they are a pair less than in

Comparative Anatomy. While the quilled duckbill (echidna, ornithorhynchus hystrix) has only 15 ribs, the common duckbill (ornithorhynchus paradoxus) has 17; the horse and quagga have 18, the rhinoceros 19, the elephant and tapir 20, and in the Unau or two-toed sloth they amount to 23, which is the greatest known number. On the whole, the most prevalent number is 13. In the carnivorous animals they are narrow and dense in structure. In the herbivorous they are large, broad, and thick. In the armadillo the two first ribs are large compared with the others. In the two-toed ant-eater, which has 16 pairs, they are so broad that they are imbricated over each other like the plates of a corslet, and render the parietes of this animal's chest exceedingly solid. In the two species of duckbill (ornithorhynchus paradoxus and hystrix; echidna of Cuvier), the true ribs, in number 6, consist of two portions—a long or vertebral joined to the spine, and a short or sternal attached to the sternum. These portions are united by cartilage, so as to resemble the ribs of birds. The 9 or 10 false ribs terminate before in broad, flat, oval plates of bone, which are mutually connected by elastic ligaments.

The sternum, which is broad in the ourang and pongo, is narrow in the other species of ape, and consists of seven or eight pieces. In the vampire and all the bat family it is narrow, but presents before or below rather a prominent azygous ridge or keel (carina), and an anterior extremity, broad on the sides, like a T, for receiving the clavicles. In the mole the clavicular extremity of the sternum is continued before the ribs, and is flat on the sides for receiving the two short clavicles. In the hog the sternum is broad behind and narrow before. In the rhinoceros, horse, and elephant, it is prolonged before and flat on the sides. In the CETACEA it is broad and thin, especially before.

Cranium. Though the QUADRUMANA have 8 cranial bones, the sphenoid often consists of two portions, one forming the orbital wings and the anterior clinoid processes, the other the temporal or large wings, the posterior clinoid processes, and the basilar fossa. The two parietal bones are early united into one in the CHIROPTERA and the other ZOO-PHAGA, in which, however, the frontal remains biparted by a middle suture. The temporal tympanum is separated from the rest of the bone by a suture, which is seldom obliterated in the feline, canine, and ricerra genera. The temporal tympanum is also separate in the RODENTIA, and the frontal united. The parietal is united in some, as the hare, the porcupine, cabia, marmot, rat, and squirrel; separate in the mouse, fat dormouse, and rabbit. The frontal and parietal bones of the elephant are early united with the other cranial bones, and form a vault without trace of suture. In the hog, tapir, and hippopotamus, the two parietal bones form one piece, while the frontal bone is biparted; and though in the rhinoceros both are biparted, the frontal is early united into one portion. The sphenoid bone of the animals of this tribe long consists of two pieces, one forming the orbital wing; the other the temporal wings, which, it is to be further observed, are the smallest, in opposition to their proportional dimensions in man. In the Ruminants and SOLIDUNGULA the frontal remains long parted by its middle suture; but the two parietals are represented by a single bony vault. The tympanum is always distinct from the temporal bone. In the seal and walrus the parietal and the frontal consist of two pieces. The lamantin has only one bony arch, representing the two parietal and the squamous part of the temporal bones, while the temporal tympanum is detached from the rest of the bone. In the other CETACEA the parietal bones are at an early period united to the occipital and temporal bones, so that the five form one solid portion. The auditory or pyramidal bone is always de-

tached from the temporal, and adheres to the cranium by soft parts only. The sphenoid is not only long separate, but consists of several portions.

Though, among the QUADRUMANA, the cranium of the ourang-outang approaches that of man in shape, it differs nevertheless in the connections of the constituent bones. The temporal wing of the sphenoid bone is very narrow, does not reach the parietal, and touches the frontal only by its upper extremity, so that the temporal bone is partly articulated with the frontal. The temporal suture is not imbricated, but serrated. The same mode of connection is observed in the mandrill, macaca (s. cynocephalus), magot, and guenon (Cercopithecus), or tailed monkey tribe. In the American monkey the temporal wing of the sphenoid touches neither the frontal nor the parietal bones; but the temporal bone is articulated directly with the malar by its flat portion. In the American monkeys the frontal bone does not touch the temporal wing of the sphenoid, and the parietal is articulated to the malar. In the howling ape (simia belzebuth) the connections are as in man.

The connections of the cranial bones are in the ZOO-PHAGA the same as in man. In the RODENTIA the sphenoid is joined to the frontal and temporal, without touching the parietal; and the orbital and temporal fossæ are very small. In the armadillo, pangolin, and sloth, the connections are as in the RODENTIA; but in the ant-eater the parietal bone, continued below the cranium, is united to the sphenoid at the posterior part of the orbito-temporal fossæ.

In the elephant, though the cranial bones are at an early period consolidated into one, the auditory is always distinct from the temporal bone. In the hog, tapir, rhinoceros, and hippopotamus, the sphenoid is united to the parietal bone, and its temporal wings occupy a small space only of the orbital and temporal fossæ. The orbital wings, though larger, appear small externally. The auditory bone, though distinct, is, however, united by its base to the margin of the auditory canal of the temporal bone. The sphenoid of the ruminants is articulated, as in man, with all the cranial bones; but its orbital wing, which is extensive, is principally concealed within the cerebral cavity, and covered by the orbital part of the frontal bone. In the CETACEA generally, all the sutures which remain after early life are squamous or imbricated.

The outline of the frontal bone in the ourang-outang is more irregular than in man, and the orbital arches are less surbased. In the American monkeys its outline is triangular, and terminates in a point towards the vertex. In the others of this family (Simia), this bone is almost elliptical, and the orbital arches are nearly straight; and in the whole family these arches form, as in man, the anterior border of the frontal bone, in consequence of the narrowness of the root of the nose. In the makis it begins to widen, and the eyes become oblique,—a circumstance which gives their frontal bone a rhomboidal shape.

The frontal bone in the ZOO-PHAGA, and in all the subsequent MAMMALIA, except the CETACEA, forms an irregular prismatic or cylindrical surface with three faces—a superior, bounded before by the muzzle, behind by the cranial convexity and two lateral, descending into the orbital and temporal fossæ on each side. The hedgehog, mole, shrew, ant-eater, some of the phoca, the morse or walrus, and the rhinoceros, have no proper orbital arches; and the frontal bone, though broad behind, is contracted and nearly cylindrical between the orbits. In the hippopotamus, the ruminants, and the one-hoofed animals, it enlarges, and forms a vault over each orbit. Lastly, in the CETACEA it is narrow from before backward, resembling a fillet stretched across the cranium, but descends beneath the maxillary bones to form the floor of the orbit.

Comparative Anatomy. The parietal bones of the orang-outang differ from those of man only in their temporal margin being nearly straight. Those of the ape are narrower, and become more oblique-angled as the cranium is flattened. In the ZOO-PHAGA and EDENTATA they are almost rectangular. The single parietal of the RODENTIA is nearly quadrilateral; but it is sometimes flat, sometimes rounded, sometimes surmounted by a crest. Of the single parietal bone of the ruminants, that of the stag, most of the antelope genus, the sheep and the goat, is broad, and sends on each side a narrow process into the temporal fossa before the occipital arch; in the camel it is narrower, and bears a longitudinal crest; and in the ox and antelope hūbalus it is placed behind the occipital crest, and resembles a fillet surrounding the back of the head transversely. In the SOLIDUNGULA the single parietal is nearly quadrilateral, and placed before the occipital crest.

The occipital bone. The occipital bone in the lower mammalia is remarkable for five characters. 1. The proper occipital surface, instead of being oblique or horizontal, and inferior or basilar, becomes vertical and posterior. 2. The plane of the occipital hole forms with that of the orbits an angle constantly diminishing, becomes parallel to the orbital plane, and at length crosses it above the head. 3. The plane of the occipital condyles, instead of being transverse and horizontal, becomes oblique, and at length vertical. 4. The basilar or cuneiform process is not only horizontal, but forms with the occipital a right angle. And, 5. The mastoid process, which in man and the ape forms part of the temporal, belongs in the other mammalia to the occipital. In the polar bear, however, the mastoid process constitutes part of the temporal bone.

From the 1st, 2d, and 3d characters, it results that the head of quadrupeds is not balanced on the spine, but is suspended by muscles, tendons, and ligaments, especially the strong cervical, which connects the occipital spine to the spinous processes of the cervical and dorsal vertebrae. This ligament, therefore, though feeble and indistinct in man, is strong, particularly in quadrupeds with heavy head or long neck, in order to counteract the disadvantage of the long lever. It is strongest in the elephant, and is almost wholly ossified in the mole—a condition requisite for the burrowing faculty exercised by that animal.

The temporal bone. The temporal bone is naturally distinguished in the MAMMALIA into two parts; a flat or proper temporal, corresponding to the squamous part of the human temporal bone, and the pyramidal, acoustic, or auditory, corresponding to the pyramidal or lithoid portion of the human subject. The first only, which is proper to the skull, claims attention here. In the orang-outang and most of the genus simia it forms a trapezium with the longest side above, and the height of which varies with that of the skull. In the American apes it is smallest in this direction. In the ZOO-PHAGA the proper temporal bone is as in the ape. Being narrow in the RODENTIA behind, it is a little rounded in the short-muzzled EDENTATA, the RUMINANTIA, and PACHYDERMATA.

The ethmoid is, strictly speaking, the olfactory bone, and shall be mentioned under the organs of sense. The sphenoid, among other offices, may be regarded as the essential ophthalmic bone.

The facial bones. The facial bones of the lower MAMMALIA differ from those of man; first, in the number of separate pieces; and, secondly, in the form and proportional horizontal extent.

Intermaxillary bones. The difference in number consists in each superior maxillary bone being divided into a maxillary bone proper, and an anterior or labial portion; which being interposed between the proper maxillary, are commonly denominated the intermaxillary (ossa intermaxillaria). As it bears also the

superior incisors, it is named by Haller the incisive bone (os incisivum); but since it is found not only in the ruminants, which, excepting the camel, are void of incisors, but in the EDENTATA and CETACEA, this denomination is less applicable than the former. It may be doubted whether these should be regarded as additional bones, as they are generally represented by zootomists; for they are in truth merely the incisive or anterior portion of the superior maxillary bones. In other respects, the difference between the human and the animal superior maxillary bone is, that in the former each bone is in one piece, in the latter it is in two. Even in the human fetus the trace of the separation may be recognised; and we have seen it in the human skull some years after birth. Conversely, it is early obliterated in some quadrupeds. Thus, though distinct in the orang-outang seen by Cuvier, it was not found by Tyson or Daubenton, and is wanting in one preserved in the Hunterian museum. In a young specimen of the jocko also, noticed by Cuvier, no trace of the intermaxillary suture was observed. It appears also to be wanting in the perforated bat, the horse-shoe bat, and the three-toed sloth.

Mutually united on the mesial plane, the intermaxillary bones are united to the maxillary by sutures, which pass from the outer angles of the latter, near the incisive holes, towards the palate, where they intersect. In form and size it varies in the different orders and genera. Small in many of the ZOO-PHAGA and the walrus, it is large in the RODENTIA, in the hippopotamus, porpoise, and cachalot, and prominent in the wombat. In the duckbill it consists of two unicumiform portions, united by a broad intermediate cartilage.

The peculiarity of the animal face consists in the horizontal elongation of the two jaw-bones. In the monkey of the tribe this elongation is trifling; and all that is remarked is, that the palate and maxillary bones are more elongated in proportion to their height, and that their anterior part, instead of being vertical, is more or less inclined forwards. The degree of this elongation, which differs in different genera, may be estimated by the acuteness of the facial angle.

The narrowness of the interorbital space is another Interorbital character of the animal countenance. In the guenon and tal region. American ape it is a mere septum; but in the orang-outang, magot, and howler, it is larger, by reason of the nasal fossa ascending to this height. From these the face of the ZOO-PHAGA is distinguished by the following circumstances. 1. The breadth of the ascending maxillary processes throws the orbits to the sides; 2. these orbital surfaces form the anterior wall instead of the floor of the orbit; 3. the malar bone is united neither to the frontal nor to the sphenoid bone, and forms only the zygomatic arch and the lower margin of the orbit; 4. the orbit, closed neither behind nor below, communicates freely with the temporal fossa; and, 5. the palate bones are much elongated and form a considerable space of the internal wall, to which the ethmoid bone does not contribute. In the RODENTIA the interorbital space is still larger, by reason of the size of the intermaxillary bones throwing the maxillary backwards and to the sides, where they form the inner orbital wall, in which the palate bones occupy only a small space. The anterior wall is formed by a process of the maxillary, which contributes to form the zygomatic arch, while the malar is suspended in the middle between the process and that of the temporal bone. Very similar is the face of the elephant, except that the height of the alveoli from the tusks, thrusting the nose upwards, and shortening its bones, alters entirely the expression of the head of this animal.

In the sloth, in which the face is short in proportion to

the skull, the malar bone attached to the maxillary only, is not united to the zygomatic process of the temporal. In the long-muzzled EDENTATA, in which the face is conoidal, the maxillary bones extend to the orbits, and are separated by a broad lacrimal bone, while a long palate bone forms the inner wall of these fossæ. The zygomatic arch, which is interrupted in the ant-eater and pangolin, is completed in the Cape ant-eater and the armadillo. In the tapir and rhinoceros the maxillary bone passes beneath the orbit; and the nasal bones form a sort of vault, which supports in the first animal the trunk, and in the second the horn.

In the CETACEA the maxillary and intermaxillary bones form a sort of flattened beak, distinguished into four parallel bands, of which the maxillary, which are external, alone bear teeth in three genera, provided with the latter organs. The nasal fossa is a vertical opening before the cranium, surrounded before and laterally by the intermaxillary bones. The maxillary ascend in the same manner, and cover that part of the frontal bone which forms the orbital vault, but do not themselves contribute to the formation of this cavity. The nasal bones are two minute tubercles implanted on the frontal bone above the narrow aperture. The malar is in the shape of a style, suspended by cartilages beneath the orbit; and the latter cavity is completed behind by a process of the frontal, which joins the zygomatic of the temporal bone, and below which the orbital and temporal fossæ communicate.

The direction of the orbits, the shape of their base or facial border, and their relation to the temporal fossæ, are important circumstances in the animal face and cranium. In the simia the angle of the orbital axes is rather smaller; and the shape of the margin, which is quadrilateral in the jocko, becomes oval in the orang-outang and American monkeys. The angle of the axes enlarges in the other Mammalia; and the base or anterior margin becomes nearly circular in the ZOOPHAGA, RODENTIA, EDENTATA, and PACHYDERMATA; but the arch is incomplete behind. In the Ruminants and SOLIDIPEDA, however, in which it is also circular, the border is complete. In the CETACEA the orbital vault is semicircular, their axes are rectilinear, and there is no floor.

In the human skull the junction of the malar bone with the frontal and sphenoid completes the orbit externally, and prevents it from communicating with the temporal fossæ; and the same arrangement is observed in the simia. In the CARNIVORA, RODENTIA, EDENTATA, and PACHYDERMATA, however, in which the malar bone is united neither to the frontal nor the sphenoid, the orbit is not only incomplete on the external posterior border, but communicates freely with the temporal fossæ. In the Ruminants is observed an arrangement intermediate between that of the QUADRUMANA and that of the CARNIVORA. The malar bone, united to the frontal, completes the orbital ring; but as it is not united to the sphenoid, it allows the orbits and temporal fossæ to communicate. The orbit of the mole is so superficial, that it can scarcely be said to exist.

The lower jaw of the mammiferous quadruped differs from that of man chiefly by the following circumstances. The triangular flat surface which constitutes the chin, and which is most distinct in the Caucasian race, begins to become faint in the negro, and is altogether lost in the monkey tribe. In the orang-outang, indeed, the animal character of the lower jaw appears distinct in the vertical convexity of the anterior arch of the jaw, and the retreating of its lower margin. In the lower QUADRUMANA the anterior maxillary arch is still more retreating, and the maxillary rami form a more acute and elongated angle. These animal characters are still more conspicuous in the CAR-

NIVORA, most of the PACHYDERMATA, Ruminants, SOLIDIPEDA, and RODENTIA. The ascending ramus also becomes short in the PACHYDERMATA and several of the CETACEA, more so in the ZOOPHAGA, and is almost extinct in several of the RODENTIA, for instance the paca, beaver, and porcupine, and the armadillo, ant-eater, and duck-bill, among the EDENTATA. In the ZOOPHAGA, however, in which the prehensile and masticatory muscles are large and powerful, the ramus becomes broad, and its coronoid process is extensive. The angle which the ramus forms with the body of the jaw, and which is almost right in the adult human subject, becomes obtuse in the lower animals, nearly at the same rate at which the ramus disappears; and indeed the transition of the angle into a straight line implies the disappearance of the ramus. This, therefore, is the character in the EDENTATA and CETACEA, in which there is neither ramus nor coronoid process, after these parts have been seen for the last time in the amphibious MAMMALIA.

When the mammiferous cranium is considered generally, and the relative direction and proportion of the cranial and facial part of the head examined, we recognise more distinctly the characters by which the lower orders of that class are distinguished from man. This character consists in the position of the occipital bone and hole, the position and direction of the facial bones in relation to the frontal, the elongation of the former, and large size which they present in relation to the cranial.

In the human subject, it has been already observed, the position of the occipital bone is oblique and horizontal, and the plane of the occipital hole is horizontal, while its position is anterior. In most quadrupeds, while the bone assumes a vertical position, the hole becomes posterior, and its plane vertical or oblique, in proportion as the face is elongated. The plane of the occipital hole forms with that of the horizontal a considerable angle, which Daubenton undertook to determine, by drawing one line through the plane of the aperture, and another from its posterior margin through the lower edge of the orbit. (Mém. de l'Acad. des Sciences de Paris, 1764, p. 568.) In the horse this angle is about 90°, while in the orang-outang it is only 37°, and in the lemur 47°. In other respects, however, it furnishes an imperfect result, since in most quadrupeds which differ very much it ranges between 80° and 90°.

The direction of the face in relation to that of the Camperian cranium, determined according to the method of Camper, line and furnishes more accurate results. While in the human subject it varies, according to the races, from 70° to 80°, in the orang-outang it is only 65°; in the American and long-tailed monkeys about 60°; in the macaca and baboon about 45°; and, lastly, in the mandrillo, the most vicious and ferocious of the monkey tribe, only 30°. In some species in which the ear is elevated and the guttural fossa deep, for instance in the pongo and alouate or howler, the small size of this angle does not indicate proportional elongation of muzzle; and to rectify this inconvenience, it is requisite to draw the basilar line of the facial angle parallel to the base of the nostrils. With this modification, however, the Camperian line admits of correct application to the human race and QUADRUMANA only, in which the frontal sinuses are small and not prominent. In quadrupeds, for instance the carnivora, several of the ruminants, and in the elephant, the frontal sinuses are so large and prominent as to affect the results given by the facial angle very materially. In other orders, again, for instance the RODENTIA and the morse, the nose occupies so much space that the cranium is inclined backwards without its walls being free before; and it is impossible to know where the facial line passes. These, therefore, must be measured by the inner surface. Lastly,

Comparative Anatomy. in the CETACEA the pyramidal elevation of the cranium above an elongated but flattened face renders the facial line much more vertical than it ought to be. The following measurements show the angle subtended by a line drawn parallel to the base of the nostrils, and another passing by the anterior margin of the alveoli, and touching the convexity of the cranium, whether the point of contact be concealed by the face or not.

Method of Cuvier. Aware of the imperfect results obtained in this method of measurement, Cuvier proposes to estimate the relative proportion of the cranial and facial part of the head, by comparing the respective areas exhibited by a longitudinal and vertical section of both. In this section the area of the cranium is to that of the face sometimes in a ratio of majority, sometimes of minority, occasionally of equality.

In the European the area of the cranial section is about four times that of the face, excluding that of the lower jaw. In the negro the cranial area remaining the same, that of the face augments about one fifth, whereas in the Calmuck it augments only one tenth. In the orang-outang the proportion is still less. In the American ape the facial is almost half the cranial area. The ratio is that of equality in the mandrill and most of the CARNIVORA, except the varieties of short-muzzled dogs, as the pug, in which the facial is rather less than the cranial area. In the RODENTIA, PACHYDERMATA, Ruminants, and SOLIDUNGULA, the facial is larger than the cranial area. Among the RODENTIA, in the hare and marmot, it is a third larger; in the hedgehog double; in the Ruminants almost double; in the pig a little more than double; nearly triple in the hippopotamus; and almost four times in the horse. In the mose and elephant the face is rendered large by the height of the alveoli; and it may be regarded as augmenting the organs of the senses. The cranium of the CETACEA is very convex, and the face very flat, and the proportional area of the latter is thereby diminished. The facial area of the dolphin may be a third larger than the cranial.

The outline of the cranial section in the human subject is oval, that of the facial section forms a triangle, with the longest side contiguous to the cranium, and the smallest without, while the angle formed by the latter with the third side or palate is the facial. In the lower animals, this triangle, which may be named the facial, becomes so much elongated, that the cranial side, which is the longest in man, becomes the shortest of the three in the cynocephalus and mandrill, and continues so in the other quadrupeds.

The pelvis. The basin (pelvis) of the MAMMALIA in general agrees with that of man in forming a part common to the trunk and the lower extremities. It differs, however, altogether in the direction which it takes, which is obliquely backwards, with its anterior opening or brim forwards and downwards, and in the bones being smaller and much narrower; both of which characters are connected with its not being used in the lower animals to oppose the gravitating weight of the abdominal and pelvic viscera. In the apes, most similar to man, the coxal bones are much elongated; and in the ZOOPHAGA, their superior or ilial

part is not much broader than their pubal. In the Ruminants, the PACHYDERMATA, and SOLIDUNGULA, the ilial portions again become broad; and in the elephant and rhinoceros especially, this width, combined with the length of the ilio-pubal rami and the concave abdominal surface, concurs to give the pelvis of these animals the prodigious capacity by which they are distinguished. The pelvis of the amphibious MAMMALIA differs from that of the ZOOPHAGA only in being narrow and elongated, and in the pubis being thrown behind. In the ant-eater, mole, and shrew, the pubal junction is open as in birds; and in the two latter genera, the bones are so narrow that the sexual and urinary organs are placed without its circumference. In the sloth genera, e. g. the Ai and Unau, the pelvis is wide, the cotyloid cavities turned upwards, and the ischial tuberosity is united with the sacrum, so as to convert the ischiatic notch into a hole. This sacro-ischial junction is also observed in others of the EDENTATA, as the ant-eater and armadillo, and in the phascocoloma (didelphis ursina). The marsupial animals present two minute bones, one on each side, connected by movable articulation to the pubis, and which are employed to sustain the mammillary pouch (marsupium) or nipple-bag, in which the young are reared after exclusion from the uterus. These bones, which are distinguished by the name of marsupial (ossa marsupialia), are oblong and flattened. Lastly, in the CETACEA, in which the pelvic extremities are wanting (Plate XXXIV. fig. 2), the pelvis is also so far deficient, that instead of consisting of the sacrum and coxal bones, it is represented by a small bony appendage, suspended in the soft parts on each side of the anus, and which, meeting at angles on the mesial plane so as to form a bone like the letter V, are merely a rudiment of the ischial bones.

The shoulder of the MAMMALIA generally differs from that of man by the absence or the proportions of the collar-bone, and by the shape of the scapula.

The collar-bone exists in the QUADRUMANA nearly as The collar-bone. in man; but it is wholly wanting in the three orders of UN-GULATA—PACHYDERMATA, the Ruminants, and the SOLIDUNGULA—and in all the CETACEA. Between these extremes it is found in various forms in the intermediate orders of the UNGUICULATA. Among the ZOOPHAGA it exists in the CHIROPTERA, especially the bats proper, in which it is strong and large; the INSECTIVORA, as the hedgehog, shrew, and mole genera; and in the didelphis or opossum among the marsupial animals. In the others of this order, as the bear, raccoon, coati, weasel, otter, dog, cat, and seal, the collar-bone is represented by clavicular bones, suspended among the muscles, touching neither the sternum nor acromion; and in some species it is altogether wanting. The RODENTIA may be distinguished into two subdivisions, as they are provided with or void of collar-bones. The first comprehends the squirrel, beaver, and mouse, with the helamys, marmot, and aye-aye or cheiromys genera. The second consists of the porcupine, hare, and cavy genera, as the guinea-pig and paca, in which the collar-bone is rudimental only, suspended among the soft parts. It exists in most of the EDENTATA, as the sloth, armadillo, ant-eater, and the gigantic fossil animal named megatherium; but it is wanting in the pangolin, and was believed to be wanting in the echidna and duckbill till Cuvier demonstrated its existence. In the CETACEA there is no vestige of collar-bone.

From these facts it results that the clavicle exists in all animals, the fore legs of which are frequently or habitually protruded, either to seize, as apes and the RODENTIA; or to fly, as bats; or to dig and burrow, as the mole; or to rake the ground, as the hedgehog and ant-eater. In the mole particularly, the collar-bone, instead of being long, is

Comparative Anatomy. a broad, thick, short, quadrangular bone: and while it is connected to the acromion by a ligament, it is articulated to the humerus by a large facet. The collar-bone of the echidna and ornithorhynchus is very singular. It consists of a broad central bone, surmounted by two transverse branches spreading out on each side, so as to give the whole bone some resemblance to the letter T, but sinuated so as to make the diverging branches like the Greek $\backslash gamma$. In young animals this bone consists of three portions. The two diverging branches are genuine collar-bones, and may be regarded as a bifurcated bone; while the middle is supported on the sternum, and has articulated to each side a part of the scapula, corresponding to the coracoid process. The collar-bone, indeed, is a powerful buttress, which prevents the arm-bone from being thrust too much forward.

The shoulder-blade. Of the shoulder-blade or scapula, which is present in all red-blooded animals with thoracic extremities, and hence in all the MAMMALIA, the principal point is to remark the varieties which its shape presents. Though in man, most of the QUADRUMANA, the CHIROPTERA, and the elephant, the vertebral margin or base of several authors on human anatomy is the longest, it becomes the shortest in most quadrupeds, especially those which, like the Ruminants and the SOLIDUNGULA, have long legs and narrow chest. In most of them, also, this margin, instead of being straight, is rounded, as in the CARNIVORA and RODENTIA. In the CARNIVORA without collar-bone, the hedgehog and didelphis, the acromion is less prominent; there is another eminence directed backwards, almost perpendicular to the spine. The coracoid process, also, which is present in the CHIROPTERA, the hedgehog, and didelphis genus, is wanting in most of the zoophagous tribe. In the hare the acromion terminates in a long slender process, rising at right angles and bending backwards, which may be named the recurrent. In the Ruminants and SOLIDUNGULA, not only are this and the acromion, but even the coracoid, wanting. The scapula, again, of the hog and rhinoceros is remarkable for the disappearance of the spine at the glenoid angle; while from its middle proceeds a prominent process towards the costal or inferior margin. In the mole the scapula is long and narrow, like a cylindrical bone, placed parallel to the spine,—an arrangement which, together with the shortness and thickness of the clavicle, already mentioned, is evidently connected with the burrowing habits of this animal. Lastly, in the echidna and ornithorhynchus, which in so many characters of organization approach the AMPHIBIA on the one side, and the BIRDS on the other, the scapula is a single sinuous bone, attached by one extremity to the sternum and middle part of the clavicular bone, with the other loose; and in the middle an articular cavity, in which the head of the humerus is placed, and which evidently corresponds to the glenoid. In this instance, therefore, the clavicle and scapula may be regarded as united into a single bone.

The humerus. The humerus, which exists in all the animals with thoracic extremities, undergoes considerable variations. In the lower animals generally it is much shorter than in man; and it is invariably shorter in proportion as the metacarpus is elongated. Thus, in animals with what is named a cannon bone, that is, one metacarpal, as in the horse and the ruminants, the humerus is so short that it is concealed in the soft parts as far as the cubit. In the CETACEA it may be said to attain its maximum of brevity. In the bat and sloth it is long in proportion to the rest of the body.

The humerus of the mole is perhaps the most extraordinary of all those of the mammiferous animals. Not only is it articulated with the scapula by a small head, but it is connected with a facet of the clavicle by another belonging to the great tuberosity, and between which and

the head is a deep pit. The crest of the small tuberosity is so large, that it represents a square placed vertically, so that the linea aspera is above. The rest of the body of the bone, which is very short, is arched above, so that the cubital extremity is directed upwards. From this arrangement it results that the cubit is elevated above the shoulder while the palm is turned downwards,—a disposition necessary for the burrowing habits of the animal.

In the simia the radius and ulna are arranged as in man, except that in the eynocephalus, mandrill, magot, and ulna, guenon, the coronoid process of the ulna is narrower, and the radial facet deeper. In the other MAMMALIA the ulna very generally disappears or becomes rudimentary only. In the bat family and the colugo (galeopithecus) the ulna is wanting or is represented by a slender style placed below the radius. These animals are therefore destitute of the power of pronation and supination. In the ZOOPHAGA, the radius and ulna, though separate, are void of rotatory motion; and the olecranon is compressed, and continued farther back than in man. In the PACHYDERMATA the radius is before and the ulna behind, and, though distinct, there is no rotation. In several of the RODENTIA, for instance the marmot, porcupine, &c. the coronoid process is small, and in others, e. g. the cavy, hare, and mouse family, it is altogether effaced. In the Ruminants the ulna is united immovably to the radius; and in the SOLIDUNGULA it is represented by an olecranon adhering to the posterior surface of that bone. In the CETACEA, though both bones are present, they are much flattened.

The carpus of the ape genus contains one bone more than that of man, situate between the pyramidal and large bone, and which seems to result from the trapezium being divided into two parts. Conversely, in the ZOOPHAGA, but especially in the dog, cat, hedgehog, shrew, bear, and seal, the scaphoid and semilunar are united into one large bone. Those which have a vestige of thumb, as the hyena and glutton, have the trapezium very small. The mole has not only 9 carpal bones, as the ape, but a large sickle-shaped bone, which is attached to the radial margin of the fore paw, and which gives it the shape proper to the habits of the animal. The toes are further very short. Of the RODENTIA, the hare resembles the ape; but in the beaver, marmot, squirrel, and cat, the scaphoid and semilunar make one bone; while in the porcupine the supernumerary bone is between the pisiform and metacarpal of the fifth toe. In the two-toed ant-eater there are only 6 carpal bones, 4 in the first row and two in the second; in the three-toed sloth there are only 5, 3 in the first row and 2 in the second; the pangolin has 7; the cache-came has 8 and a rudimental small toe; the elephant 8, 7 wedge-shaped and one elongated, corresponding to the pisiform; and the other PACHYDERMATA 8. In the rhinoceros, which has only 3 toes, the trapezium only is wanting; but there is a supernumerary bone on the margin of the scaphoid, and on that of the unciform, as in the porcupine. The first range consists, in the Ruminants and SOLIDUNGULA, of 4 bones; in the former, excepting the camel, the second consists of 2, and the latter of 3. Those of the CETACEA, which are much flattened, are 3 in the first row and 2 in the second.

The MAMMALIA generally have as many metacarpal bones as toes, that is, never fewer than 3 or more than 5, with the exception of the Ruminants, in which these bones are in early life consolidated into one named the cannon bone. In animals which walk on the tips of the toes, or which use them as organs of prehension, the metacarpal bones are lengthened to nearly double; and hence in all these animals the metacarpus is erroneously named the fore leg, and therefore it has been imagined, that in several

Comparative Anatomy. of our domestic animals the different parts of the lower extremity are articulated in opposite directions to those of man. Thus the fore leg of the horse, deer, sheep, and dog are in truth the metacarpus of these animals; and what is vulgarly named the fore knee or cannon bone of the horse, is actually the carpus or wrist-joint. It is therefore convex on the dorsal, and concave and inflected on the palmar aspect, exactly as the carpus of the human subject.

In the three-toed sloth, the three bones of which the metacarpus consists are mutually consolidated at the base and with the rudiment of a fourth toe. In the CETACEA, the metacarpal bones, which are much flattened, are also mutually united.

Fore toes. In the MAMMALIA generally, if we include imperfect or rudimental phalanges concealed in the skin, there are never fewer than 3, nor more than 5. The Unguiculated animals generally have 5, perfect and imperfect. The character of the perfect fore toe or finger is to consist of 3 rows or phalanges, excepting the first of the radial side, which has only 2. In the QUADRUMANA this is separate, and opposable to the other toes, constituting a thumb, and giving this tribe of animals a prehensile organ entitled to the epithet of hand. It is, however, shorter and less perfect in other respects than the genuine thumb of the human hand. In the coaila (simia paniscus) it is converted into a rudimental bone, concealed under the skin.

In the ZOOPHAGA, which have no power of grasping minute objects, the thumb or first toe is parallel to the others, and, though equal in length to these in the ursine family, it is shorter in the mustela, viverra, canine and feline genera. In the latter, which have the power of erecting the claws, to prevent them from being blunted in walking, the shape of the middle and ungual phalanx is remarkable. The former is triangular prismatic, with two lateral and a plantar or palmar inferior surface. The third or ungual phalanx is shaped like a hook, consisting of two parts. One, directed forwards, sharp and pointed, receives the nail or claw, in a long groove like a sheath. The second part of the hook, which is behind, rises vertically from the lower part by which it is articulated, and is produced into two processes, to which are attached the erecting muscles of the claw, which are flexors of the phalanx.

Among the RODENTIA there is a perfect but short thumb in the hare, beaver, and jerbois; a two-phalanx but concealed one in the squirrel, mouse, and rat family, porcupine, paca, agouti; and a one-phalanx concealed one in the cavy, guinea-pig, marmot, &c. In the EDENTATA the number of fore toes varies much; in the Tamanoir, and Tamandua or four-toed ant-eater, the thumb-toe is obliterated; in the Ai or three-toed sloth, both that and the fifth toe are obliterated; and in the two-toed ant-eater, and Unau or two-toed sloth, these, with the second toe also, are obliterated.

The elephant has 5 perfect toes, all concealed under the thick, callous hide of the foot. In hoofed animals with 4 toes, for instance the hog, tapir, and hippopotamus, the thumb-toe is in the shape of a small rudimental bone.

In the Ruminants the single metacarpal bone (Cheselden's figure of the Deer, Plate I.) is articulated with two digital phalanges, which constitute one of the distinguishing characters of this order—the cloven foot. In some genera, at the root of these two perfect toes are two small bones, often covered with horn, which represent two other toes. The last or ungual phalanx is always trilateral in shape. In the horse and the SOLIDUNGULA generally, the two lateral toes are represented only by two bony styles, named the splint bones, situate on the two sides of the metacarpal or cannon bone. The three phalanges of

the single toe which constitutes the foot are distinguished as the pastern bone, which is the first phalanx; the coronet, which is the middle or second; and the coffin bone, which is the third or ungual phalanx, which has the shape of the hoof, rounded before, convex above, and flat below. To the back of the pastern joint are connected two sesamoid bones; and to the coffin bone is attached another, named the shuttle bone. In the CETACEA, all the phalanges, which are flattened, and often cartilaginous, are united in the fin or paddle.

The thigh-bone, which is single in all the classes, follows the type of that of the human frame in general figure and parts. In the MAMMALIA it is, however, proportionally shorter, and its length diminishes as that of the metatarsus augments. In the Ruminants and SOLIDUNGULA, for instance, it is so short that it is concealed by muscles against the belly; and hence it is too often overlooked and confounded with the leg. In other respects the general characters are, that it is not arched; that, excepting in the bear and some of the simia genus, e. g. the ourang-outang, it is shorter than the leg-bones; that its neck is very short, and more perpendicular to the axis of the diaphysis than in man; and that the great trochanter is raised above the head, which is directed inwards. In the simia it is quite cylindrical, and void of linea aspera. In the tapir the middle part is found flattened; and at the external margin there is a prominent crest, terminating in an unicum process. In the rhinoceros the great trochanter and the unicum process are so elongated as to unite almost, and form a hole between them and the diaphysis. The unicum process is observed also in the horse, beaver, and armadillo. The thigh-bone of the seal is so short, that the half of its length consists of the two articular extremities.

Though the leg-bones of the MAMMALIA bear a general similitude to those of man, the tibia alone is constant; and the fibula, after becoming unusually slender, and changing its position from the outside to the posterior part of the tibia, is converted into a mere appendage, and at length disappears entirely. Thus, though it is distinct, and occupies its usual position in the simia, in the CHIROPTERA it is extremely slender; and since the femora are directed backward, the fibulae are turned towards each other. In several of the EDENTATA, for instance the phagrin, armadillo, and sloth, it is large, curved, and remote from the tibia. In the dog family and the RODENTIA it is altogether behind the tibia. In the mole and murine genus it is consolidated to the lower third of the tibia, leaving an empty trilateral space above. In the rhinoceros, elephant, and hog, the fibula is flattened and united to the whole length of the tibia. In the ruminants it is represented by a small bony appendage, placed on the outer margin of the astragalus, below the tibia, and forming the external or fibular ankle. Lastly, in the horse and SOLIDUNGULA, the fibula is reduced to a styloid rudimental process, which is firmly consolidated in the adult animal to the upper part of the tibia.

Between the tarsus of man and that of the other MAMMALIA the following are the principal differences.

In the simia the fibular facet of the astragalus is nearly vertical, and the tibial is very oblique; and the calcaneum wants the tuberosity, except in the pongo. In the ordinary bat family the calcaneum is elongated into a styloid process, concealed in the substance of the membranous ubiform expansions; but in the roussette (pteropus) the tuberosity projects beneath the foot.

In the RODENTIA the calcaneum is produced considerably backwards, while the scaphoid, which consists of two parts, forms a tubercle on the sole. Among the EDENTATA the three-toed sloth is peculiar in having a tarsus,

Comparative Anatomy. consisting of four bones only, the astragalus, calcaneum, and two cuneiform bones, the first of which is articulated not only with the tibia, fibula, and calcaneum, but with the large cuneiform bone, without any intermediate scaphoid bone. Its connection with the tibia is by means of a convex articular surface, which rolls on the external part of the tarsal end of the tibia. From this mode of articulation it results that the foot of the sloth admits neither of being elevated nor depressed, but simply of performing lateral motions of adduction and abduction, to which it owes the power of clasping the trunks of trees and climbing, but which renders progression difficult and laborious.

The hog has a scaphoid with three ordinary cuneiform bones, and a rudimentary great-toe bone beneath the first. In the tapir and rhinoceros there are only two cuneiform bones. All the animals already enumerated have the same number of metatarsal bones as of toes.

In the Ruminants the cuboid and scaphoid bones are united, unless in the camel, in which they are distinct. At the outer margin of the pulley of the astragalus is a bone which represents the lower head of the fibula, and which is farther articulated to the upper surface of the os calcis. In this side also there are only two cuneiform bones, which are united in the giraffe. The two metatarsal bones are always united, as in the metacarpus, into one, which forms a posterior cannon bone. The SOLIDUNGULA resemble the camel in this, that the scaphoid is distinct from the cuboid bone, and that there are two cuneiform bones, while the peroneal rudiment and the corresponding articular surface of the calcaneum are wanting. The metacarpal are also consolidated into a single piece, named the hinder cannon bone, each side of which is provided with a minute bony style.

The toes of the QUADRUMANA and the MARSUPIALIA are longer than those of man; but the great toe is shorter than the others, and its metatarsal bone is susceptible of separation and opposition, as the thumb or thumb-toe of the hand. Hence Cuvier, in his first classification, distinguished the latter by the name of Pedimana. The Aie-ai among the RODENTIA appears to possess the same faculty. Among the ZOOPHAGA the great toe remains always conjoined with and parallel to the others; and in the canine and feline genera it is obliterated. Among the RODENTIA, that of the beaver is nearly equal to the other toes; those of the marmot, porcupine, and the murine genus, are shorter; in the paca it is almost obliterated; it is reduced to a single bone in the Cape gerboa; and the leporine genus have no trace of it. In the cavy, agouti, and guinea-pig, the great and small toes are each reduced to one bone. In the gerboa (mutus jaculus) and alactaga (mus sagitta) the three middle metatarsal bones are united into a single one similar to the cannon bone of the Ruminants and SOLIDUNGULA; and while the two lateral toes are distinct, though short, in the former animal, they are obliterated altogether in the latter.

Among the EDENTATA, the ant-eater, orycteropus, pangolin, and armadillo, have five toes, of which the great is the shortest in all. In the sloth the great and small toe are reduced to one small bone. The other metatarsal bones are united at their base. The toes have only two phalanges, of which the ungual is the largest.

In the subsequent families the metatarsal bones deserve particular attention. In the elephant and PACHYDERMATA, their tarsal extremity has a flat surface, and the phalangeal consists of a convex tubercle, which presents below a prominent line in the middle of the bone. In the SOLIDUNGULA this line is above and below both. In the Ruminants, in which the cannon bone consists of the two metatarsal bones, the line of union is represented by a deep line like the tract of a saw. The elephant has 5

perfect toes; the hog 4; the tapir and rhinoceros 3; the Ruminants have two perfect toes on one metatarsal bone, and two small ones attached behind its base. The SOLIDUNGULA have one perfect toe, and two imperfect, which are reduced to a single styloid bone. In these animals the body is supported in walking by the last or ungual phalanx alone; and hence the term foot is not of the same import as in the human subject and animals similarly constructed. While indeed man supports his person in progression on the os calcis and the posterior or metatarsal phalanges, in the other mammiferous animals the former bone touches not the ground, but is always elevated above it a considerable height. All the zoophagous or unguiculated animals, excepting the plantigrade, support themselves chiefly on the ungual and middle phalanges both of the fore and hind foot; and neither the posterior phalanges nor the calcaneum touch the ground, as is easily demonstrated on observing the gait of the hedgehog, dog, fox, cat, or similar individuals of the same family. The animals distinguished by the name of PLANTIGRADE are believed to support themselves on the entire foot. But though the foot is certainly spread on the ground more freely than in those already mentioned, by the bear, glutton, badger, and others, it appears that not the heel, but the metatarsus, is allowed to touch the ground in progression. In the Ruminants and SOLIDUNGULA, as already mentioned, the only part of the foot which is applied to the ground is the ungual phalanx; and it is well known that the horse supports himself on the plantar surface of the coffin bone only.

Lastly, in the Amphibious Mammals, while the extreme brevity of the humerus and femur unfit them for progression on land without extreme awkwardness and difficulty, the expanded shape and oblique position of the metacarpal bones and phalanges, the length of the tibia and fibula, and the greater length of the first and last than the middle metatarsal phalanges, all concur to give these animals great facility in swimming. (Cuvier, Ossemens Fossiles, tome v. partie i. septieme partie.) In the CETACEA, again, while the total want of pelvic extremities renders motion on land quite impracticable, the fin-like disposition of the metacarpus and metacarpal phalanges, with the great strength of the lumbar, and the length of the coccygeal vertebrae, peculiarly qualify them for locomotion in the waters.

The number of vertebrae of which the different regions of the spine consist, is not less variable in BIRDS than in the MAMMALIA. Some idea of these variations may be formed from the number exhibited in the following table by Cuvier.

In this table the most remarkable circumstance is the great number of cervical vertebrae, which are much more numerous than in the Mammalia. They vary from 9, the number in the sparrow, to 23, which is that of the cervical vertebrae of the swan. The most common number is 11, which is that of 10 genera. The next most frequent is 12, 13, and 14, which are equally the numbers of 9 genera. The next is 15, which is that of 8; 10 occurs in 6, 18 in 4, and 16 in 3. In the stork and crane they are 19.

The next remarkable circumstance is, that the dorsal or costal vertebrae are greatly fewer than in the Mammalia, never exceeding the number of 11, and being more frequently about 7 or 8. Thus, while they are 11 in the cassowary, swan, and sheldrake, 10 in the goose, bernacle, and grebe, and 9 in the sparrow, lark, humming-bird, parrot, crane, avocet, oyster-catcher, cormorant, and black-diver, they are 7 or 8 in all the other genera, and only 6 in the bullfinch.

There are no lumbar vertebrae strictly so named, for those which extend from the chest to the tail are consolidated into one piece with the iliac bones. The tail, which is short, consists of from 7 to 9 vertebrae.

The part most variable in proportional length is the neck. It is so much longer as the feet are elevated, ex-

cept in some of the swimmers, in which it is greatly longer, because they require to seek their food below the surface of the waters on which they float.

The bodies of the cervical vertebrae are articulated not by plane facettes, which would admit obscure motion only, but by portions of cylinders, which allow extensive motion. The 3d, 4th, or 5th superior vertebrae allow of anterior inflection only, and the others of posterior inflection. This gives the necks of birds an alternate serpentine inflection; and it is by rendering the two arches, of which this curvature consists, straight or convex, that the animal elongates or shortens his neck. The articular processes of the superior vertebrae are directed upwards and downwards; those of the lower are turned anteriorly and posteriorly.

Instead of transverse processes, the cervical vertebrae of birds are provided with a tubercle above, and the anterior extremity of which terminates in a narrow style, descending parallel to the body of the vertebra.

Only the most superior and inferior vertebrae have distinct spinous processes, and these have anterior as well as posterior ones. The middle ones have before two crests, which form a half-canal, and behind a tubercle, often bifid, or, when they are elongated, two rough lines. The atlas, which is articulated with the occipital bone by a single facet, has the shape of a minute ring.

As the neck of birds is movable, the back is fixed. The spinous processes of these vertebrae are in mutual contact, and they are connected by strong ligaments. Most of these processes are generally consolidated into a single continuous crest, extending along the whole back. The extremities of the transverse processes terminate in two apices, one directed forwards, the other backwards; and occasionally they are consolidated into a continuous mass like the spinous. That this arrangement is requisite for the trunk to remain fixed during the violent motions which take place in flying, is rendered probable by the fact, that in birds which do not fly, as the ostrich and cassowary, the spinal column retains its mobility.

The last dorsal vertebrae are often placed on the crest of the iliac bones, and they are then united, as the lumbar, on the large piece of the iliac bones, from which it results that the number of vertebrae can often be estimated in no other mode than by that of the holes of the nerves which issue from the chord.

The caudal vertebrae are most numerous in the species Caudal which move the tail with most energy; for instance, the vertebrae magpie and swallow. They have spinous processes below as well as above, and very long transverse processes. The last of all, to which the pinous are attached, is longest, and has the shape of a ploughshare or a compressed quoit. In the cassowary, which has no visible tail, the last bone is conical; in the peacock, on the contrary it has the shape of an oval plate, situate horizontally.

It was early observed by the original zoologist and tra- veller Pierre Belon, that the crania of birds were void of sutures; and that in a few only were these lines of distinction into separate bones recognised. The explanation of this peculiarity is found in the history of the ossification of the head in young birds, which shows that the cranium consists at that period of separate bones, corresponding in number and situation to those of quadrupeds. Thus, there are two frontal bones, which are continued forwards to form the vault of the orbits; two small parietal bones behind the frontal; a temporal bone on each side of the skull; a sphenoid united to the occipital, even in subjects in which the other sutures are distinct; or a sphenio-occipital bone, which is early united with the temporal.

These sutures, however, are distinctly seen only in

Comparative Anatomy. young birds and those recently hatched; for the bones are very early united, and in the adult bird the cranial sutures are invariably obliterated. Thus, in the domestic fowl and turkey the skull is one piece; and the only trace of suture that remains is a linear depression in the middle of the frontal bone, indicating the original formation in two halves. In the recently hatched bird, also, the sphenoid is separated from the occipital bone by a transverse suture, extending from the one ear to the other. The occipital bone is at the same time a ring, consisting of four parts; a superior, two lateral, and an inferior which is small. The sphenoid, which forms the greater part of the base of the cranium, is nearly trilateral, with a small anterior process, to which the palatine arches are articulated. It has no pterygoid processes, and does not touch the posterior aperture of the nostrils. The temporal bone, though void of zygomatic process, has a pointed style, which contributes to form the posterior margin of the orbit. The frontal bone, after covering part of the cranium, is continued forwards in a broad, thin plate, which forms the vault of the orbits, while these cavities are separated by a thin vertical bony plate which descends at right angles from the frontal bone, and is connected behind with the sphenoid. The long eminences observed on the heads of the cassowary, curlew, pintado, and some species of hocco, are produced from this supra-orbital part of the frontal bone; and their interior, which consists of loose diploe, communicates with that of the same bone.

Facial bones. The face in birds is rarely so firmly consolidated as the cranium. It is composed of two lacrimal bones, forming the anterior margins of the orbits, and united on the mesial plane; two nasal bones anterior to the lacrimal; two bones corresponding to the superior maxillary, and forming the external lateral parts of the upper half of the bill; two inter-maxillary bones; two anterior palate bones, corresponding to those of the MAMMALIA; two posterior palate bones corresponding to the pterygoid processes of the sphenoid; and the lower jaw a paraboloid bone, consisting of two rami united before, where they are covered by the horn of the lower half of the bill. Besides these, there is in

The quadrilateral bone. the whole class an irregular-shaped bone, common to the cranium and lower jaw, and connecting these two together. This bone, which has been rather improperly named the square, quadrangular, or quadrilateral bone (os quadratum), consists of a body with curvilinear hollow margins, terminating in two elevated and rather pointed processes, one of which is connected with the cavity named tympanum, while the other, projecting into the orbit, affords attachment to several muscles. The anomalous character of this bone has perplexed several of the most distinguished zoologists; and while Geoffroy gives it the name of os tympano-styloideum, Spix considers it analogous to the annular process of the temporal bone, which in the human foetus is separate; and Carus regards it as representing the incus, to which it bears a remote resemblance in shape and in one of its connections.

Maxillary bones. Both maxillæ are void of teeth; but the hard, horny matter of the bill covering the margins and extremities of each jaw, and constituting the mandibles (mandibula), is manifestly constructed to perform for BIRDS what teeth do for the MAMMALIA. But the most remarkable peculiarity of the facial bones of this class is, that the upper jaw admits of more or less motion. In the majority of instances this is effected by the jaw being united to the cranium by means of thin, flexible, elastic, bony plates; but in the parrot family the upper jaw is entirely distinct, and is connected by a proper articulation.

The base of the palatine surface of the upper jaw is divided into 4 branches, which diverge backwards. The

two external ones, which correspond to the zygomatic arches of the MAMMALIA, and which are very slender, are articulated to the quadrangular bone which moves on the temporal before the ear. The two intermediate ones, which have been already stated to correspond to the pterygoid processes, and which are parallel, are placed beneath the septum of the orbits, and are articulated by their posterior extremities with a small bone, variable in shape, but named omoid by Herissant, which is also articulated with the quadrilateral. From this arrangement results a singular species of broken lever, not dissimilar to the parallel joint of the piston and lever of the steam-engine, and the effect of which is, that whenever the lower jaw is depressed by its proper muscles, it necessarily causes the quadrilateral bone to perform a slight rotatory motion, in consequence of which, by means of the omoid bone, the upper jaw is at the same time elevated on the elastic plates; and as soon as the lower jaw is raised, the elasticity of these plates forces down the superior one.

The upper jaw is immovable in a few instances only, and of these the calao or rhinoceros bird is one.

The breast bone (sternum) is a trilateral, boat-shaped bone, concave internally, convex with a middle longitudinal crest externally, with the base of the triangle above, and the apex, which is also incurved backwards, below. The middle longitudinal crest, which is occasionally named the keel (carina), is shaped something like a spherical triangle, with the broadest side above, the base before, and the apex behind; and its prominence forms large spaces on each side for the attachment of the pectoral and other muscles used in flight. In the male wild swan (anas cygnus), in some species of curlew, in the crane, and in the guinea-fowl, this crest forms a cavity for the reception of the windpipe. In the ostrich and cassowary, which do not fly, the sternum is void of crest, and is merely arched strongly.

The ribs, which rarely exceed 10 pairs, may be distinguished into sterno-vertebral and vertebral. Though the latter are generally before, they are sometimes also behind. The vertebral end terminates in two diverging processes, one of which is articulated with the vertebral body, the other with the transverse process. The sternal extremity consists of a bony process, which performs the part of the sterno-costal cartilages of the MAMMALIA by uniting the rib to the sternum. The ribs of birds, however, are further distinguished by presenting near their middle a flat long process, projecting from the rib backwards at an acute angle, and resting on the rib immediately below, so that each rib is supported not only on the vertebrae and sternum, or the vertebrae alone, but on the next rib below. These processes are obliterated in the lower ribs.

The coxal bones constitute one piece with the sacrum and lumbar vertebrae. The ischial portion is united with the sacrum, and the ischiatic notch is converted into a hole. The part which corresponds to the os pubis of the MAMMALIA is not consolidated before so as to form a symphysis, but proceeding directly backwards, terminates in a styloid process, variable in length and slenderness. The only exception to this mode of structure occurs in the ostrich, in which the pubal bones are united below. The infra-pubal or oval hole is present in the whole class notwithstanding. It is worthy of remark, however, that in young birds this and the ischial aperture are still notches, in consequence of the deficient ossification of the parts.

The direction of the pelvis in birds is nearly that of the spine, that is, obliquely backwards, and deviating but little from the horizontal line.

The wings or thoracic extremities are connected to the trunk by three bones, the collar-bone or clavicle, the scapula, and the bifurcated bone. The collar-bones, which

Comparative Anatomy. are straight, strong, and cylindrical, are articulated by a large head with the anterior and lateral part of the sternum, in which its motion is rather limited. It forms before and laterally two short processes, one anterior-inferior and internal, articulated with the bifurcated bone; the other posterior-superior and external, uniting with the scapula, and forming a cavity, in which the head of the humerus is lodged.

Shoulder-lade. The scapula is a long bone, flattened, but narrow, and slightly incurved, with the convex side turned towards the spine, to which it is nearly parallel in position. The head or anterior extremity is thick and extensive, oblique from before backwards, and is articulated behind with the clavicle, before with the humerus. The free extremity is thin, flattened, and sharp. The whole bone is not dissimilar in shape to a scimitar.

Bifurcated bone, or tarsale. Besides these, which BIRDS possess in common with the MAMMALIA, we find an azygous bone, situate on the mesial plane, denominated in ordinary language the Merry thought, and, from its shape, the fork-like or bifurcated bone. It consists of two long, rounded, converging branches, united at an acute angle, and forming a broad process, flat in the vertical direction, and by which it is articulated to the anterior extremity of the crest or carinated part of the breast-bone. To the posterior or free extremities of the divergent branches are articulated the humeral ends of the collar-bones, which are thus enabled to sustain the violent motions of the humerus during flight. The branches of the bifurcated bone are separate in the ostrich, and each is united with the clavicle and scapula of the same side, so that the three bones form only one, much flattened, and with a hole towards the sternal extremity. In the cassowary the bifurcated bone is reduced to a mere rudimentary process at the inner margin of the head of the clavicle. From these facts it results, that the bifurcated bone is particularly useful in the energetic and continued efforts of the wings in flight, and not only serves to keep the clavicles apart, but, by lengthening the distance between the collar-bones and sternum, enables the animal to use a longer lever. It is freest, strongest, and most elastic in the birds which fly best. In birds which do not fly, and which use the wings merely to sustain the equilibrium, as the ostrich and cassowary, it is reduced to almost nothing, or it is in such a rudimentary and imperfect form, that it cannot keep the collar-bones apart.

Thoracic extremities. The bones of the thoracic extremities, or those of the wings, correspond in general to those of the MAMMALIA. They consist of a single cylindrical humerus, articulated with the scapula and collar-bone above, two bones of the fore wing corresponding to the ulna and radius, two bones of the carpus, two of the metacarpus, consolidated by their extremities, one styloid bone as a thumb, a long finger consisting of two phalanges, and a short one consisting of one. The thumb supports the bastard pinions, the large finger and metacarpus the primaries, while the small one, which is covered by the skin, is destitute. In several of the web-footed divers, for instance the duck and penguin (alca impennis and spheniscus), these bones are flattened like thin plates.

Pelvic extremities. In the pelvic extremities the thigh-bone is provided with one trochanter only, is shorter than that of the leg, and is almost invariably straight; and is arched only in the cormorant, duck, and doo-chick. In the ostrich its diameter is about four times that of the humerus. The tibia differs from that of the MAMMALIA chiefly at its lower extremity. While the fibula adheres to it like a slender appendage as far as the middle, the tarsal extremity terminates in two trochlear condyles, with an intermediate pulley-like groove. The tarsus and metatarsus are repre-

sented by a single bone of considerable length, and the head or tibial end of which consists of a middle prominence and two lateral depressions, and which, therefore, moves in cardinal opposition, but does not admit of extension beyond the straight line. Though variable in proportion to the length in different orders, this bone is very long in the order GRALLÆ (GRALLATORES). It terminates below in 3 pulley-shaped processes, to which are attached the bones of the 3 anterior toes, with an internal margin for that of the great toe. In the ostrich there are 2 processes only, corresponding to the two toes. In the penguin tribe, however, the tarsus and metatarsus consist of 3 bones, separate from each other in the middle, but united at the tibial and digital extremities. To the tarso-metatarsal bone of the cock, and others of the GALLINACEOUS tribe, is attached the spur, a conical pointed excrescence of hard horny matter.

SECT. III.—OSTEOLOGY OF THE REPTILES.

The number of vertebrae, and all the other attributes of the spinal column, vary more in this class than in all the others.

In the CHELONIAD or Tortoise family there are 7 cervical, 8 dorsal, connected with the shell in an immovable piece, so as to have neither processes nor articular facettes; from 3 to 5 lumbar and sacral, consolidated in like manner; and about 20 caudal or coccygeal. (Plate XXXIV. fig. 5.)

In the SAURIAL or Lizard tribe, the number 7 predominates in the cervical, being that of the crocodile and most lizards. In several, however, there are 8, as in two of the monitor genus, the American safeguard, the lizard of Fontainebleau, the dragon, the iguana, the anolis, and the gecko and scinc; and in a few, as the Nilotic monitor, and an undetermined species of monitor, they amount to 9. In the chameleon there are only 5 cervical vertebrae. Here, however, a singular peculiarity is observed. Instead of the cervical vertebrae being, as in the MAMMALIA, distinguished by being unconnected with ribs, to those, from the third to the seventh inclusive, short ribs, unconnected with the sternum, are attached. The atlas and axis, therefore, alone are proper cervical vertebrae; but the general analogy is observed in the cervical ribs being exceedingly short and almost rudimentary. The dorsal vary from 11, which is that of the crocodile and iguana, to 29 and 30, which are the numbers in the New Holland scinc and Nilotic monitor. In the American safeguard, cordylus, stellio, crested basilisk, dragon, guana, and great anolis, they are 16; in the chameleon, black safeguard, and ameiva, 17; in the tupinambis, spotted gecko, and golden scinc, 18; in the green lizard and spotted guana (polychrus) 19; in the Fontainebleau and gray lizard 20; 21 in the Levant scinc and undetermined monitor; and 22 in the Java and New Holland monitor.

The BATRACHOID or Ranine reptiles are void of ribs, and it is impossible therefore to distinguish the first three orders of vertebrae from each other. In general, however, there are from the nape to the peleis 8 vertebrae, all provided with long transverse processes, and which are longest in the last. The sacrum is represented by a long flattened but pointed bone, without coccyx. In the Surinam toad (rana pipa) the last vertebra is consolidated with this bone; and the transverse processes of the second and third vertebrae are so much larger than the others, that they resemble rudimentary ribs. In the Salamander family there are from the head to the sacrum 14 vertebrae, all alike in shape except the first, which receives the occipital bone, and the last, which is articulated with the sacrum. These two are distinguished by wanting rudimentary ribs, which are small elongated bones, movable, and articulated with the trans-

Comparative Anatomy. verse processes, which are directed backwards. The articular processes are large and imbricated, the posterior resting on the anterior, so as to resist the motion of the spine backwards. The sacrum consists of one vertebra only, but the coccyx or tail is composed of 27.

In the SERPENTINE tribe the vertebrae may be said to attain the most extensive numerical development. With the exception of the head and rudimentary ribs, they constitute the whole skeleton. (Fig. 3.) From the head to the tail their shape is the same, and may be distinguished into body, articular and transverse spinous processes. In some species, for instance the boa, the spinous processes of the back are so much separated as to allow mutual motion to a considerable extent. In others, conversely, for instance the rattlesnake, these processes are so long and broad as to touch each other, while the oblique processes, which form their bases, are imbricated over each other. In consequence of this arrangement the motion of the spine is limited behind, but more extensive on the ventral surface. The vertebral bodies, which move easily on each other, are provided with a sharp spine directed towards the tail, which somewhat limits motion in this direction.

The first vertebra differs from those of the rest of the body in supporting short or rudimentary ribs; there are therefore no cervical vertebrae and no proper neck in the serpent family. The caudal vertebrae are distinguished by not supporting ribs, and by their spines both dorsal and ventral being double, and forming two rows of tubercles. The articulation of the bodies of these vertebrae is peculiar. On the anterior part of the body is a round hemispherical tubercle, while the posterior presents a corresponding cavity, so that each vertebra forms a cup and ball joint with the following one.

The number of costal vertebrae varies from 32, which is that of the blind worm (Angnis fragilis), to 204 in the ringed snake (Coluber natrix), 244 in the snake, and 252 in the Boa constrictor, and which is perhaps the greatest known number. Of intermediate numbers, the Amphisbaena has 54, the viper (Coluber berus) has 139, the rattlesnake 175, and the cobra di capello 192. The caudal vertebrae vary in number from 7, which is that of the Amphisbaena, to 112, which is that of the Coluber natrix. Of intermediate numbers, the blind worm has 17, the rattlesnake 26, the boa 52, the viper 55, and the cobra 63; from which it appears that the number of the caudal is not in proportion to that of the costal vertebrae.

Cranium. Of the heads of the CHELONIADS, the most remarkable characters are, that the facial line is horizontal, and quite continuous with the cranial line; that the orbits, though complete without, are continuous behind with the temporal fossae; that the parietal and occipital bones are compressed laterally, while the latter terminates above in a sharp spine, projecting behind. The occipito-parietal and occipito-temporal sutures are distinct. The cranial cavity is small compared with the volume of the skull.

These characters are not less remarkable in the SAURIAL or LACERTINE Reptiles. The cranium of a crocodile measuring from 13 to 14 feet is scarcely capacious enough to admit the thumb; and Cuvier estimates the area of the cranial section, which is oblong, at about $\backslash frac\{1\}\{2\}$ of that of the whole head. In these animals, indeed, the bones of the superior and inferior jaws are so much prolonged, and occupy so large a proportion of the head, that small space is left for the proper cranial cavity, which indeed is an immediate continuation of the vertebral. In these animals, also, the anatomist can trace, much more distinctly than in the more perfect, the resemblance between the cranial bones and the vertebral. In the CHELONIADS, and SAURIAL especially, the occipital bone is very distinctly a cephalic vertebra.

Still more manifest is this arrangement in the RANINE or BATRACHOID and SERPENTINE or OPHIDIAL Reptiles. In the former, as exemplified in the frog, the occipital bone, which forms the posterior cranial vertebra, consists of four pieces, and has two articular processes. The middle cranial vertebra is represented by the parietal bones above and the posterior part of the sphenoid below, while between it and the occipital or posterior is contained the temporal as the organ of hearing. The third or anterior cranial vertebra is represented by the anterior part of the sphenoid bone below and the two narrow frontal bones above. The face, which may be regarded as the organ of the senses, is elongated anterior to the head, somewhat after the manner of the CHELONIAD family; while an approximation to the BIRDS is indicated in the articulation of the lower jaw, which is connected to the head by the intervention of a quadrilateral bone.

In the Serpentine family, the cranium of which is very similar in other respects, the most remarkable deviation is in the want of ethmoid bone. The lower jaw is connected to the cranium by an intermediate bone, corresponding to the quadrilateral, but of an oblong shape, and something like a collar-bone.

The chest of the REPTILE class varies much in the mode of formation. While true ribs are recognised in the SAURIAL family only, the BATRACHOID reptiles have a sternum without ribs, the SERPENTINE ribs without sternum, and the CHELONIAD ribs united into the dorsal shell, and a sternum expanded into the abdominal one.

In the SAURIAL family the ribs correspond in number to that of the costal vertebrae already mentioned, that is, 12 in the crocodile and iguana, two of which are not connected to the sternum, 17 in the chameleon, 18 in the tupinambis, and 27 in the monitor. The SAURIAL reptiles, however, are peculiar in having from 1 to 6 ribs attached to the cervical vertebrae, and the opposite ends of which are not connected to the sternum. These, which have been named cervical ribs, form a transition to the rudimentary ribs of the SERPENTINE family, which are larger in the neck than elsewhere. The sternum of the crocodile consists of two parts,—an anterior or thoracic, which is osseous, supporting the two collar-bones,—and a posterior or abdominal, which is cartilaginous, and extends to the pabis, and furnishing to the abdominal parietes eight cylindrical cartilages. In the East India crocodile it appears that these lateral processes are converted into a single broad piece of cartilage on each side. (Fig. 4.)

The ribs of the CHELONIAD family are represented by the dorsal and the dorsal shell, which consists of eight broad incurvated plates, identified behind with the dorsal vertebrae, and terminating before in the margin of the shell, and which are doubtless genuine ribs. In the ordinary land-tortoise (Testudo Græca) these are seen in the shape of elevated bony ridges, proceeding from the head of each rib in a transverse concave bend to the margin of the dorsal shell. On each side of these ridges the bone is depressed, and is united at its lowest point by a genuine suture with the adjoining ones. These sutures, however, are not continuous with those of the sterno-abdominal shell, but meet it in the intermediate points. (Fig. 5.) The sterno-abdominal shell consists, in like manner, of several transverse pieces consolidated into one. The ordinary number is eight on each side of the mesial plane, and a ninth azygous, generally placed in the centre of the shell. In a specimen, however, of the tabular tortoise (Testudo tabulata), in our possession, the number of the sterno-abdominal pieces is 11, of which 8 are in pairs, united on the mesial line from before backwards, and 3 azygous at the posterior tip of the shell. In young animals it is easy to recognise the unions of these constituent bones, which consist of sutures, ex-

Comparative Anatomy. exactly similar to those of the cranium in the MAMMALIA. So feeble is the union, that it often happens that the abdominal shell especially separates at the lines of junction, in the attempt to detach it from the dorsal.

The BATRACHOID REPTILES, though void of ribs, are provided with a sternum, which before is a cartilaginous process, terminating on a disc placed below the larynx, where it receives the collar-bones, and forms behind a broad plate placed below the abdomen, and giving attachment to the muscles. In the Salamander tribe, which are without sternum, the ribs consist of twelve pair of small rudimentary processes, articulated with the vertebrae, but admitting of very limited motion.

Lastly, in the SERPENTINE family, though there is no sternum, the upper vertebrae are provided with costal processes, quite rudimentary. The great number of these costal rudiments, amounting in the rattlesnake to 175, in the cobra di capello to 192, in the coluber natrix to 204, and in the boa constrictor to 252, and the freedom of their anterior extremities, enable the animals of this tribe, which are destitute of locomotive members, thoracic or abdominal, to employ the spinal column and the ribs as organs of progressive motion. On this point the reader will find some interesting observations by Sir Everard Home (Phil. Trans. 1812, p. 163). In the region of the neck, where the ribs acquire peculiar length, they are employed in erecting that region, and producing the expansive swelling peculiar to this tribe of animals. It is an important link in the same series of facts, that in the animal absurdly named the flying lizard (draco volans), the five posterior ribs are recurved and elongated to form the bony skeleton of the membranous sails by which the animal supports itself in its desultory flight from tree to tree.

It is in the SAURIAL family that the locomotive extremities of Reptiles ought first to be studied. In these we find an elongated scapula without spine, and one short flat bone, constituting the clavicle, united to the sternum. In the iguana and chameleon this bone is broad and nearly quadrilateral, while in the tupinambis it is large and oval-shaped, with its greatest length from before backwards, and with two unossified points.

In the RANINE tribe, while the scapula consists of two articulated pieces, the upper towards the spine, each shoulder is provided with two collar-bones attached to the two extremities of the sternum, and the two anterior of which correspond to the bifurcated bone of birds. The sternum, collar-bone, and first part of the scapula, form one piece. In the salamander, in which the same consolidation is observed, the scapular portion is most distinct and directed to the spine, while of the clavicular portion the part connected to the sternum stretches below the chest, but, without uniting with that of the opposite side, the right glides over the left,—an arrangement which facilitates the dilatation of the chest during inspiration.

A nearer approach still to the bifurcated bone than is seen in the RANINE may be recognised in the CHELONIAD family. In these animals three bones are united to form the humeral cavity. The first is a flat, trilateral bone, situate below the abdominal and thoracic viscera, close to the abdominal shell, and which, notwithstanding its situation, is evidently the scapula. The second is a bone about the same length, flat, and like the feather of an oar at one extremity, which is free, round in the middle, and flattened in the opposite direction at the other end, which is firmly united at a right angle to a long slender cylindrical bone. At the angle of union of these two bones is part of the glenoid cavity, which is complete in the small end of the scapula. The first of the two bones is the collar-bone proper; the second is the lateral branch of the bone, which forms the bifurcated, and which is occasionally

united with its fellow. (Plate XXXIV. fig. 5.) The abdominal shell we have already stated to represent the sternum or breast-bone.

The humerus in the SAURIAL and CHELONIAD family is arched and incurvated in a serpentine direction. It is articulated with a radius and ulna, which are succeeded by three rows of carpal bones, one row of four metacarpal bones and digital phalanges, varying in number in different genera. In the skeleton of a fossil animal belonging to the SAURIAL tribe, originally delineated by Collini, and afterwards by Cuvier, and named by him the Pterodactyle or Winged (Pterodactylus, OSSEMENTS FOSSILES, tome v.), the metacarpal bone and phalanges of the index are prolonged to about twenty times the ordinary length, for the purpose, apparently, of giving attachment to the membranous web by which the animal occasionally elevated itself into the atmosphere. This animal, which, like the dragon (draco volans) of modern times, must have combined the contradictory characters of a flying reptile, may be regarded as forming the link between the REPTILES and BIRDS, as the Ichthyosaurus does between REPTILES and FISHES.

In the pelvis of the CHELONIAD family it is remarkable Pelvis. that the pubal and iliac bones appear to change places. Thus the ilium on each side is a narrow bone proceeding backwards to the sacral part of the spine, which is received between its posterior aperture; while the pubis appears in the shape of a broad, trilateral, flat bone, uniting before with its fellow on the mesial plane, behind with the ilium, and below with a flat, thin, quadrilateral bone, corresponding to the ischium, with which it forms the oval aperture. The inner of these three bones presents, as usual, the cotyloid cavity. It is further to be observed, that the two iliac bones, and consequently the whole pelvis, are movable on the vertebral column. (Plate XXXIV. fig. 5.)

In the SAURIAL Reptiles the pelvic bones are arranged Pelvic ex- tremities. and shaped nearly as in the CHELONIAD. In the RANINE the iliac bones are much elongated, and the pubal and ischial are consolidated into one piece, the symphysis of which forms a rounded crest.

The femur is short, thick, and incurvated sinuously, with the convexity before towards the tibial end, and the concavity towards the pelvic. Trochanters, though present in the CHELONIAD, are wanting in the SAURIAL and RANINE Reptiles. In the leg we find both tibia and fibula distinct, and of nearly equal size, in the CHELONIAD and SAURIAL family, but conjoined in the RANINE family. The tarsus consists of five bones, and sustains four or five metatarsal ones, on which are supported three rows of phalanges. The metatarsal bones, which vary in length, are longest in the crocodile and others of the LACERTINE tribe. In the RANINE, again, the astragalus and calcaneum are the bones of greatest proportional length.

The anatomical characters now enumerated are proper to the skeletons of Reptiles at present existing on the surface of the earth or in its waters; and in these we find a gradual transition from the SAURIAL and CHELONIAD, by means of the SERPENTINE, to the finny inhabitants of the ocean. Even the BATRACHOID Reptiles, in the early period of their existence while tadpoles, we shall have occasion to see, approach to the FISHES; and in one singular genus, if not two, the Proteus anguinus and Siren Lacertina, the characters of the Reptile are combined with those of the Fish, in having at once lungs or internal respiratory cells, and gills or external ciliated branchiae. The transition thus indicated is still more strongly demonstrated in the osteological characters of two Genera of animals now extinct, so far as is yet known,—the Ichthyosaurus and the Plesiosaurus.

From the specimens of the Ichthyosaurus hitherto discovered, it appears that the number of vertebrae varies from 80 to 90 or more; in one entire specimen they amounted to 104 (Conybeare and De la Beche); that they are flattened, with the transverse diameter greater than the longitudinal, and the two articulating surfaces of the bodies calycoid or cup-shaped as in Fishes. Though the annular part is distinct from the body, it is united to its sides. The spinous processes, which are long and prominent, form a continuous ridge above the spine, and are connected to each other by a process from the front of the one spine, which is inserted into a pit in the back of the other. Instead of proper transverse process, a certain number of the vertebrae are provided with two tubercles on each side of the body, of which the superior, convex, is articulated to the tubercle of the rib, while the other, which is concave, receives the head. In the inferior part of the vertebral column, these two tubercles, after approximating, are eventually identified into one.

The ribs, which are numerous, and extend from the occiput to the pelvis, are slender and trilateral in shape, bifurcated above, and attached to the vertebrae by a head and tubercle. In the perfect specimen of Mr de la Beche they amount to 31, and of these 17 appear to be cervical or anterior false ribs, with single tubercles; thus affording another mark of resemblance to the SAURIAN family in osteological characters.

The bones of the head, distinguished by the extraordinary size of the orbit, are similar to those of the Saurian Reptiles. The sternum, collar-bone, and scapula, though also similar to those of this family, bear a much closer resemblance to the figure of these parts in the Echidna and Ornithorhynchus. The humerus is short, thick, and sinuated; the bones of the fore arm flat, and probably constituting part of the fore or thoracic fin. The Carpus consists of three rows, the first containing three bones, the other two, four each. These are followed by five or six rows of flattened, irregularly cuboidal bones, gradually diminishing in size and number to the tips, and which represent at once the metacarpus and phalanges of the fore paw, used apparently chiefly as a fin or paddle. The pelvic extremities appear to have been less strong and perfectly constructed than the thoracic. The femur is smaller and shorter than the humerus; the tibia and fibula are flattened like the ulna and radius; the tarsus consists of two rows only, the first containing three, and the second five bones; and this in like manner terminates in five ranges of flattened bones, gradually diminishing in size, and which represent the metatarsus and metatarsal phalanges of the hind paw or paddle.

From the specimens hitherto discovered of the Plesiosaurus, it appears that the total number of vertebrae amounts to 90, of which 35 appear to be cervical, while the other 55 are dorsal and caudal, the regions of which are proportionally short. The head of this animal also is small and compressed, nor has it the large orbit of the Ichthyosaurus. Each rib consists of a vertebral and sternal portion, united at an obtuse angle, the former articulated by a single head to the transverse process, and the latter connected with its fellow by a transverse slip, so that the lower or abdominal ribs appear to have surrounded the abdomen with a complete cincture. The anterior part of the chest is occupied by two trilateral bones uniting in the middle, which, from their connection with the scapula, are believed to be the coracoid bones; and above these is a transverse piece, with a middle notch and lateral sinuated elevations, which is regarded as the sternum; while the scapula extends on each side like a buttress between the two. It is not improbable, nevertheless, that the middle portions named coracoid bones are the ster-

num, and the transverse bone the clavicles; and it is worthy of remark, that not only this bone, but the middle piece, closely resembles in figure and disposition those of the Echidna and Ornithorhynchus. The pelvis consists of three bones, a vertebral or superior, corresponding to the ilium, narrow and slightly incurvated; an anterior, ascending forwards, and broad, separating the pubis; and a posterior, short, forming the ischium. The humerus and femur are longer than in the Ichthyosaurus. There is a very short radius and ulna, and tibia and fibula, articulated with five carpal and tarsal bones; and the rest of both paddles consists of successive rows of flattened but long bones, contracted in the middle, and expanded at the extremities, representing the metacarpal and metatarsal digital phalanges. (Home, Phil. Trans. 1816, 1818, 1819, 1820; De la Beche and Conybeare, Geological Transactions, vol. v. p. 559; and Cuvier, Ossements Fossiles, vol. v. part ii.)

SECT. IV.—OSTEOLOGY OF THE FISHES.

The SERPENTINE or OPHIAD Reptiles present in their osteological characters an approximative transition to those of Fishes. While in the former order the skeleton is reduced to the spinal column, ribs, and head, in the latter class the spine and head only are left; and in some tribes the transition is still more distinctly marked by the presence of ribs.

The vertebra of a fish is distinguished from that of any Vertebrate other animal by the shape of its body. The cephalic and caudal, or anterior and posterior surfaces, are hollow cup-like cones, so that the union of each two vertebrae forms a double conical cavity, united by the base, containing a substance composed of concentric fibro-cartilaginous layers, with intermediate albuminous or gelatinous matter. By this cartilage the vertebral bodies are united; and on this the motions of the spine are effected. This motion, however, is chiefly lateral; for the spinous processes are so long, and the articulation so complex, that antero-posterior inflection or extension is nearly impracticable.

In the cartilaginous fishes, for instance the shark, sturgeon, and lamprey, the vertebral bodies form simple tubes, which, from the extreme elasticity of the constituent cartilage, propel the contained fluid to a considerable distance. Thus Sir E. Home saw the fluid projected to the height of four feet from the intervertebral cavities of the shark. (Phil. Trans. 1809.) In this order, also, the spine is infinitely more flexible, and its resilient power, when bent by the muscles, is almost incredible. On each side, also, the vertebrae are excavated, to form a canal for lodging the large blood-vessels.

The vertebrae of fishes are numerous, and not easily distinguished into classes. They may, however, be distinguished into two, according as the spinous process is above only, or above and below at once. Those with the dorsal spine only are denominated dorsal or abdominal vertebrae, and have commonly at the sides transverse processes for the attachment of the ribs. Those with the dorsal and ventral spines are distinguished as the caudal vertebrae. The last caudal vertebra is generally trilateral, flattened in the vertical direction; and its tip is marked with articular pits, which indicate the attachment of the small elongated bones which sustain the caudal fins.

The number of vertebrae varies. In the uranoscopus or star-gazer there are only 25, in the balista 17, and in the four-spined trunk-fish (ostracion) only 13; while in the sturgeon the number is 84, in the eel 115, and in the shark 207.

Though Fishes have no chest, and require none, since their respiratory organs are gills, all of them are not void of ribs. The ray, shark, syngnathus, tetraodon, diodon, cy-

comparative anatomy. clopterus, fistularia, &c. have indeed no vestige of ribs. But in the sturgeon, balista, eel, uranoscopus, pleuronectes, sea-wolf, and dory, they are in the shape of short rudimentary processes; in the trigla and loricaria their sides are horizontal; in the perch, carp, pike, and chelodon, they encompass nearly the whole upper region of the abdominal cavity; and, lastly, in the silver-fish (zeus vomer), the herring, rhomboidal salmon, &c. they are united to a sternum. In the little animal named sea-horse (syngnathus hippocampus), several series of osseous tubercles of the skin, surrounding the body like belts, are supposed to represent false ribs. The sternum is limited to a small number of FISHES. Besides those already mentioned, in the dory there is a series of minute flat bones disseminated along the lower edge of the belly, which is supposed to represent a rudimental sternum.

The head. In size and number the ribs vary, though in the silurus, carp, and chelodon they are of largest proportional size; in the herring they are as fine as hairs.

The head in the finny tribes is more an object of zoological than anatomical description. The chief points to be remarked are, that the cranium forms but a small part of the head; that the orbits are separated by a septum, sometimes membranous, occasionally, as in the wolf-fish, bony; and that there is on each side a large movable bone, corresponding to the quadrilateral of BIRDS, not square, however, but oblong, which supports not only the lower jaw and palatine arches, but the gill-cover. In the cartilaginous fishes the sutures are early obliterated, and the cranium consists of an inseparable mass of cartilage. In the bony fishes the cranium is separable into numerous pieces, and in the perch they amount to 80. In the cranium of fishes the anatomist recognises more distinctly than in the superior orders the formation according to the vertebral type. Small in proportion to the whole head, the cranium appears like a direct continuation of the vertebral column. In the osseous division of the class especially, the cranium may be distinguished into the occipital or posterior vertebra, the sphenoparietal or middle, and the frontal or facial vertebra. The cavity thus formed is very small; yet small as it is, it is not exactly filled by the brain, between which and the bones there is interposed a pellucid fluid, contained in fine cellular tissue. The cranium of the osseous fishes also is widest between the ears, because the organ of hearing is contained within its cavity with the brain. In the cartilaginous it is quite different.

Though FISHES are destitute of extremities similar to those possessed by the other three classes of the VERTEBRATA, they are not, however, without locomotive members. The thoracic extremities are represented by the pectoral fins, and the pelvic by the ventral. In short, it may be said that the bones of the thoracic and abdominal extremities are converted into osseous rays in the finny tribes.

In the Ray genus, in which the wing-like disposition of the pectoral fins gives the body a rhomboidal shape, they consist of numerous radiating cartilaginous lines, all attached to a cartilage parallel to the spine, divisible into two or three others, and articulated above to another adherent to the spine. Below there is a strong transverse bar common to the cartilages of both fins, and separating at once the sternum and clavicle. This transverse bar is also seen in the shark tribe; but their pectoral fins, which are much smaller, are not articulated with the spine.

In the osseous fishes, and in many others usually referred to the cartilaginous division, e. g. the balista, the pectoral fins are fixed to an osseous belt, which surrounds the body behind the gills, and which supports the

posterior margin of their aperture. This belt consists of a single bone on each side, articulated to the posterior-superior angle of the cranium, and uniting below the breast with that of the opposite side. This bone, which may be regarded as a scapula, varies in shape and the angle which it forms with its fellow in different species. In fishes flattened vertically, the angle of union is acute; in those which are depressed, the angle is so obtuse as to form nearly a straight line. In many fishes, especially those of the order THORACICI, e. g. pleuronectes, cottus, zeus, chelodon, perch, &c., in the small unicorn (balista), and others, the superior part forms a large spine, which descends immediately behind the fin, and to which the adductor muscles are attached. This spine, which is movable, has been improperly named a clavicle.

The rays by which the membrane is supported are not directly articulated to this belt, but are connected by a row of minute flat bones, which may be compared to the carpus in the other three classes. When the first ray of the pectoral fin, however, is thorny, as in the harness-fish (loricaria), and some species of silurus, it is articulated directly with an osseous belt; and it is remarkable that some fishes, as the silurus and stickle-back, have the power of retaining this spinous ray erected against the body as a means of defence. This is effected by a cylindrical tubercle, on which the spinous ray is articulated by a hollow, bounded before and behind by an elevated process. When the spine is erected, the anterior process, entering a hole in the cylindrical tubercle, is locked in it by the spine revolving slightly on its axis, so that it cannot be inflected unless by the spine revolving in the opposite direction.

The pectoral fins are so long that they answer the purpose of wings in several species of trigla, the trigla hippopotamus, the flying gurnard (trigla volitans), the springing gurnard (trigla evolans), in the scorpana volitans, the tropical flying fish (exocetus volitans), and some others. Their situation also is liable to vary. In the exocetus they are near the gills, but in the blennius and others they are remote. Lastly, they are totally wanting in a small number only, as the lamprey (petromyzon), the hag-fish (myxine, Lin.; gastrobranchus), the muræna, the eel genus, the sphagobrachius, &c.

The abdominal or ventral fins, which correspond to the pelvic extremities of the other classes, are so denominated or ventral because in the majority of fishes they are situate below the belly, and nearer the anal outlet than the pectoral. By this circumstance a numerous order are distinguished by the name of Abdominal Fishes (ABDOMINALES).

In a small number of fishes, comprehending the gadus, blennius, kurtus, callionymus, trachinus, and uranoscopus, the ventral fins are placed under the throat, below the aperture of the gills, and before the pectoral fins. This order is therefore distinguished by the name of JUGULARES.

In the most numerous order of all, the ventral fins are situate behind and below the pectoral fins. These have therefore been denominated Thoracic Fishes (THORACICI).

The ventral fins consist of two parts—one formed of rays covered by a double membrane, apparent externally, and constituting the proper ventral fin; the other internal, representing the coxal bones of the pelvis, always supporting the pinnal rays, and often articulated with the bones of the trunk. It is never articulated, however, with the spine, nor does it form an osseous belt round the abdomen. The bones of which it consists are generally flattened, varying in shape, and in mutual contact by the internal margin. In the shark and ray genera only is there a single transverse bone, nearly cylindrical, to the extre-

mities of which the fins are attached. The direction of the pelvic plane to the walls of the abdomen varies according to the shape of the body of the fish. In the flat fishes they are directed obliquely, and their inner margin forms the keel of the belly. In fishes with a broad or cylindrical belly they form a plane more or less horizontal.

In the JUGULAR and THORACIC FISHES, the pelvic bones are always articulated with the base of the belt which sustains the pectoral fins; and they vary much in shape and situation.

In the trachinus, uranoscopus, cottus, sciaena, chiodon, and perch, these two bones are united by their inner margin. In the cuckoo-gurnard, in which they are united by the posterior tip only of their internal margin, they are broad, flat, and oval. In the sole and flounder genus (Pleuronectes), in which the fins are attached to their anterior tip, they are united in a quadrangular pyramid, the apex of which is directed backwards and upwards, and the base forwards. In some of the stickle-backs these bones are altogether separate, and being long, receive in their middle a movable spine, which supplies the place of the ventral fin. In the dory (zeus faber, L.) they are flat and triangular, in mutual contact by their whole surface. In the silver-fish (zeus romer) they are small and cylindrical.

In the whole of the ABDOMINAL order, on the contrary, the pelvic bones are equally unconnected with the bones of the shoulder and with the osseous belt of the pectoral fins, and are confined to the middle-inferior part of the belly, not far from the anus. In general these two bones are separate from each other, and are retained in their situation by ligaments. In the carp, in which they are elongated, they touch only by their posterior third. In the herring, in which they are small and approximated, they are continuous with the minute bones of the sternum. In the pike they are broad and trilateral, approximated by their anterior tips, separate behind where the fin is attached. In the silurus, in which they are united, they form a round and often spinous shield before, while the fins are attached to the exterior-posterior margin. Lastly, in the cuirassier or harness-fish (loricaria), the pelvic bones are united in one piece, the posterior notch of which forms the anal aperture, while the fins are attached to its external margin.

The proper fin consists of a certain number of osseous rays, simple or bifid, supported by one or two rows of minute bones placed between them and the pelvic bones. On these small bones the constituent rays move, diverging or converging like the rods of a fan, while the whole fin may be inflected or extended by the minute bones moving on the pelvic, so as to adduct or abduct the fin.

In the cartilaginous fishes the structure is different. To the tip of each pelvic bone are articulated two principal cartilages, one external, forming a kind of toe with seven or eight joints; the other internal, supporting all the other rays of the fin, which often exceed thirty in number.

If we suppose these bones, like the minute ones of the pectoral fin, to represent the tarsus of the other three classes, it must follow that, in the locomotive extremities, the humerus, with the ulna and radius, and the femur, with the tibia and fibula, are obliterated. It is not unimportant to observe, that the general structure of the VERTEBRATA tends through various transitions to this termination. In the AMPHIBIA the long bones of the extremities are shortened by removing the diaphysis, and leaving their articulating ends only. In the CETACEA the pelvic extremities are removed altogether. In the CHELONIAN and SAURIAN REPTILES the same long bones of the extremities are much abridged; and in the ICHTHYOID REPTILES, now extinct, but sharing by their structure a form of animal existence partaking of the reptile and fish at

once, and perhaps intermediate between the two, this abbreviation is carried perhaps to its greatest possible degree, in leaving the articular ends only of the four locomotive extremities. Lastly, this reduction is merely preparatory to that exhibited in the whole class of FISHES, in which the three longitudinal bones so conspicuous in the higher classes of animals are completely obliterated, and those representing the hand or forepaw and foot are articulated directly to the shoulder and pelvic bones.

Besides the bones already mentioned as constituting the skeleton, there are observed in the osseous fishes minute bones, generally fork-like in shape, disseminated through all the muscular parts of the body. The purpose of these bones, which, as being totally insulated from the other parts of the skeleton, are denominated ossicula musculorum, is chiefly to afford points of support; and they are probably to be regarded as rudimental representatives of osseous parts, more completely developed in the higher animals.

It is further a curious circumstance, that the skeleton, which is so symmetrical in all the other classes and orders, begins to exhibit a deviation from this first in the skeleton of the finny tribes. In the Sole genus (Pleuronectes) this deviation is very conspicuous. Both eyes are placed on the same side of the mesial plane; and the side on which the eyes are placed is broader than the opposite one. The former is bounded by a convex margin, the latter by a concave one. The orbit towards the former is large, the other small and imperfect. Conversely, it is to be observed, that in the latter the maxillary and intermaxillary bones are larger than in the former. The sides of the inferior jaw are less discordant; and though in the Sole and Plaice those of the eyeless side are more straight and elongated than those of the other, in the Turbot (Pleuronectes maximus) they are nearly symmetrical.

CHAP. II.—COMPARATIVE MYOLOGY.

Though this is the proper place to consider the peculiarities of the muscular system of animals, the limits assigned to this sketch will not allow us to enter into details. We shall merely, therefore, take a cursory view of those points in which the myology of the lower animals differs from that of the human subject.

In general, in the lower animals, especially the MAMMALIA, BIRDS, and REPTILES, the muscles correspond in situation to those of the human subject; and whatever modifications they undergo consist in changes of figure, and in some few instances in changes in attachment. The former kind of changes may be in all cases pretty accurately estimated by the osteological characters of the class, order, or genus; for when the position, shape, or direction of a bone is altered, in the same proportion nearly are the attached muscles altered in their attributes.

Though in the lower animals, however, the zootomist Defessey traces muscles in general quite analogous to those of the human subject, in several instances this analogy ceases to be observed. In general the muscles of the lower animals are less numerous than those of the human subject; and this deficiency in number, though not much observed in the QUADRUMANA, is very remarkable in all the inferior orders of the MAMMALIA, and still more in the BIRDS and REPTILES. In general, also, these variations are most conspicuous in the locomotive extremities. Thus the small pectoral muscle, which is present in the QUADRUMANA, is wanting in the CARNIVORA and the whole of the Ungulated Animals and the Reptiles. The short supinator is present in the Canine and Feline genera, but the long is wanting; and both are absent in the CHIROPTEA, RODENTIA, PACHYDERMATA, RUMINANTIA, and

Comparative anatomy. SOLIDUNGULA, and in the whole class of BIRDS. Both pronators (teres and quadratus) are present in the QUADRUMANA and CARNIVORA, but wanting in the CHIROPTERA, Ruminants, and SOLIDUNGULA. The Rabbit; and perhaps the RODENTIA generally, have the pronator teres; but as the radius is not very movable, its influence is trifling.

Physiological peculiarities of the sole. In the mole the rhomboideus is inserted into the cervical ligament, which is ossified; and it therefore elevates the head and neck on the scapula with singular force. This is effected still more remarkably by the occipital part of the rhomboideus, the fibres of which being parallel to the spine, pass below the proper rhomboideus to be attached to the transverse ligament and the middle of the cranium. The strong, thick, quadrangular collar-bone has two muscles, a supraclavius and a subclavius. The large pectoral is very thick, and nearly as large as in birds.

The common extensor of the fingers or fore toes is the only muscle which is common to man and all the quadrupeds. Of the proper extensors the horse has two on the side of the common extensor, but acting as an extensor of the fore pastern; another between the common extensor and the extensor of the pastern, and which seems merely an appendage to the former. The proper extensor of the index is wanting in the RODENTIA, Ruminants, and SOLIDUNGULA; and while the two latter orders are destitute of the long and short extensors of the thumb, and the feline, canine, ursine, and leporine genera have the former, they are destitute of the latter. Lastly, the lower animals are wholly destitute of the short muscles of the hand, which in man produce flexion, abduction, adduction, and opposition. In the CHIROPTERA only is there one extensor, and flexors of the fore toes.

Among the muscles of the pelvic extremities the gluteus maximus, or large muscle of the buttock in man, diminishes much in the QUADRUMANA; and in the other orders is reduced to a very small size. The buttock in the MAMMALIA generally consists chiefly of the gluteus medius and minimus; and while the gluteus maximus is in the horse in a great part aponeurotic, the g. medius is so large as to produce those forcible and sudden extensions of the hind leg which constitute the kick.

In the leg the sartorius of the horse, the animal in which the muscles have been most studied, is large, and is distinguished by the name of the long adductor, in opposition to the gracilis, which constitutes the short adductor. The muscle representing the biceps of man is in all quadrupeds a uniceps, and the single head is attached to the ischium only. In the horse and dog it has been denominated the vastus longus.

The gastrocnemius externus et internus (gemellus), which constitute the calf of the human subject, diminish considerably in the lower animals; and the soleus, which is placed below them, also becomes small, and is particularly slender in the Ruminants and SOLIDUNGULA.

The following muscles are wanting in the whole class of BIRDS. The diaphragm, the recti abdominis, and the pyramidales; the muscles of the dorsal part of the spine, the splenius, the brachialis externus, or third head of the triceps; the supinators of the fore arm or wing, as already mentioned, all those corresponding to the short muscles of the hand and fingers; the quadratus lumborum, the psas parvus, the psas magnus, iliacus internus, obturator externus, and the extensor longus pollicis pedis.

Two muscles, which occupy the situation of the pronators, act as flexors, showing the connection between the actions of inflexion and pronation, and the occasional substitution of the latter for the former.

In this class, also, the gluteus maximus is of a pyramidal shape, while the true pyriformis is wanting. The

gluteus minimus, which is attached to the anterior edge of the iliac bones, is the iliacus. In place of the pectineus there is a slender muscle, which extends to the knee, over which its tendon passes, and gliding behind the leg, its tendon is bifurcated, one slip going back to be inserted into the posterior part of the metatarsus, the other to be united to the perforated flexor of the first and last toe. This muscle, which is named the accessory femoral flexor, is the one by which birds are enabled to clasp a perch during sleep.

In BIRDS the great pectoral is a remarkable muscle in point of size. It consists indeed of three muscles, the large pectoral, the middle, and the small, which occupy the sides of the vertical crest of the sternum, and constitute what is named the breast of the animal; and which are chiefly employed in the energetic motion of the wings in flying. These muscles are sometimes so large that they weigh more than all the other muscles of the animal together. In birds which fly much they are dark coloured and firm; in those which fly little, as the domestic poultry, they are white coloured, and in general soft. The same distinction is observed in the muscles of the two extremities. In birds much on the wing these muscles are dark coloured and firm, while those of the legs are comparatively lighter and more tender; and, conversely, in birds little on the wing and mostly on the legs, as the domestic poultry and many of the Grallae, the waders, swimmers, &c. the muscles of the wings are light coloured and tender, while those of the legs are dark coloured, firm, and strong.

The flexor muscles of the leg and toes of BIRDS merit Mechanic notice. They consist of muscles corresponding to the long flexors, which are divided into three masses. The first consists of five portions, three of which may be regarded as constituting a single common perforated flexor. It rises by two bellies, one attached to the external condyle of the femur, forming a perforated tendon, which receives one of those of the muscle corresponding to the peroneus; the other to the posterior surface of the femur, forming the tendons of the index and small toe. This muscle is further connected by intermediate fibres with the accessory femoral flexor,—a muscle placed on the internal surface of the thigh, and sending its tendon over the knee; and as the tendons are inserted into the ungual phalanges, when the accessory femoral bends the thigh the flexors of the toes inflect them also, and retain them in the inflected position. By means of this arrangement birds are enabled to clasp a perch or other small body when roosting, without continued muscular effort, and thereby to sleep on the perch. This mode of explanation, which was originally given by Borelli, has been controverted by Vicq d'Azyr; but apparently not on good grounds.

Among the class of Reptiles, while the muscles of the OPHIDIAL family are confined to those of the vertebrae and rudimentary ribs, in the CHELONIAN these are obliterated, and the muscles of the neck, head, and tail, and those of the locomotive extremities alone, are left. In the other two classes of reptiles the muscles are in general analogous to those of the MAMMALIA.

There are not many instances of muscles which, though unknown in man, are found in the lower animals. Of muscle, these the most remarkable are the cutaneous muscle (panniculus cutaneus), and the suspensory of the eye. The former was absurdly maintained to exist in the human subject, especially by Nicolaus Massa; but it is manifest that the assertion was derived from the dissection of the lower animals only. It was not long after demonstrated by Charles Etienne, that no fleshy pannicle or cutaneous muscle exists, such as is found in the lower animals; and

that the only cutaneous muscles in man are the latissimus colli, the epicranus or scalp-muscle, and those which are attached to the face, and which by their motion give expression to the countenance. The cutaneous muscle even is not found in the QUADRUMANA, nor does it exist in the pig. In various other animals, however, it is found in different degrees of distinctness. It is very well marked in the hedgehog and porcupine;—by its means they have the power of erecting their spines, and rolling themselves up;—and in the armadillo and the ant-eater tribe. In the mole, also, we have seen it pretty well marked.

It is an interesting fact, that Galen originally observed that the lower animals possess a seventh muscle of the eye, or one more than man. The suspensory or infundibular muscle (musculus choanoides), as it has been named, from its shape, especially in the Ruminants and SOLIDUNGULA, has the apex fixed to the margin of the optic hole, and its base inserted a little behind the four straight muscles. In the ZOOPHAGA and CETACEA it consists of four parts, so that these orders appear to be provided with 8 straight muscles. In the rhinoceros it consists of two portions.

There is yet another part, the muscles of which can scarcely be said to exist in the human subject, but which attain a very great degree of development in the lower animals. The coccyx of the human subject is expanded in the lower animals into a highly flexible prolongation denominated the tail (cauda), variable in length, but always consisting of separate vertebrae, articulated and movable on each other. While the coccyx of the human subject possesses two muscles only, the ischio-coccygeus and sacro-coccygeus, which are so insignificant in size that they scarcely serve to move the part, the caudal vertebrae of the lower animals are moved by muscles greatly larger, more numerous, and more powerful.

The tail is to animals a much more useful and powerful organ than the coccyx to man. It is a member which peculiarly belongs to them; and though in ordinary circumstances pendulous, it is made to assume a variety of motions of which no other organ is susceptible, and to perform duties which would be otherwise impracticable. With many, as the long-tailed monkeys, the sloths, the ant-eater, and the squirrel tribes, it is indispensable as an organ of prehension. The majority of animals, as the Ruminants, SOLIDUNGULA, &c. use it as a whip or lash to drive away insects. The lion, tiger, and others of the feline tribe, lash their sides with it when enraged. The Cetaceous swimmers employ it as a rudder and oar in the waters. The beaver uses it as a trowel, to enable him to construct his clay-built dwelling. An organ employed so variously must consist of a muscular apparatus rather complex.

The different motions of which the tail of the MAMMALIA is susceptible may be referred to three heads,—one by which it is extended or elevated, another by which it is inflected or depressed, and a third by which it is made to beat the sides. The combination or succession of these motions gives rise to secondary ones more complex in character. It may be twisted on its axis, or turned in a spiral direction. These motions are effected by three classes of muscles.

1st, The muscles which raise the tail are situate above; they are musculi sacro-coccygei superiores. Commencing at the base of the articular processes of the 3 or 4 last lumbar vertebrae, or those of the sacrum and the caudal vertebrae, by fleshy slips, they are connected to tendons, which are inserted into the base of the first of the caudal vertebrae, which are void of articular processes. The second tendon goes to the next following vertebra,

and so on to the 13th, each contained in a ligamentous groove, which forms an investment. The muscles of both sides acting together, elevate or incurvate the tail upwards. Comparative Anatomy.

The interspinalis and spinalis obliquus or lumbo-sacro-coccygeal are the continuations of the interspinales dorsi et lumborum. The spinous processes, however, becoming indistinct, or being represented by two tubercles, the attachments vary.

2d, The muscles which depress or inflect the tail downwards take their origin within the pelvis, and are prolonged to various extents along the inferior surface of the tail. Of these there are four pairs, the ileo-coccygeal of Vicq d'Azyr, the inferior sacro-coccygeal, the inter-coccygeal muscles, and the pubo-coccygeal of the same author. The insertions of these muscles vary in different genera, according to the number of vertebrae of which the tail consists. The pubo-coccygeal is wanting in the raccoon, but it is distinct in the dog and opossum. The effect of the ileo-coccygeal and it, is to depress the tail and apply it forcibly to the anus.

3d, There are only two muscles which carry the tail to the sides of the animal—the ischio-coccygeus externus, and the intertransversalis of Vicq d'Azyr; the former proceeding from the pelvic surface of the ischium below and behind the acetabulum, to the transverse processes of the caudal vertebrae, the second extending in a continued band between all the transverse processes.

The tail, therefore, in the MAMMALIA, consists of a series of successively decreasing vertebrae, moved by eight pairs of muscles.

In FISHES it is not easy to trace any analogy between the muscles and those of the other classes. Though the spine, head and fins, have appropriate muscular bundles, the natural or fascial distinctions are less evident than in the other three classes. It is important, however, to remark, that while the muscles which move the spinal column are placed in these classes, partly before, and chiefly behind the vertebrae, those of FISHES are placed on each side. Hence the lateral motion of the spine, which is inconsiderable in MAMMIFEROUS animals, BIRDS, and REPTILES, becomes very conspicuous in the finny tribes, especially in the motion of swimming, while the antero-posterior inflection or extension is altogether insignificant.

It is almost superfluous to remark, that, in the greater part of the finny tribes, the muscles are white or pale coloured. In a few only, for instance the salmon, trout, gwinad (coregonus), herring (clupea harengus), carp (cypinus), and some others, are the muscular fibres of a pale flesh red. The circumstances on which these differences depend are not known; but it is supposed that in the latter sorts the proportion of oleo-albuminous matter is more abundant than in the former. The proportion of albumen, however, in the muscles of fish, seems in general to be small. They abound in gelatine and isinglass; and in some of the cartilaginous fishes especially, the greater part of the muscles seem to consist chiefly of gelatine in various degrees of consistence. This is particularly the case with the lamprey, the hag-fish (myxine glutinosa), and even with the sturgeon. The sterlet especially (aci-penser ruthenus), a small species of sturgeon found in the rivers of Russia, both European and Asiatic, abounds in gelatine; and the presence of this principle enables the inhabitants to use it in the preparation of a species of soup, the sterlet, which is esteemed a great delicacy. In some of the genus Pleuronectes this principle is also very abundant. Thus the Plaice (Pleuronectes Platessa), sole (P. Solea), and especially the turbot (P. Maximus), contain a considerable proportion of gelatine. On the proportion of this principle depends the quality of fish used

as an article of food in nourishing without exciting. All fishes which abound in gelatine uncombined with oleo-albuminous matter may be safely used as articles of food; while those in which the latter ingredients predominate are invariably eaten with the risk of disordering the stomach and producing indigestion.

CHAP. III.—COMPARATIVE ZESTHIOLOGY, OR THE COMPARATIVE ANATOMY OF THE ORGANS OF SENSATION.

SECT. I.—THE ORGAN OF SMELL.

The organ of smell consists of the nasal cavities, those of the ethmoidal and turbinated bones, and the frontal, sphenoidal, and superior maxillary sinuses, all of which communicate with the nasal. The whole of these parts are invested by fine periosteum, lined by mucous membrane. The ethmoid bone is the essential organ of smell; and the others appear simply to multiply the extent of the membrane.

The ethmoid bone consists of a perforated plate, with a middle vertical one attached at right angles to it, and lateral portions composed of thin bony plates convoluted with various degrees of complexity and minuteness in different orders and genera of animals. These convoluted plates form what are denominated the ethmoidal cells. They may be represented as numerous tortuous canals, proceeding from the perforated plate forwards and outwards, approximating mutually, and forming numerous communicating cavities. Such nearly is the structure of these plates in the EDENTATA, RUMINANTIA, SOLIDUNGULA, PACHYDERMATA, and CARNIVORA, the last of which have more complicated cells than the first. In the dog they are numerous and extensive. In the RODENTIA, for instance the porcupine, they are few—not above 3 or 4 on each side.

In BIRDS the internal side of each nostril is occupied by three orders of plates; the inferior turbinated or spongy bone; the middle, consisting of one plate convoluted on itself two turns and a half; and the upper, shaped like a bell, adhering to the frontal and lacrimal bones. These form three tortuous passages, varying in size and tortuosity in different genera. Though generally cartilaginous, these turbinated bones are membranous in the cassowary and albatross, and bony in the calao and toucan.

In the nostrils of REPTILES there are convoluted prominent plates, which, however, are merely membranous productions, unsupported by any bone.

In the FISHES, in like manner, there are membranous folds, the disposition of which is tortuous. They are, however, more regularly arranged than in the other classes. In the cartilaginous fishes they consist of semilunar folds placed in parallel tracts on each side of a broad plate, which divides the one side of the nasal cavity from the other. In the sturgeon, however, they are arranged in diverging plates, which are subdivided into more minute ones, like the branches of a tree. In the osseous fishes generally they consist of radiating plates disposed round a prominent central tubercle.

In these three classes the olfactory nerve is distributed to the membrane much in the same manner. This nerve, however, does not proceed farther than the superior turbinated bones; and the middle and inferior appear to be supplied with filaments of the fifth pair, the naso-optic branch of which is distributed to the nose in all the vertebrated animals. In the MAMMALIA, further, the sphenopalatine ganglion sends several filaments to the posterior part of the nasine membrane.

By most zootomical authors the trunk of the elephant has been described as an organ of smell peculiar to that

animal. We are satisfied, however, from observing the motions of this body in the living animal, that it is an organ, not of smell, but of prehension. Cuvier, after adopting the ordinary view, has relinquished it, and, on the ground of personal inspection, admits that the sense of smell in the elephant is confined, as in other animals, to that portion of the nasal cavities which is contained within the bones of the head. The trunk of the elephant, therefore, will with greater propriety be noticed under a subsequent head.

The nasal cavities of the CETACEOUS animals are not so much organs of smell as channels of respiration, and must also be noticed afterwards. It is sufficient here to remark, that in these animals the part of the cranium corresponding to the ethmoid bone is penetrated by no aperture, or, in other words, is not an ethmoid bone. It has therefore been asserted that the CETACEA have no olfactory nerve, and no sense of smell. This, however, is by no means established. Blainville and Jacobsen have observed in the dolphin nerves which they regard as olfacent; and Treviranus delineates nerves of the same character. By Otto and Rudolphi, on the contrary, who have had frequent opportunities of dissecting the dolphin and whale, the existence of these nerves is denied.

Though almost all the invertebrated animals give proofs of the existence of the sense of smell, in none of them do we find any organ in which this sensation appears with certainty to be exercised. That these animals possess the faculty of smell, is inferred from the fact, that insects recognise their food at a distance; that male butterflies scent the female even when inclosed in cages; and that the ordinary flesh-fly deposits her eggs on tainted meat, and occasionally on fetid plants, in the belief that they are the proper nidus, though in the latter case the larvae perish for want of the necessary sustenance.

Since odorous particles are evidently applied to the olfactory membrane of all aeropnoic animals by the medium of the atmosphere, and since the organ of smell is therefore situate in connection with the wind-pipe, it was conjectured by Baster, that, in insects at least, the organ of smell is situate at the entrance of the tracheæ or air-tubes. This conjecture derives some probability from the fact, that the inner tracheal membrane in these animals is soft and moist, and that those in which it is expanded into convoluted lacunæ and tortuous vesicles, for instance beetles, flies, and bees, are remarkable for the nicety of their sense of smell.

The antennæ, in which this sense has been placed by some anatomists, appear to be rather organs of touch than of smell.

In the MOLLUSCA the whole cutaneous covering seems to combine the character of an organ of touch or tact, and of smell. Like an extensive pituitary membrane, it is soft, villous, moist, and liberally supplied with nerves. The ARTICULATA and Zoophytes seem much in the same state. But on all these points information is rather conjectural than positive.

SECT. II.—THE EYES; THE ORGANS OF VISION.

All red-blooded animals, without exception, are provided with two movable eyes, consisting of the same essential parts as those of man, forming globular organs, and placed in the cranio-facial cavities named orbits. In none are there more or fewer; and the exceptions to the general rule, either in relation to the presence of these organs, or number, are only apparent. Among the MAMMALIA, Blind indeed, there are two instances of blindness,—in the zenni though not or blind rat (Mus typhlus, Lin.; Spalax typhlus, Pall.), and eyeless the golden mole (Talpa Asiatica, Lin.; Chrysochloris, Læcep. and Cuvier). But in neither of these animals are

Comparative Anatomy. the eyes absolutely wanting; they are merely very minute, and covered by a thin fold of hairy skin, in which there is said to be no aperture. Much in the same manner the murana coccilia, and the hag-fish (myxine, Lin. gastrobranchus cæcus), though provided with eyes, are deprived of the use of these organs by the opacity of the conjunctiva. In the Anableps (Cobitis Anableps, Lin.), the cornea and iris are biparted by transverse bands, so as to give the animal the appearance of having two pupils in each eye, though the crystalline lens, vitreous humour, and retina, are single. This animal affords the only example of this structure among the vertebrated animals; but a similar arrangement is observed in the Cephalopodous MOLLUSCA and cuttle-fish family.

Figure of the eyeball. The general shape of the eye depends on the medium in which the animal lives. It is nearly spherical, or approaching the spherical shape, in man and the quadrupeds moving along the surface of the earth; that is, in the lowest and most dense region of the atmosphere. The cornea merely forms a slight convexity, in consequence of being the segment of a smaller sphere than the rest of the eyeball; yet in the porcupine, opossum, &c., this difference is inconsiderable. To show the degree of this convexity, it is merely requisite to compare the axis or antero-posterior diameter with the transverse diameter of the ball, as exhibited in the following table:—

According to the measurements of the younger Soemmering, the axis of the human eye, taken in a beautiful Tyrolese girl of 20, is to the transverse diameter as 100 to 95; that of the eye of the magot (simia inuus) as 85 to 84; and that of the bat (vespertilio auritus) as 12 to 11. In the raccoon (ursus lotor) and lynx (felis lynx) alone the axis is exactly equal to the diameter. In all the other vertebrated animals, it is, as in the measurements of Cuvier, less than the transverse diameter at the rate of from 9 to 33 or 45 per cent. In the horse it is as 186 to 212, in the seal as 130 to 142, in the Indian elephant as 135 to 180, and in the black whale (balæna mysticetus) as 20 to 29. In the owl it is as 17 to 18, in the golden falcon as 14½ to 16, in the ostrich as 18 to 19½, and in the swan as 7 to 10. In the REPTILES and FISHES it is always less at the rate of from 3 to nearly 10 per cent. In the cuttle-fish, which may be taken as a general example of the invertebrated classes, it is much greater, the axis being to the diameter as 80 to 57. (D. W. Soemmering de Oculorum Hominis Animaliumque Sectione Horizontali Commentatio. Goetting, 1818, fol.)

In FISHES and the CETACEA which inhabit the sea, the anterior part of the eyeball is much more flattened, and in many fishes it resembles a hemisphere with the plane surface before and the convex behind. In the ray genus the superior part is also flat, so as to give the eye the appearance of the quadrant of a sphere, cut through two large circles perpendicular to each other. In some fishes, especially the burbot (Gadus Lota), the cornea is convex.

In BIRDS which occupy the elevated regions of the atmosphere, the deviation from the spherical shape is in the direction opposite to that of fishes. On the anterior part, which is sometimes flat, sometimes shaped like a truncated cone, is chased a short cylinder, closed by a very convex, and occasionally hemispherical cornea, always belonging to a much smaller sphere than the posterior convexity.

Aqueous humour. These differences in shape depend on the proportion between the density of the medium in which the animals live

and that of the aqueous humour. In the higher regions of the atmosphere, in which the air is very much rarefied, the refracting power of the aqueous humour is much more considerable than at the surface, occupied by quadrupeds; and hence it is more abundant in the former than in the latter class. Its refracting power, however, would be almost extinguished in a watery medium, from which it could differ but little in density; and hence it is either trifling or absolutely wanting in the inhabitants of the deep. In the cuttle-fish family it is entirely wanting.

The crystalline lens in FISHES, which is nearly spherical, projects through the pupil, and leaves little room for the aqueous humour. The lens is also very convex in the CETACEA, the AMPHIBIOUS MAMMALIA, the diving birds, as the cormorant, and the marine and aquatic REPTILES.

Affecting the oblate spheroidal shape in the MAMMALIA, it becomes extremely so in man, and still more in BIRDS. Its consistency is greatest in animals in which it is most convex; and hence it is matter of common observation, that the crystalline of fishes is particularly firm. It also contains rather more albumen than the lens of the MAMMALIA. The crystalline lens occupies least proportional space of the eyeball in man, and most in fishes.

The comparative spaces occupied by each of the humours may be understood from the following table, in which the axis of the eye, or the space occupied by the whole three humours, is represented by unity.

On the proportion of the total volume occupied by each of the three transparent parts there are few accurate facts. It may be remarked, however, that the human eye among the MAMMALIA is that in which the vitreous humour is proportionally most abundant. It is estimated to be twenty times more copious than the aqueous. In the ox it is only ten times, and in the sheep nine times the quantity of the aqueous.

In the MAMMALIA generally, the sclerotic is comparatively elastic, soft, and yielding; but in all animals in which the eye deviates from the spherical shape, as the CETACEA, FISHES, and BIRDS, this membrane is strengthened by greater solidity and thickness of tissue, or supported by accessory parts of a hard unyielding structure.

In the eye of the whale these two parts, the hard and soft, are naturally distinguished in a very striking manner. The lateral parts of the sclerotic are nearly an inch thick, and very hard. The posterior part is about one and a half inch thick, and softer, because the spaces between the firm fibres are filled with oily substance. The posterior part presents for the optic nerve a canal one and a half inch long, the walls of which are formed chiefly by fibres in direct continuity with the dura mater,—the only fact, it may be observed, which favours the statement of the ancient anatomists, that the sclerotic is derived from the dura mater. The sclerotic of the porpoise, though only two or three lines thick, has the same structure as that of the whale. In the seal it is thick before and thicker behind, but the middle zone is thin and flexible.

The sclerotic in BIRDS is thin, flexible, and elastic behind, with a bluish glistening aspect, and without distinct zone of fibres. The optic nerve enters, not by a hole, but an ob-
lique canal. The anterior part consists of two plates, between which is encased a zone of thin, hard, oblong, osse-

Comparative Anatomy. os scales, varying in number from 11 or 12 to 14 or 15, imbricated over each other so as to give the anterior part of the eyeball a great degree of hardness, and a figure unsusceptible of change. Though these plates are nearly flat in most birds, and form an annular zone slightly convex, they are broad, arched, and concave internally in the owl genus, and form a bell-shaped tube, with the posterior aperture oval and the anterior round. This may be denominated the osseous ring (annulus osseus, zona ossea). In the ostrich it is narrow and flat.

Among the REPTILES, the CÆLONIADS possess an osseous zone, consisting of plates inclosed in the membrane without being continuous with its substance. They are also found at the lateral part of the sclerotic in the chameleon and some of the SAURIAN Reptiles, as the Crocodylus Sclerops and Lucius, the monitor, and the iguana (D. W. Soemmering). It is also an important character in the structure of the eye of the Ichthyosaurus, which indicates the connection of that animal with the SAURIAN tribe, that its sclerotic was provided with an osseous zone, consisting, as in these, of 13 separate pieces.

In FISHES the sclerotic is cartilaginous, homogeneous, semi-translucent, elastic, and, though thin, firm enough not to collapse. In the ray it is expanded into a tubercle, by which the eye is attached to a peduncle or stalk. The sclerotic of the sturgeon is so thick that it resembles a cartilaginous sphere, with the external part hollowed for the humours and membranes.

In the CEPHALOPOD MOLLUSCA it forms behind a truncated cone, with the apex at the bottom of the orbit containing the gangliform swelling of the optic nerve, and several glandular parts, with the eye before.

Ornea. The cornea has often been represented to be merely a continuation of the sclerotic; and though this is easily disproved by accurate dissection of the human eye and that of our ordinary domestic animals, its inaccuracy is much more manifestly demonstrated in the animal world at large. In the whale and rhinoceros the margins of the two membranes penetrate reciprocally. In man and the ox the corneal margin is enclosed within the sharp imbricated edge of the sclerotic. In the tope-fish (equulus milandra, Lin.; galeus, Cuv.) the cornea is observed distinctly passing within the sclerotic in the manner of imbrication. The cuttle-fish is destitute of cornea; and as there is no aqueous humour, the crystalline lens is covered by a fine thin membrane, extended beneath the conjunctiva.

In all animals provided with eyelids, the mucous membrane, after being folded behind the eyelids, is reflected forwards over the sclerotic and cornea, in the form of a thin, transparent membrane. In those void of eyelids, as most fishes are, the skin, passing into mucous membrane, is continued directly over the cornea, without forming any angular fold, and adheres strongly. This is very distinct in the eel, which may be flayed without leaving any trace at the side of the eyes, except a round, translucent spot. The same peculiarity is remarked in SERPENTS and in the cuttle-fish family. In the zenni, golden mole, blind eel, and hag-fish, it has been already stated that the cornea is covered by opaque mucous membrane.

Coroid cat and Ruyschian membrane. The choroid coat exists in all animals yet examined. It is always very vascular. The inner layer, which has been distinguished by the name of tunica Ruyschiana, can scarcely be said to exist in man, small quadrupeds, and birds. In the large quadrupeds, however, especially the CÆTACEOUS animals, it appears in the form of a distinct simple membrane like epidermis. The lateral and anterior parts of the membrane are always invested by a semifluid, viscid substance, of different shades of black or brown black (pigmentum nigrum); chocolate brown in the hare, rabbit, and

pig; deep red brown in some birds; and purple red in the calmar. The absence of this dark-coloured pigment, which is not unfrequent, is observed in albinos, both human and animal, for instance white rabbits and white mice. The transparency of the Ruyschian membrane then shows the choroid of its natural red colour; and the pupil is red and contracted, and the eye intolerant of light.

In the ZOOPLAGIA, RUMINANTIA, PACHYDERMATA, SOLIDUNGULA, and CÆTACEA, the concave or inner surface of the Ruyschian membrane is diversified with colours of metallic lustre, more or less brilliant and something iridescent. In the ox it is of bright metallic green, changing to sky-blue; in the horse, buck, buffalo, and stag, it is a silvery blue passing to violet; in the sheep of a pale golden green, sometimes bluish; in the lion, cat, bear, and dolphin, of a pale gold yellow; and in the dog, wolf, and badger, of a pure white, surrounded by blue. This coloured part of the inner choroid surface, which occupies chiefly the side opposite to that on which the optic nerve enters, is named the tapetum. The use of it is by no means obvious. The explanation of Monro in reference to the tapetum of the ox, that it represents more distinctly to that animal the colour of his natural food, is not only frivolous, but inapplicable to the other genera.

The tapetum is wanting in all BIRDS and FISHES, excepting the ray, in which there is at the bottom of the eye a beautiful silvery-coloured space, produced by the transparency of the Ruyschian tunic, through which the tint of the choroid is seen.

In FISHES generally the choroid consists of two distinct separable membranes; the external, the proper choroid, white, silvery, or golden, very thin and not vascular; and the inner or Ruyschian, black, and consisting of a network of vessels. Between these two membranes is a body of a bright red colour, consisting of numerous tortuous vessels, convoluted and inclosed in pulpy filamentous tissue. Its general shape is that of a thin cylinder, encompassing the optic nerve like a ring, which, however, is incomplete at one side. This is the choroid gland,—a body about which there has been some difference of opinion, but which appears to be glandular rather than any thing else. Its vascular structure is well seen in the globe-fish, perca labrax, and cod, in which they are very large, and form numerous anastomotic communications. They are generally covered by a white, opaque, viscid fluid.

The choroid gland is wanting in the CARTILAGINOUS FISHES, the eye of which approaches more nearly to that of the MAMMALIA in this as in other circumstances. The choroid of the ray and shark genera is a threefold tissue of vessels, thick and consistent; the tunica Ruyschiana is very thin and semi-transparent; and between these is a layer of silvery matter with metallic lustre.

In the cuttle-fish genus, though between the sclerotic and choroid there are several glandular bodies, there are none between the choroid and Ruyschian tunic. The choroid is thick, soft, and vascular; the Ruyschian thin, firm, and dry; and though there is no tapetum, the whole interior surface of the eye is covered by deep purple, semi-fluid, viscid varnish.

Ciliary processes are found in all the MAMMALIA, BIRDS, Ciliary processes and even in the cuttle-fish among the MOLLUSCA; but they are wanting in almost all fishes.

The indented border of these processes is more distinct, and is converted into a genuine fringe in the large animals, as the ox, horse, rhinoceros, and whale, in which the angle applied to the capsule is more acuminate than in other animals. In the CARNIVORA, particularly the lion, the base of the plates is shorter in proportion to the other sides than in the previously mentioned animals, so that the opposite angle is more prominent; nor is the

Comparative Anatomy. border indented. In all the species every third or fourth plate is shorter than the others, but without determinate order.

The ciliary plates of BIRDS are mere serrated striae, without sufficient prominence to make them undulate in fluid. In the owl they are fine, closely set, and numerous; in the ostrich they are larger and more numerous; and in all, their extremities adhere firmly to the capsule of the lens.

In the tortoise the ciliary processes are so short that they are recognised only by the impression left on the vitreous humour; in the crocodile, however, they are distinct, and terminate each by an angle nearly right. They are indistinct in the toad, and imperceptible in the ordinary lizards and serpents.

The ciliary body and processes are large and distinct in the tope-fish; but if they are seen in any other of the cartilaginous fishes, they are wholly wanting in the osseous, in which the Ruyschian tunic is directly continuous with the uvea.

The utility of these processes in retaining the lens in its position is nowhere so distinctly seen as in the eye of the cuttle-fish family, and especially the many-feet of the ancients (polypus octopoda). In these the ciliary processes form a large diaphragm or zone, in the aperture of which the crystalline lens is truly chasmed. They penetrate a deep annular furrow which surrounds the lens, dividing it in two unequal hemispheres, and cannot be detached without laceration.

The iris, of the same intimate structure as in man, is of a deep tawny or brown in the MAMMALIA, and marked with fewer coloured striae than the human iris. In BIRDS it is of a uniform lustreless colour, varying according to the species, bright yellow, red, or clear blue. In the microscope it appears like a net-work formed by the intersection of numerous very minute fibres. The uvea is so fine that when the viscid varnish is removed it becomes transparent, and the iris appears of the same colour on both sides. In FISHES, conversely, the iris is so thin and transparent that the uvea is seen through it, of a golden or silvery brilliance, showing its direct connection with the choroid. Intermediate in metallic splendour is the iris of the REPTILES. The vessels, however, are greatly more conspicuous, especially in the crocodile.

The central aperture or pupil, though round in man, the QUADRUMANA, many of the CARNIVORA, and BIRDS, is not of that shape when contracted in all animals. In the feline family it consists of two elliptical segments, which form angles above and below, and which approach mutually so as to form a slit nearly vertical. In the RUMINANTS it is oblong transversely, and forms at its greatest contraction an oblong or transverse slit. In the horse, in which it is also transverse, its upper margin is distinguished by a five-pointed festoon. In the whale it is oblong transversely, and in the dolphin it is heart-shaped. The pupil of the crocodile resembles that of the cat; in the frog and gecko it is rhomboidal; and round in the tortoise, chameleon, and common lizards. In the ray among FISHES the upper margin forms several radiating slips like the branches of the palm-tree, gold-coloured without, dark within. In the dilated state these slips are folded backwards between the upper margin of the pupil and the vitreous humour; but when the eye is pressed they are erected, and close the pupil like a blind.

The motion of the pupil is voluntary in the parrot, and is indistinct in most of the FISHES.

The pupillary membrane is well known to exist in the fetuses of all the MAMMALIA; but it is not determined whether it is found in the chick of birds.

On the subject of the retina in the lower animals the

most important point is the structure of the melanoplectic or pectiniform membrane (pecten, marsupium nigrum) of BIRDS. In this class the optic nerve forms not a round disk as in the MAMMALIA, but a narrow white line, the margins and extremity of which are in continuity with the retina. Along this line is suspended a plicated or convoluted membrane, very fine and vascular in structure, like the or marsupium choroid, from which, however, it is quite distinct, and entering a depression of the vitreous humour almost like a wedge. Its vessels, which proceed from a proper branch of the ophthalmic artery, are distributed in a minute arborescent form, among the folds of which the marsupium consists; and from these vessels the black viscid pigment with which its folds are covered appears to be secreted. The plicæ or membranous folds vary in number. In the cassowary they are only 4; in the brown owl (S. aluco) 5; in the common owl, ostrich, Guiana macaw, and merganser, they are 7; in the flamingo 9; in the falcon and swan 11; in the vulture and goose 12; in the duck, large heron, woodcock, and coot, 13; in the stork and partridge 15; in the crane 17; in the pheasant 20; in the turkey 22; in the jackdaw 25; and in the thrush 28. According to the observations of the elder Soemmering, to whom we are indebted for these numbers, the number of folds, though variable in different species, is the same in the same. In most birds the folds are arranged in a pectiniform order. On the use of this organ different opinions have been entertained by Petit, Haller, and Home; but all of them are conjectural.

In the REPTILES and many FISHES, between the optic nerve and the retina is a small tubercle, from the margins of which the latter membrane appears to rise; and radiating fibres are perceived more distinctly than in most quadrupeds. In many other FISHES the connection of the retina with the optic nerves resembles that of BIRDS. Thus, in the salmon, trout, herring, mackerel, cod, dory, and moon-fish, the optic nerve, after passing through the Ruyschian tunic, appears to be parted into two long white processes, which, following the outline of this membrane parallel but not contiguous to each other, are connected with the retina by their opposite margins.

In all animals provided with ciliary processes the retina terminates at and is connected with the gray pulpy zone denominated ciliary ligament. In those without ciliary processes, as the FISHES, it terminates suddenly at the attached or large margin of the uvea.

In several of the reptiles the retina presents the yellow spot of Soemmering. The principal peculiarities of the humours have been already mentioned.

Of the appendages the most important are the lacrimal gland and nictitating membrane.

In the Ruminants the lacrimal gland consists of two or three bodies, each composed of granules, each provided with a separate short excretory duct. In the hare and rabbit, in which the gland is large, there appears to be only one excretory duct, which perforates the upper eyelid near its posterior angle.

A gland peculiar to certain species, and wanting in man, that of Harder, is situate at the external or nasal angle, and presents an aperture under the third or nictitating eyelid, from which issues a thick viscid fluid. It is found in the RUMINANTS, the RODENTIA, the PACHYDERMATA, and in the sloth genus.

The caruncle exists in the RUMINANTS as in man, and appears to consist of numerous aggregated follicles. It is wanting in the RODENTIA.

In the CETACEA, as in most animals which live under water, there is neither gland nor lacrimal passages; and they are represented apparently by lacunæ below the upper eyelid, which discharge a thick mucilaginous fluid.

BIRDS, though destitute of caruncle, have both lacy-

Comparative anatomy. Lacrimal glands and that of Harder, the latter large, oblong, and flesh-coloured, placed betwixt the levator and adductor, and discharging by a single canal, opening at the inner surface of the third eyelid, a thick yellow fluid. The lacrimal, which is small, round, and very red, is provided in general with two or three canals, which, though small, are distinct. In most of the GRALLÆ and PALMIPED Birds there is, in place of the lacrimal gland, a hard granular body, occupying the upper part of the orbit, and following in situation the curvature of the eye. It has, nevertheless, no visible excretory duct.

In the turtle there is a reddish granular lobulated body, of considerable size, extending beneath the temporal vault. In the tortoise, frog, and toad, there are two small blackish glands without apparent excretory ducts. Neither in Serpents nor Fishes has any glandular apparatus in the eye been recognised.

The third eyelid or nictitating membrane. Though in man and the monkey tribe the eyelids consist of two semilunar cutaneo-muscular folds, with a minute mucous duplicature at the nasal angle, the latter acquires such a development in the lower animals as to constitute a genuine third eyelid, often distinguished by the name of nictitating membrane. This duplicature is semilunar in shape in the Ruminants, EDENTATA, and PACHYDERMATA. In the rhinoceros it is thick and fleshy; but of this the CETACEA present no trace. In the BIRDS, on the contrary, in which the eye is covered by the elevation of the lower eyelid, which is also the largest, the third eyelid is large, and covers the eye like a blind drawn before it; yet it is in some degree translucent, for it is evident that birds see objects through the membrane. In the owl and goatsucker the eye is closed by the depression of the upper as well as the elevation of the lower eyelid.

Though the SERPENTINE reptiles are void of eyelids entirely, in the crocodile, tortoise, and BATRACHOID, there are three, as in birds, the third being vertical in the two former orders, and horizontal in the latter. In the SAURIAN and CHELONIAD, also, the third, which is translucent, moves from before backwards by means of a single muscle, and may cover the whole eye. In the lizard genus the eyelids consist of a circular veil drawn before the orbit, and perforated by a horizontal fissure, which is shut by a sphincter, and opened by a levator and depressor. The gecko has no movable eyelid.

Compound eyes of the insects. In insects, the eye consists of innumerable hexagonal surfaces, slightly convex, and mutually separated by minute furrows, containing fine hairs variable in length. Each of these hexagonal surfaces, which constitute a hard, elastic, very transparent membrane, may be regarded as a cornea or crystalline lens, convex externally, concave within, and thicker in the centre than on the margins. Immediately behind is an opaque, viscid coating, varying in colour in the different species, analogous to the choroid pigment of the vertebrated animals, and completely obstructing the transmission of light. Beneath this varnish are short, whitish filaments, corresponding in number to the corneal surfaces, and mutually joined like mosaic or tessellated pavement, separated only by the dark-coloured pigment, and which appear to correspond to the retina of the VERTEBRATA. Behind these again is a delicate, dark-coloured membrane, which appears to correspond to the choroid; and exterior to this is a membrane continuous with the optic nerve, and which seems to be a general retina, forming, by subdivision of its parts on the anterior part of the choroid, the divided or multiplied retina. This is the structure of what are named compound eyes. That of the simple eyes of insects is too minute to be accurately demonstrated; but analogy gives probability to the inference that they are not dissimilar.

In warm-blooded animals generally, that is, in the MAMMALIA and BIRDS, the labyrinth or essential part of the organ consists of three semicircular canals, with a globular enlargement to each (ampulla), a cavity common to these canals named vestibule (vestibulum), and a conical tapering canal, divided into two compartments by a longitudinal septum. This may be named the bilocular cone (conus bilocularis). These parts consist of membranous substance inclosed in the bony walls of the pyramidal or auditory bone. In all the MAMMALIA the bilocular conical canal is convoluted in a spiral form, and hence is denominated, as in man, the cochlea—a name, however, which is applicable to it in this class only.

The organ of hearing in the MAMMALIA consists of the same parts nearly as in man. In some, indeed, for instance the guinea-pig (cavia), and porcupine, the cochlea makes three turns and a half; and, conversely, in the CETACEA only one and a half. In most of the ZOOPHAGA, and in the hog, elephant, and horse, the cochlea is much larger in proportion than the semicircular canals; but in the hare it is small, and in the mole very small. In the CETACEA, while the cochlea is very large and fully developed, its spiral is on the same plane throughout; and the semicircular canals are so small, that their existence was long denied by Camper, till they were demonstrated by Cuvier in a fetal whale. In general the labyrinth of the MAMMALIA is greatly smaller in proportion to the head than in BIRDS.

This part, which is membranous, is inclosed in the solid compact substance of the temporal pyramid, so closely that its existence appears to be identified with the latter. Researches, however, on the labyrinth in the fetus of the MAMMALIA, and especially in those of whales, demonstrate the fact that it is in a completely membranous form, distinct from the bony inclosure; that in shape and constituent parts it exists previous to the bony inclosure; and that the latter is afterwards moulded round the different parts as they acquire their full development. It is also to be observed, that in the mole the semicircular canals are seen within the cranium without preparation, and the cochlea is merely inclosed in fine cellular tissue. In the bat family, also, both parts are seen without bony inclosure.

The tympanum forms a cylindrical or spheroidal cavity in Tympanum most of the MAMMALIA. In most of the DIGITATA the num. mastoid process consists of a slight prominence of the tympanum only as it is identified with the latter; but in the cavia, guinea-pig, hog, the Ruminants, and SOLIDUNGULA, it is represented by a long process of the occipital bone. In most of the ZOOPHAGA and RODENTIA the parietes of this protuberance, which are thin and hard, form by their separation a large cavity. In the hog only it is occupied by a firm cancellated structure.

All the MAMMALIA, except the ornithorhynchus, have the Tympanal tympanal bones as in man: the hammer (malleus), anvil bones (incus), orbicular bone and stirrup (stapes). The lenticular bone, which is rarely found in the adult, is probably only an epiphysis of the anvil. They are articulated with each other so as to admit of motion, and are moved by the same muscles as in the human subject—the internus mallei, externus mallei, laxator tympani, and stapedius. In the ornithorhynchus, however, there are only two tympanal bones.

In all the MAMMALIA, except the Cetaceous, the ear External is provided with a bony external canal (meatus); and aperture. most of the MAMMALIA, except the Cetaceous, have a cartilaginous funnel-shaped opening (concha) attached to the outer margin of the bony meatus, and which serves to

Comparative Anatomy. collect the sonorous vibrations, and direct them to the meatus. The other exceptions are among the INSECTIVORA, the mole, and some of the shrew genus; among the RODENTIA, the zenni or blind rat, and some of the rat-mole genus; among the EDENTATA, the pangolin or scaly ant-eater; and among the AMPHIBIA, the morse and several species of seal; and the ornithorhynchus paradoxus.

The tympanum of the CETACEA is peculiar. It consists of a bony plate, convoluted on itself like a buccinum, unless that the thick side, instead of containing a spiral cavity, is entirely solid. The opposite side is thin, with an irregular margin. The anterior extremity of the tympanum is open, and there commences the Eustachian tube, which ascends along the pterygoid process, and, penetrating the maxillary bone, terminates at the upper part of the nose. This direction of the tube and position of its orifice is so much more necessary, because, since these animals have no external bony meatus, and the ear-hole scarcely admits a pin, the vibrations of the air reach their organ of hearing entirely by the Eustachian tube, and because the Eustachian tube also in these animals conveys odorous impressions to the part in which the sense of smell appears to reside. The aperture by which it communicates with the nose is provided with a membranous valve, which prevents the water from entering when the animal expels it by his nostrils.

In BIRDS, of the three semicircular canals the vertical is largest, and obliquely directed forwards and outwards; the second is horizontal and turned outwards; and the third, which, like the first, is vertical, crosses the second, and is turned in the direction opposite to that of the first. The vestibule is small and nearly spherical. The bilocular cone, which is obtuse at the aper, is situate obliquely backwards and outwards below the inferior part of the cranium. The longitudinal septum consists of two narrow cartilaginous plates connected by a thin membrane. The posterior canal is short, and, as in the MAMMALIA, is separated from the tympanic cavity by the membrane of the fenestra rotunda; while the anterior, which is larger, communicates directly with the vestibule. The whole of these parts are inclosed, as in the MAMMALIA, in the compact bone of the pyramid.

The posterior and inferior walls of the tympanic cavity are formed by part of the occipital bone; the lateral aperture is large, and the cavity superficial; and its anterior part is closed by the posterior superior cornu of the quadrilateral bone and a membrane. The inner wall presents the two apertures—the oval or vestibular, and the round or cochlear. In this class, however, while the upper is round or triangular, the lower is distinctly elliptical,—a disposition the reverse of what is observed in man. The Eustachian tube or tympano-guttural canal is entirely osseous. The tympanic cavity contains only one ossiculum, consisting of two branches; the first attached to the tympanum, corresponding to the malleus; the second closing by an oval or triangular plate the vestibular aperture, and therefore corresponding to the stapes of the MAMMALIA. By Carus the incus is supposed to be represented by the quadrilateral bone.

The external meatus is short, and opened by a simple aperture, while the absence of external ear is compensated by a ring or zone of fine elastic feathers with thin barbs, between which the air passes very easily. In the owl tribe it terminates in a large cavity, the margins of which are covered by a smooth valvular fold of skin.

Comparative Anatomy. The ear of the REPTILES is remarkable for the last appearance of the bilocular cone, and the first of the saccular apparatus which is found in the Fishes. In the crocodile and lizard this part appears, as in Birds, in the shape of a conical tube, divided by a cartilaginous partition into a

double canal, one separated by the membrane of the round hole from the tympanic cavity, the other communicating with a membranous sac containing three very small friable stones, not harder than starch. There are also three semicircular canals of considerable size, each forming a large circumference. In the frog and toad, while the three canals form almost a complete circle, the sac contains an amylaceous friable stone; but the bilocular cone is no longer observed. In the salamander, also, in which the three canals form together a sort of equilateral triangle, the sac which is below contains a single amylaceous stone. The same arrangement is observed in the Cartilaginous Fishes, unless that the sac contains two amylaceous stones, nearly oval in shape, suspended in a gelatinous semifluid pulp. In the Osseous Fishes it is a little different. The three semicircular canals terminate in a membranous sac, which is divided by septa into compartments which contain one, two, or three small stones suspended in gelatinous fluid. These minute stones, however, instead of being soft, friable, and amylaceous, as in Reptiles and cartilaginous Fishes, are as hard as rock, and white as porcelain. These parts are situate on the sides of the cranial cavity, and are fixed to it by cellular tissue, vessels, and osseous or cartilaginous processes. This sac, in the fluid of which the extremities of the auditory nerve are distributed, is believed to correspond to the bilocular cone of the higher classes.

These membranous cavities are contained, in the bony fishes, in the general cavity of the cranium; and while only the middle of the canals is inclosed in the bone of the cranium, the extremities and the sac are quite free. The sturgeon, which belongs to the cartilaginous order, is the first in which the canals are entirely inclosed in the cranial cartilage; but even in this a membrane is interposed between the cranium and sac, which is free. In the ray and shark genera, again, these organs are entirely inclosed in the cartilage of the head.

The tympanic cavity, in like manner, is modified, and eventually disappears as we descend in the scale. Though present in the tortoise, crocodile, and lizard tribe, it is superficial and open; it becomes membranous behind in the RANINE tribe, and communicates directly with the back of the mouth; and in the SERPENTINE reptiles it entirely disappears, so that the handle of the osseous plate by which the oval aperture is closed is suspended in the soft parts with its free extremity below the skin, near the articulation of the lower jaw. In the lizard tribe, also, Cochlear the round or cochlear aperture is seen for the last time, aperture In the Chelonid, for instance, the Batrachoid, and the Serpentine, this aperture disappears, and the oval or vestibular alone is left; and in the salamander both disappear, and there is no communication between the external part of the cranium and the labyrinth. This arrangement is continued in the fishes. Vestibular aperture disappears

In the molluscous animals the labyrinthine membrane is a simple sac, globular or ovoidal, containing pulpy matter, in which is suspended a small body, which is osseous in the sepia and amylaceous in the many-feet (polypus), in which the filaments of the auditory nerve are distributed.

Our limits do not allow us to enter into the detailed description of the organ in the other INVERTEBRATED animals.

SECT. IV.—THE ORGAN OF TASTE.

Though the sense of taste is seated chiefly in the tongue in animals, yet that organ performs, in all the classes, so important a part as an instrument of prehension, that it cannot with much justice be distinguished by the former title only. In the present section, therefore, we must regard it as one of prehension as well as of taste.

In the MAMMALIA and BIRDS the tongue is a muscular organ invested by mucous papillated membrane, supported by a proper bone, the hyoid, which serves as a point of support in its various motions. In the RANINE REPTILES it is also muscular, attached to the margin of the lower jaw. In the salamander, however, it is attached as far as the tip, and is movable on the sides only. In the crocodile it is attached so generally, both by tip and margins, that it was long asserted that the animal was tongueless. In the stellio and iguana it is as movable as in the MAMMALIA; and in the scine and gecko to this property is added that of being bifid, or divided by a longitudinal notch into two pointed tips. In the ordinary lizard, tu-pinambis, monitor, &c. the tongue is remarkable for its great extensibility, and terminates in two long, flexible though semi-cartilaginous extremities. That of the chameleon is still more extensible, and forms, by a peculiar arrangement of vessels, a cup-like extremity. The tongue of the blind worm (anguis fragilis) and amphisbœna is also bifid at the tip. The cartilaginous fishes are void of tongue, while in the bony division of the class this organ is represented by a hard protuberance, attached to the middle branchial bone.

In some of the MAMMALIA, however, the tongue is not exclusively muscular. In the singularly long, extensible, and tortuous tongue of the giraffe, Sir Everard Home describes a peculiar arrangement of vessels, which he represents as a substitute for muscular motion. Though Sir Everard does not appear to understand the exact nature of this arrangement of vessels, all the circumstances tend to show that it is that denominated erectile. These vessels, from the account given, are large, numerous, and communicate freely; and it would be impossible to discover the reason of such a vascular system, unless for some purpose of this description. (Phil. Trans. Comp. Anat.) When the tongue is protruded it becomes perfectly black or bluish-black, evidently from the injection and detention of the blood in its elongated and anastomosing veins. By means of this mechanism the giraffe not only elongates the tongue to the distance of about twenty inches or two feet beyond the mouth, but twists it round the soft leafy twigs of the trees on which he feeds. It is not improbable that a similar vascular arrangement exists, though in less degree, in the tongue of the deer, and in the long projectile tongue of the animals of the ant-eater tribe, as the Tamanoir, Tamandua, &c.

The erectile arrangement is still more distinctly presented in the tongue of the chameleon. The researches of Mr Houston of Dublin show that the tongue of this animal consists of two parts,—a prehensile, which is anterior, and provided with a glandular apparatus for secreting the viscid fluid by which its tip is covered, and insects are entangled; and an erectile, which is posterior between the prehensile and the hyoid bone, in the form of a trellis-work of innumerable minute anastomosing blood-vessels, not very dissimilar to those of the cavernous body in animals generally, and inclosing a central tube connecting the prehensile portion to the hyoid bone. The effect of this arrangement is, that when the vascular network is injected with blood, the anterior part of the tongue is rapidly darted out at the insects on which the animal lives. The injection of these vessels, and the consequent projection of the tongue, is not independent altogether of the will of the animal; for the veins by which the blood is returned pass through a slit in the tendon of the internal cerato-maxillary muscles, which are always contracted in order to protrude the hyoid style, and thereby tend, by compressing the veins, to inject the erectile part, and project the tongue. (Trans. R. I. Acad. 1828, and Dublin Hospital Reports, vol. v. p. 487.) The same arrangement

is in all probability found in several of the lizards, the tongues of which, like that of the chameleon, are darted out suddenly, and become of a dark-blue colour at the moment of projection.

In all the MAMMALIA the tongue is invested by a papillated muco-villous membrane, in which the papillæ are of the same general characters as in man—granular, mushroom-like, or fungiform, tubercular or calycoid, and conical or acuminate. The only differences consist in the size and abundance of the fungiform papillæ, in the number of the calycoid and the mode of their arrangement, and in the shape of the conical papillæ and the mode of their termination. In the Ruminants especially, the conical papillæ are numerous, long, slightly incurvated, and each terminating in a horny but flexible style slightly incurvated backwards. The tongue of the dolphin and porpoise, examined even by the microscope, presents no distinct conical papillæ, but is covered by minute eminences, each penetrated by a small aperture.

In the tongue of the dog genus there is a ligamentous substance extended longitudinally from the hyoid bone to the tip of the member. This, which has been vulgarly distinguished by the name of worm of the tongue, and has been absurdly supposed to be the seat of hydrophobic rabies, is merely a central pillar of support for the muscular fibres to act with greater steadiness and effect, and which enables the animal to protrude and expand the tongue in lapping water or other fluids better than he could have otherwise done. A similar central ligament is found in the opossum.

The tongue of Brans is generally more or less horny, and almost cartilaginous. That of the woodpecker and wryneck is peculiar in consisting of two parts,—a basilar or posterior, loose and fleshy; and an anterior projectile, long, smooth, acuminate, and covered laterally with four or five stiff spines directed backwards, which make the organ a sort of barbed arrow. The soft, loose, or basilar part of the tongue contains the aperture of the glottis; and the surface is covered with minute spines pointed backwards, and each of which is placed in the centre of a fleshy papilla.

As a prehensile organ of very singular construction, the trunk of the elephant deserves particular notice; and it cannot be more conveniently introduced than under the present section, since it is used not only to convey food, but drink, into the mouth. The trunk may be described as a cylindrical tubular organ, consisting of integument, a sort of fibro-cartilage, muscles, fat, and a membrane of villous character internally. This tube contains two long canals continued from the nostrils, parallel to the axis of the trunk, and separated throughout by a partition of adipose substance about two fifths of an inch thick. From the extremity to the middle part of the intermaxillary bone, in which the tusks are fixed, these canals are nearer the anterior-superior than the posterior-inferior part of the tube, the latter wall being thickest; and their diameter is the same throughout. At this part they undergo a sudden incurvation, approaching the anterior surface of the intermaxillary bone, and form a semicircular bend with the convexity turned forwards. Here also they are so narrow that, without a muscular effort on the part of the animal to dilate them, fluids could not ascend beyond this point; and hence this forms the only valvular contrivance, either to impede the progress of fluids upwards, or to propel them downwards, at the will of the animal.

Above this curvature each canal is dilated before the upper part of the intermaxillary bone, and again is contracted where it bends back to enter the bony nostril; and the curvature is protected before by the nasal cartilage, which is oval, convex in the male, and flat in the female.

Both canals are lined by a dry, greenish-yellow coloured membrane, marked with superficial intersections (rugæ), inclosing rhomboidal spaces, and some venous branches.

Though the muscular fasciculi of the trunk are numerous, they may be referred to two orders,—those forming the substance or inner part of the organ, and those by which it is invested. The former, which are transverse, and cut the axis in different directions, consist of numerous small muscular packets proceeding in various directions, some running from the inner membrane to the circumference of the tube, others directly from right to left, and others crossing the two former obliquely. All these little muscles are inclosed in cellular tissue, containing white homogeneous fat; and all of them terminate in slender tendons, some of which cross the layers of the longitudinal muscles in their course to the external covering, while others are attached to the internal membrane. Cuvier calculates the number of these minute transverse muscles in the trunk of the elephant to be not fewer than 30,000 or 40,000. (Plate XXXVII. fig. 13.)

The longitudinal muscles, which are external, may be distinguished into anterior, posterior, and lateral bundles. The first extend from the anterior surface of the frontal bone, above the nasal bones and cartilages, in parallel bundles, connected by tendinous intersections downwards on the trunk. The posterior extend from the posterior surface and inferior margin of the intermaxillary bones, and form two layers which meet on the median line along the lower surface of the trunk. The lateral muscles form two pairs, one of which, descending between the anterior and posterior muscles to the middle of the trunk, may be regarded as a continuation of the orbicular muscle of the lips, or the representative of the nasalis labii superioris; while the other, which is attached to the anterior margin of the orbit, and is expanded over the root of the former, may be supposed to correspond to the levator of the upper lip.

The whole of these muscles are supplied by a very large branch of the infraorbital or second branch of the trifacial nerve, which, entering on each side between the lateral and superior muscle, is distributed to the whole of the trunk.

With such a construction, it is not difficult to understand the numerous motions of the elephant's trunk. While the longitudinal muscles are employed either to shorten the tube, to bend it upwards or downwards or to the side, or by means of the tendinous intersections to give it peculiar inflections, it is manifest that the transverse ones, which act as antagonists to the longitudinal, may also either dilate or close the canals, or incurvate or alter the direction of particular parts.

The foot of the Lacerta Gecko and the house-fly presents a prehensile apparatus of peculiar construction for walking along surfaces, in opposition to the action of gravity. In the former animal the plantar surface of each toe presents sixteen transverse slits, leading into an equal number of pouches, which by means of appropriate muscles are capable of forming an equal number of vacua, so that the atmospheric pressure is employed with muscular effort to support the animal in his unnatural position. A similar apparatus is found in the upper surface of the head of the sucking fish (echeneis remora); and something approaching to it, though less distinctly, in the foot of the walrus. (Home, Phil. Trans. 1816.)

CHAP. IV.—COMPARATIVE ANATOMY OF THE ORGANS OF VOICE.

Under this head our limits allow us to mention very few circumstances.

In the American long-tailed monkeys (sapajous) the

cuneiform cartilages form, by means of adipose cellular tissue, before the upper extremity of the ventricle of the glottis, a large cushion like a spherical segment, which, touching that of the opposite side, causes the air to whistle through the canal in its course to the mouth, and occasions the flute-like voice of some of these animals, as the weeper (s. apella) and the capuchin (s. capucina). In the howler (s. seniculus), so remarkable for its morning and the howling evening yelling, though the larynx is similar in general characters to that of the common sapajous, in having the two rounded cushions before the ventricles, the hyoid bone is arched in the form of a spherical chamber, with a large quadrilateral aperture, and each ventricle opens into a membranous sac, lying between the epiglottis and the adjoining wing of the thyroid cartilage. The air, which passes between the vocal chords, is therefore partly impelled into this osseous and elastic cavity of the hyoid bone, and probably by its resonance in this situation gives the voice of these animals the deep-toned howl by which they are known in the American forests.

Among the ZOOPIAGA, in the dog the cuneiform cartilages are large, the arytenoid small, the vocal chords well marked, and the ventricles deep. In the feline tribe the anterior ligaments, though destitute of cuneiform cartilages, are thick, and separated from the back of the epiglottis by a broad, deep furrow. The posterior ligaments, though neither free nor sharp-edged, are distinguished from the anterior by an appearance of greater firmness, more regular fibres, and by an intermediate furrow. The approximation of the anterior ligaments towards the glottis forms a sonorous vault, in which the air may be forcibly vibrated by the posterior. In the bear the cuneiform cartilages assume the shape of styles, and their posterior extremity forms a distinct eminence, not above, but without the arytenoid cartilages, while the ventricles are merely deep fissures.

The kangaroo has neither cuneiform cartilage, anterior ligament, nor ventricle; and it may even be said to be void of vocal chord, while the margins of the glottis are much separated in the middle. This arrangement appears to indicate that the animal is almost destitute of voice. In the opossum, in which there is merely a small inferior ligament susceptible of tension, voice is limited to a whistling sound.

In the SOLIDUNGULA, in which the cuneiform cartilages are completely concealed by the mucous membrane, there is neither superior ligament nor proper ventricle; but an aperture in the lateral wall of the laryngeal membrane, above the vocal chord, leads into a large, oblong, sinuous cavity, situate between this membrane and the thyroid cartilage, and covered chiefly by the thyro-arytenoid muscles, by which it may be compressed; and above the anterior commissure of the vocal chords, or below the base of the epiglottis, is an aperture on the mesial plane, leading into a cavity below the vault formed by the anterior margin of the thyroid cartilage. This cavity, which may be named the infrathyroid, is superficial in the horse, of the ass, and its aperture is large; while in the ass, with a small, round aperture, the cavity is large, capacious, and globular in every direction, and allows the latter animal to make his voice re-echo in the singularly harsh sound denominated the bray. Conversely, though the lateral cavities are equally large in both animals, the apertures in those of the ass are small, round, and situate nearer the epiglottis than the vocal chord, while those of the horse are large, oblong, and situate immediately above the vocal chord on each side. On the latter peculiarity appears to depend the neigh of the horse.

In the CETACEA we recognise neither vocal chords nor glottis, that is to say, an aperture variable in size ac-

Comparative Anatomy. cording to the will of the animal; but the superior part of the trachea, which represents the larynx, forms a hollow pyramid or funnel, rising into the posterior part of the nostrils, in which only it opens, while on the sides is left a passage for the food. This pyramidal funnel is formed by an elongated triangular cartilage, corresponding to the epiglottis, attached by membrane to the arytenoid cartilages, which also take the shape of scalene triangles, with the small side connected to the cricoid cartilage. Strictly speaking, therefore, the CETACEA have no larynx, and probably no vocal organ; and the superior part of the trachea, with the nostrils, serves merely to admit the atmospheric air for the purposes of respiration.

BIRDS are distinguished by possessing not only a glottis or laryngeal aperture similar to that of the MAMMALIA at the upper end of the trachea, but a second, denominated the inferior glottis or larynx, at the lower end, near its bifurcation. The former, which is composed of four cartilages, or six, according as the cricoid consists of one or three pieces, on the middle of the posterior part of which is a small round bone, articulated with two oblong longitudinal bones, parallel, and forming between them in the posterior wall of the windpipe a longitudinal slit, susceptible of approximation by means of muscles, is intended merely to regulate the admission of air into the windpipe, or its expulsion from that tube, and to close more or less accurately its superior orifice.

The inferior larynx consists of a membrane projecting from each side of the inferior aperture of the trachea. This aperture is divided into two, occasionally by an osseous anterior-posterior middle bar, occasionally by the angle at which the two bronchial tubes unite. Since the first bronchial arc has the same curvature as the last tracheal ring, the second and third, which are arcs of larger circles, are less convex without, but more prominent within, than the former. Over this prominence the tracheo-bronchial membrane forms a fold, which, half closing on each side the inferior tracheal aperture, forms a plate susceptible of vibrating by the motion of the air, and producing sound. This apparatus, which constitutes what is named the inferior larynx, or rather glottis, is of two kinds, one void of proper muscles, the other provided with muscles.

In the former kind of larynx the state of the glottis is altered only by those muscles which depress and elevate the trachea. The depressors are two pairs, the sternotracheal and the glosso-tracheal, the latter attached to the bifurcated bone and trachea. There are no proper elevators; but the windpipe is raised by the mylo-hyoid muscle through the ligaments which connect the hyoid bone to the superior larynx. In the quiescent or relaxed state, and while the trachea is depressed, the bronchial rings approximate, and the second and third even gliding below the first, the glottis may be elongated. When the trachea is elevated by these pairs of muscles, the bronchi are at the same time dragged upwards, and the second and third arcs are separated from the first; and while the external prominence of the glottidal membrane diminishes in length, its tension is augmented. These forms of larynx without proper muscles may yet be subdivided into two sorts, as they have or have not lateral pouches, membranous or osseous. These are observed in the male duck (anas) and the merganser (mergus), but never in the female; and to this perhaps the harsh and deep tone of the voice of the male bird is to be ascribed. The larynx without muscles and without pouch is observed in all the gallinaceous order without exception.

The forms of larynx provided with proper muscles may be distinguished into three subdivisions.

The first, which has only one proper muscle on each side, is observed in the whole of the falcon genus, e.g. the

eagle, hawk, falcon, buzzard, sparrow-hawk, and goss-hawk; in the owl genus and the majority of the waders and swimmers, as the heron, bustard, woodcock, lapwing, rail, coot, gull, cormorant, and some of the passerine birds. In these birds, in which the motions of the lower larynx are necessarily limited, the voice is not variable or extensive in its notes.

The second form of larynx has three pair of proper muscles, a constrictor of the glottis, an auxiliary constrictor, and a laxator or opener of the glottis. This kind of larynx is observed in the whole of the parrot genus and psittacoid birds generally, as the toucan, macaw, calao, &c.

In the third kind of musculo-membranous larynx there are no fewer than five pair of muscles, the longitudinal levator of the demiannular cartilages, the posterior levator of the same cartilages, the small levator, the oblique levator, and the transverse levator. This quinquemuscular larynx is found not only in all the birds properly named whistlers or warblers, as the nightingale, hedge-sparrow, blackbird, thrush, goldfinch, lark, linnet, canary, chaffinch, &c., but in others whose tones are more monotonous, as the swallow, sparrow, stork, crossbill, &c., and even in some the tones of which are harsh and positively disagreeable, e.g. the jay, magpie, crow, raven, &c.

The differences remarked in the notes of these three divisions of birds with the quinquemuscular larynx depend not so much on anatomical peculiarities as on the timbre of their larynx, and on the mobility of the trachea in relation to the larynx, and on the tracheal membrane having dilatations and contractions.

On the whole, the inferior larynx of birds is to be regarded in three lights: 1st, As the reed of a wind-instrument, like a hautboy or clarionet, in which the notes vary as the lower glottis varies in its position to the windpipe; 2d, as an instrument susceptible of uttering different tones, according to the distance between the mouth-piece and vent, or as the windpipe is elongated or shortened; and, 3d, as an instrument capable of uttering different notes by varying the diameter of the mouth-piece, or as the superior glottis is widened or contracted.

The only bird in which the inferior larynx is wanting is the vulture.

The vocal organ of the REPTILES consists of the superior larynx only, analogous to that of the MAMMALIA. This is a cartilaginous apparatus, composed in general of five distinct pieces at least in the large individuals of the SAURIAL genera, as the crocodile and alligator, and forming a broad cavity behind, before a narrow slit, bounded by two vertical pillars. The glottis, however, is entirely membranous; and there are neither vocal chords nor ventricles. There are nevertheless two muscles, one for opening and another for shutting the glottis. When these act, and the air is made to vibrate against the anterior pillars, it gives a slight whistling sound only. In the iguana, tupinambis, lizard, tortoise, and serpent, the arrangement of parts is nearly the same; and these animals, therefore, can utter only slight hissing sounds. In the chameleon the pillars are furnished with a tense, vibrating membrane, a fleshy tubercle which contracts the glottis, and a membranous pouch opening below, between the lower laryngeal cartilage and the first tracheal ring. In the frog tribe, so remarkable for their croaking noise, the vocal chords are large and prominent. The males have also two membranous pouches, opening by a small aperture, not in the larynx, but deep in the lateral part of the mouth. When the frog croaks these pouches are inflated, and swell the skin on each side below the ear. Though these sacs are wanting in the female frog, and toad both male and female, as well as the tree-frog, there is beneath the throat a single pouch on the median plane.

A general view of the nervous systems in different classes of animals shows that the only common part is an azygous tubercle, situate at the anterior extremity of the spinal chord, connected by means of two lateral bundles or peduncles to the rest of the system. This mass, which corresponds to what is denominated the cerebellum in man, is connected in the vertebrated red-blooded animals with several pairs of tubercles, forming generally a larger mass than itself, and connected to the rest of the system by two longitudinal bundles or limbs, which mingle with and intersect those of the cerebellum. These anterior-superior tubercles, which constitute the brain proper, present numerous modifications in figure, disposition, and magnitude, and even in presence, in the different orders of the animal world.

In the vertebrated animals the brain or central part of the nervous system consists of the vertebral or funicular portion, named the spinal chord, and the cranial or cephalic portion, properly denominated the brain.

The funicular portion has the shape of a cylinder flattened on its superior or dorsal and inferior or sternal surfaces. It consists of two similar chordiform productions, united on the mesial plane, and marked at the line of junction by linear longitudinal furrows, of which the superior is most deep and distinct. This, however, which was observed by Blasius in several animals, is denied by Bellingeri, who asserts, from various observations, that though in some parts the superior furrow penetrates more deeply than the inferior, in the general course of the chord the latter is the deepest. (De Medulla Spinali Amotations Anatomico-Physiologicæ. Auctore C. F. Bellingeri, Aug. Taurinorum, 1823.) This discordance appears to depend on the circumstance that the sternal or inferior furrow is in truth deeper and more distinct in man, the monkey tribe, and a few MAMMALIA, than the dorsal; whereas in most others, as the coati, raccoon, horse, &c. the dorsal furrow is of the same depth as the sternal; and in the mole, murine, and leporine genera, it is deeper and more distinct than the sternal. In BIRDS and the cartilaginous FISHES the sternal furrow is deepest; in Reptiles the dorsal and sternal are nearly equally deep; and in the osseous FISHES the dorsal is the deepest. On each side of the chord, also, there is a slight longitudinal depression commonly called the lateral (sulcus lateralis).

The breadth and thickness of the spinal chord vary in different regions. The lower cervical portion is in general the broadest and thickest; at the upper dorsal region it is somewhat more slender, becoming thicker again at the lower dorsal region; and after again becoming more slender in the lumbar region, it is expanded into a filamentous brush-like termination of different lengths in different classes, and even in different orders. These enlargements at the superior dorsal and lumbar regions are believed by Serres to correspond with the origins of the thoracic and pelvic extremities; and each swelling he represents to predominate over the other, as the animal habitually employs the one or the other kind of members. It is remarkable in illustration of this principle, that in the CETACEA, which are void of pelvic extremities, the lumbar enlargement is wanting; and in the amphibious MAMMALIA, the pelvic extremities of which are feeble, this enlargement is also inconsiderable.

The spinal or funicular brain consists of white cerebral matter on the surface and gray matter in the centre, the former being most abundant, unless in the sacral region, where both are nearly equal in quantity, or the gray rather predominates. In the MAMMALIA the gray matter is in greater proportion to the white than in the other three

classes, in which the white matter progressively aug-
ments.

The sternal and dorsal longitudinal lines indicate the original formation of the chord in two lateral portions, with an intermediate cavity denominated the spinal canal. This exists only during the formation of the chord in the human fetus, in some of the MAMMALIA when adult, and in the other three classes. Its disappearance in the human fetus, and in that of various MAMMALIA, is represented by Serres to depend on the progressive deposition of gray cerebral matter on the inner or central surfaces of the component pillars of the chord. Though obliterated in man, in whom the gray matter is abundant, it is not hermetically sealed in the monkey tribe, in which some traces of it are left. In the AMPHIBIA and CETACEA it is larger than in the monkey tribe; its diameter augments in the CARNIVORA, feline, canine, and ursine, in which the gray matter is thinner than in the former; and in the RODENTIA it is said to be largest of all among the MAMMALIA. Lastly, in the birds, reptiles, and fishes generally, in which the gray matter is scanty and the white predominates, the spinal canal is large and distinct.

The lateral regions of the spinal chord are connected with a double series of nerves by means of two rows of nervous filaments, an anterior and posterior, separated by a longitudinal membrane of fine white tissue, with a serrated or festooned border. This membrane, which is named the denticulate, is the same in the MAMMALIA and BIRDS. That these nerves do not issue from the spinal chord, must be inferred not only from the phenomena attending the original development of the nervous system, but especially from what is observed of their comparative size, and that of the chord in the inferior classes. In that of fishes especially, while the chord is small and slender, and by no means fills the vertebral canal, it is remarkable that the nerves which supply the voluntary muscles are exceedingly large. Thus, in many both of the cartilaginous and osseous divisions, as the sturgeon, dog-fish, ray, wolf-fish, cod, &c. the nerves which supply the pectoral fins are large, broad chords, two or three of which seem to contain more substance than the whole spinal chord itself.

The chord is expanded at its cephalic end into a thick Bulb of the eminence denominated the spinal bulb, the surface of which presents three pairs of eminences. These eminences are developed in various degrees in the different classes and orders. In some of the MAMMALIA the anterior pyramids or pyramidal bodies are more distinct than in man, for instance the ape tribe, the CETACEA, CARNIVORA, Ruminants, and RODENTIA. They are small in most Birds, all the Reptiles, and the Cartilaginous FISHES. In the osseous division they assume the appearance of two parallel chords at the base of the brain. The olivary bodies, less prominent in the ape than in man, are still less so in the CETACEA, and progressively through the AMPHIBIA, CARNIVORA, RUMINANTIA, RODENTIA, and INSECTIVORA, and the other three great classes of vertebrated animals. Conversely, the restiform bodies or posterior pyramidal eminences, from man through the ape tribe, the CETACEA, AMPHIBIA, to the Ruminants, CARNIVORA, and RODENTIA, increase in size.

Though the brain of the MAMMALIA presents the same parts, and is arranged nearly in the same order, as that of man, it varies in its proportions to the rest of the body; in its proportions to the cerebellum and spinal bulb; in general figure; in the presence, absence, and number of convolutions; in the configuration of its central surface; in the communication of its central with its external surface; and in the manner of its connections with the cerebral nerves. As it is impossible in this sketch to examine all

Comparative Anatomy. these circumstances fully, we shall confine our attention to the notice of a few only.

It is not easy to ascertain the proportion of the mass of the brain to that of the rest of the body. Excluding as much as possible the ordinary sources of fallacy, in small animals the brain is proportionally larger; yet in this respect man is surpassed only by a small number of animals, habitually lean, and with little muscle, as bats, small birds, &c. While the proportion of brain in man to that of the whole person varies from a 22d to a 35th part, that of the monkey tribe varies from a 22d to a 42d part; and in the baboon it is only the 104th part of the body. Among the MAMMALIA, the RODENTIA have in general the largest proportion of brain, and the PACHYDERMATA the smallest; and while the hare has a brain about the 300th part of the size of the body, that of the elephant, the most sagacious of animals, is about the 500th part of the size of his body. It is also remarkable, that while the brain of the horse is only a 400th part of the size of his body, that of the ass amounts to a 254th part. The Reptile brain becomes excessively small, that of the turtle being rather more than the 5000th part the size of his body; and in some of the Fishes, not all, it appears to attain the maximum of decreasing proportion, that of the tunny being so small as the 37,000th part of his body, while the brain of the carp is so large as to approach the proportion of the elephant. It may be doubted whether, under such circumstances, any precise conclusions can be drawn from results so variable and so little to be expected.

The proportional weight of the brain to that of the cerebellum is, excepting in the case of one species of ape, the simiri, greater in man than in any other animal. The ox is equal to man in this respect, and the dog approaches him. The animals most remote are the RODENTIA, as the beaver, rat, and mouse, &c.

The convolutions, which are so numerous and so deep in man, diminish both in number and size in the QUADRUMANA and CARNIVORA, and are nearly obliterated in the RODENTIA. In the UNGULATED animals, however, and especially in the Ruminants and the horse, the convolutions are numerous; and even in the dolphin among the CETACEA, they are numerous and deep. In all the MAMMALIA the cerebellum is foliated.

On the whole, the peculiar character of the brain of man and the ape family consists in the existence of the posterior lobe and digital cavity. The brain of the ZOOPHAGA is remarkable for the small size of the nates or anterior pair of the bigeminate eminences in proportion to the testes or posterior pair. In the RODENTIA the organ is distinguished for the large size of the nates, and the want of superficial nature of the convolutions. In the UNGULATED division of animals, i. e. PACHYDERMATA, RUMINANTIA, and SOLIDUNGULA, the brain is remarkable for the large size of the nates combined with the number and depth of the convolutions; while that of the CETACEA is remarkable for its height and breadth, and the want of olfactory nerves. It is further to be observed as a general distinction between herbivorous and carnivorous or zoophagous animals, that in the former the nates are larger than the testes, whereas in the latter the testes are largest. Lastly, Man and the QUADRUMANA are the only animals which possess genuine olfactory nerves. In the other quadrupeds they are represented by the mammillary processes of the ancients; and in the CETACEA they have not yet been unequivocally demonstrated.

The brain of BIRDS is at once recognised by consisting of six distinct tubercles, two representing the cerebral hemispheres, two representing the optic eminences, one the cerebellum, and one the bulb of the chord. The hemispheres are void of convolutions, but the cerebellum

is marked by transverse parallel striæ corresponding to the laminae of the mammiferous brain. There is neither middle band (corpus callosum), vault, nor septum. The ceiling or vault of the aqueduct or passage from the third to the fourth ventricle is not, as in the MAMMALIA, surmounted by the bigeminate eminences, but is merely a thin plate corresponding to the valve. Each optic eminence contains a cavity communicating with the others by the Sylvian aqueduct. The anterior eminences (corpora striata) are not striated with alternate white and gray matter, as in the MAMMALIA. Between the anterior and the optic eminences are four rounded tubercles, best seen in the ostrich, which are to be regarded as entirely heterogeneous to the structure of the mammiferous brain, and connecting the cerebral structure of BIRDS with that of Reptiles and Fishes, in which also these tubercles are observed.

The Reptile brain is smooth and unconvoluted. The optic eminences, which are situate behind the hemispheres, are uncovered, and contain a ventricle communicating with the third. At the extremities of the latter are the anterior and posterior commissures, but there is neither soft commissure nor bigeminate eminences. The hemisphere presents an anterior eminence, which, however, in the brain of BIRDS is unstriated. The cerebral valve is, like that of BIRDS, unsurmounted by bigeminate eminences.

In the class of FISHES the structure becomes still more simple. The tubercles of which the brain consists are placed in a row; and their increase in number only demonstrates the decomposition of the organ, and its resolution into simple integrant parts. The two representing the hemispheres are ovoidal, unconvoluted, and contain a ventricle, in which is seen the eminence analogous to the striated bodies. The optic eminences, situate beneath the hemispheres, though small, contain each a cavity, as in the two oviparous classes already noticed. Lastly, there are in several genera, under the common vault of the hemispheres, occasionally two, occasionally four tubercles, variable in shape and proportions, but which would be analogous to the bigeminate eminences, were they not, like those already mentioned in BIRDS, situate before and above the optic chambers. In the cartilaginous fishes, in which these tubercles are not observed, the anterior or striated eminences are obliterated. The cerebellum does not cover the fourth ventricle.

Behind the cerebellum are two tubercles, which in the ray give origin to the fifth pair, and are very distinct in the pike, trout, salmon, and perch. These tubercles are peculiar to this class.

The cavities in the interior of the optic eminences in BIRDS, Reptiles, and Fishes, are observed in the foetal brain of the MAMMALIA during its early growth.

It is almost superfluous to mention, that, in the two warm-blooded classes, Mammalia and Birds, the brain, with its investments, fills completely the cranial and vertebral cavities. In the Reptiles, however, in which the brain does not approach the cranial walls, we remark the first departure from this arrangement; and in the Fishes it is so completely violated that the brain and chord occupy but a small proportion of the cranio-vertebral cavity; and between the former and the osseous walls there is a quantity of fine but very loose filamentous tissue, containing in its cells a large quantity of pellucid fluid. Though this arrangement gives this the appearance of a white jelly-like substance, it is not gelatinous, as is generally represented, but merely a pellucid fluid, sometimes pale straw-coloured, occasionally with a reddish tint, contained in numerous communicating cavities of a tissue which appears to represent the arachnoid of the warm-blooded animals.

The pia mater in the reptiles and fishes is reduced to a filamento-vascular web, accompanying the blood-vessels. The dura mater undergoes some peculiar modifications in different orders. In the duckbill a bony plate is contained between the lamina of the falx; and the same structure is found in the porpoise, perhaps in the other CETACEA. An osseous tentorium with a quadrangular aperture is found in the coatta and marten, and the feline and ursine families; and an osseous partition consisting of three parts is found in the dog, horse, Cape ant-eater (orycteropus), the wombat, and the seal. The bony tentorium is also found in the woodcock and others of the feathered class. In the red but cold-blooded animals the dura mater forms neither falciform nor tentorial process.

On the nerves or ramified chords of the nervous system a few words must suffice. In fishes the tenth or pneumogastric consists not of a common trunk, but of three orders of filaments, the first and largest of which are distributed to the gills, and correspond to the pulmonary

nerves of the MAMMALIA; the second, slender, are distributed to the muscles of the tongue and the surface of the oesophagus; and the third terminate in a large nerve which traverses the body longitudinally immediately beneath the lateral line. The phrenic nerve is wanting in birds, reptiles, and fishes.

In the MOLLUSCA the nervous system consists of a number of whitish cerebral masses distributed in different parts of the body, with one or two more conspicuous than the rest, and supposed therefore to represent the brain, placed transversely over the oesophagus, which it encompasses with a nervous collar.

In the ARTICULATA the nervous system consists of two long chords extending along the belly, and expanded at various intervals into gangliform knots or enlargements. The first of these, which is situate on the oesophagus, rarely exceeds the others in size. Among the ZOOPHYTES hitherto examined the nervous system assumes either a radiated or an arborescent form.

Though genuine teeth are found in three classes of animals only, viz. the MAMMALIA, the REPTILES, and the FISHES, yet all the orders of these classes are not provided with teeth. Thus, among the first class, the ant-eater tribe, the pangolin, the echidna and ornithorhynchus, and the whales—among the second the chelonians—and among the third the sturgeon—are altogether destitute of these organs. In all the invertebrated classes, the jaws, when present, are provided with notches varying in number. The echinodermata alone have genuine teeth, inserted in a mechanical apparatus very different from ordinary jaws.

Though in general structure the teeth of the lower animals resemble those of man, in some respects they differ considerably. These varieties consist either in some change or modification of the constituent parts of the teeth, or in the addition of some other substance to those parts.

The first variety to be noticed is of the former description.

Though in the QUADRUMANA and ZOOPIAGA the bony matter of the teeth is quite similar to that of man, in other orders this substance appears in the form of a very hard, compact, and more regularly fibrous substance than bone, and to which the name of ivory (ebur) is applied. It is chiefly in the canine or tearing teeth that this substance is found to represent the bony pillar of the teeth; and it is principally among the PACHYDERMATA, and some of the AMPHIBIA and CETACEA, that this change is observed.

The ivory of the elephant is the most tender, and that which most rapidly becomes yellow on exposure to air. It is readily distinguished from the ivory of other animals by the curve lines which radiate from the centre to the circumference of the tooth in various directions, and which form by intersection regular curvilinear lozenges.

The ivory of the hippopotamus is greatly harder and whiter, and is on that account preferably employed for the preparation of artificial teeth. A transverse section of this substance shows striae extremely delicate and regular. In this animal, also, not only the canine but the incisor teeth consist of this substance. The tusks of the Ethiopian boar (sus Ethiopicus) consist of ivory similar to that of the hippopotamus. In those of the ordinary boar, though no striae are recognised, there is sometimes a mix-

ture of brown substance disposed in layers. The ivory of the teeth of the morse, though void of striae, is compact and susceptible of polish nearly as brilliant as that of the hippopotamus; and its character is, that the central pillar of the tooth consists of minute round grains, indiscriminately aggregated, like pebbles in puddingstone. The axis or pillar of the molar teeth of this animal, which are without internal cavity, consists of similar minute grains. The ivory of the dugong is homogeneous and without striae. That of the teeth of the white whale or cachalot resembles the bone of human teeth in its satin-like appearance. The ivory of the tusk of the narwal is very compact and homogeneous in appearance.

The most singular structure of teeth among quadrupeds is observed in those of the Cape ant-eater (orycteropus). The teeth of this animal, which have the appearance of two cylinders conjoined, consist of an infinite number of minute straight parallel tubes, so that their transverse section resembles that of a rush. As these tubes are closed only at the triturating surface, it is there only that the tissue of the tooth is compact; and when the enamel is worn, the upper orifices of these tubes begin to be exposed. There is, therefore, no general cavity in the interior of the tooth. These teeth are also void of root. A similar tubular structure is observed in the two molar teeth of the ornithorhynchus, and in the teeth of some fishes.

The enamel (lamella vitrea, cortex striatus) presents peculiarities in the lower animals, as well as the bone of the tooth.

While the enamel of the human tooth is confined to the crown, in several of the lower animals, as the morse, it envelopes the tooth all round; and in the molar teeth of this animal, which, indeed, are void of cavity, it is thicker under the root than at the crown. A similar arrangement is observed in the old or adult teeth of the cachalot, which, when their cavity is obliterated by the full deposition of osseous substance, are also covered with enamel below.

The texture or constitution of the enamel is best seen in the grinders of the elephant. The section of a tooth in the germ exhibits fibres similar to those of asbestos or fine velvet. The fracture of the enamel is more distinctly fibrous than that of the bone, and the fibres are everywhere perpendicular, or nearly so, to the surface of

Comparative Anatomy. the tooth. The hardness of this substance may be inferred from the fact that it strikes fire with steel. These component fibres, however, are not always rectilinear. Most frequently they describe curves with the convexity of incurvation towards the crown and the concavity towards the root. This arrangement at least is observed in the ruminants. The distinction between the enamel and bony matter is recognised by a gray line, and another whiter which belongs to the latter substance.

The enamel varies chiefly in thickness in different animals. The tusks which project from the mouth are generally observed to be less white, less hard, and more similar to bone or ivory than the other teeth; and on this account, probably, the existence of enamel has been denied in the tusks of the elephant. It is nevertheless certain that the external layer of these tusks presents radiating fibres, though it is by no means so hard, or possesses the same grain, as the enamel of the other teeth. Enamel is more apparent, though thinner, in the tusks of the morse, dugong, and boar; and it is quite as distinct in those of the hippopotamus as in the other teeth of that animal. Lastly, the enamel of the teeth of the cachalot, which is very thick, shows in its section only striae parallel to the surface of the osseous substance.

Teeth may be distinguished according to the mode in which their component tissues are arranged into three sorts. 1st, When the enamel invests the axis all round, and does not penetrate the latter, the tooth is said to be simple (dens simplex). Such is the character of the human teeth, and those of the QUADRUMANA and ZOO-PHAGA, and several other animals, and all the reptiles.

Compound teeth. 2d, When the enamel is folded as it were round the bony part, but without enclosing it, so that the latter forms a continuous band several times folded on itself, and sections of the tooth in every direction divide repeatedly the component substances, the tooth is said to be compound or complex (dens multiplex vel compositus). A good example of this structure is seen in the grinders of the elephant. 3d, When the base or root of the tooth is simple, and the folds of the enamel and bone penetrate only to a certain depth, they are said to be semicompound. Examples of this modification of arrangement are seen in the grinders of the ruminating animals.

In the compound, and part of the semicompound teeth, the enamel is covered by a third substance; and as the latter is arranged, especially in the former sort, so as to leave intervals between it and the next layer of enamel, this substance serves to fill all these intervals, and consolidates the component lobes of the tooth even before their osseous parts are united below. This substance, which is denominated by Cuvier cement, by Tenon cortex osseus, and by Blake crusta petrosa, though less firm than either bone or enamel, is dissolved by acids more slowly than the former, and sooner becomes black in the fire. In the teeth of the elephant and cabiai it forms half their mass at least. In most genera it presents no apparent organization, and resembles a sort of crystalline tartar incrusting on the tooth. In the cabiai, however, it presents numerous pores very regularly arranged. Tenon was of opinion that it arose from ossification of the membrane which enveloped the tooth; Blake ascribed it to deposition from the opposite surface of the enamel membrane; but Cuvier ascertained that it is deposited by the same membrane and the same surface as the enamel. This accurate observer found, on inspecting the germs of the teeth of the elephant, that when the internal membrane of the dental capsule has deposited the enamel, it undergoes a change of structure, and becomes thick, spongy, opaque, and reddish, to furnish the cement, which is then deposited, not in regular crystalline fibres, but in random drops.

The teeth of the Reptiles consist of hard, compact, osseous matter, invested by a thin covering of enamel, and without cement.

The teeth of Fishes vary much in structure. They are either simple or compound. The simple teeth are those which consist of bone invested by enamel. They may be distinguished into two species, according to their mode of attachment. The first are the simple teeth, which are not implanted in alveoli, but merely attached to the gum, or fixed by articulation to the jaw, as those of the shark tribe; the second are simple teeth growing in alveolar cavities, as is observed in the majority of fishes, the pike, dory, &c. The simple teeth attached to the gum are chiefly distinguished by their fibres intersecting in the manner of the cancellated tissue of bones, and being therefore at first light, porous, and spongy, and becoming afterwards uniformly hard and compact like ivory.

The compound teeth, which consist of an infinite number of minute tubes mutually aggregated and invested by a common covering of enamel, form plates of different sizes, adhering to the bones of the jaws or palate by an intermediate membrane only. In some they affect the disposition of the quincunx; in others they occupy the whole breadth of the upper jaw at least, as in the ray as seen on the small scale, and in the same manner in larger fishes; others are in straight transverse bars; others assume the shape of a circular segment, or the figure in heraldry denominated the chevron.

In the wolf-fish the jaws are provided with eminences composed of fibres or tubes proceeding from the base to the circumference, and which are connected to the jaw by a substance more spongy than the rest of the bone.

After their first formation the teeth retain nearly their mechanical original shape in the ZOO-PHAGA, man, and the QUADRUMANA. In the two latter only their crowns begin to be worn, rendering the incisor and canine less pointed by the use of food partly vegetable; but in the zoophagous tribes they undergo no detrition whatever. In the herbivorous animals, however, the crown begins to undergo detrition more or less rapidly; and in no long time the superior layer of enamel is entirely worn off, and the surface of the tooth exposes the succession of bone or ivory, enamel, and cement. These substances are well seen in the teeth of the RODENTIA, for instance the hare; those of the PACHYDERMATA, as the elephant; the Ruminants, as the stag, sheep, and ox; and the SOLIDUNGULA, as the horse. In all these animals the enamel, which is hardest, forms prominent lines or ridges; while the bone and cement are indicated by depressions.

This detrition, which is purely mechanical, might proceed to such an extent in the herbivorous quadrupeds as to destroy the whole of the crown of the tooth, and leave the process of mastication to be performed by the jaws only. It appears to be chiefly to obviate this inconvenience that the dentition of the elephant, the Ethiopian boar, and perhaps all the PACHYDERMATA, is conducted in a successive manner through a series of six or eight sets of teeth at least. In the former animal, in which this process has been best observed, and was ably explained many years ago by Mr John Corse Scott (Phil. Trans. 1799), each half-jaw, whatever it contains, exhibits at one time only one complete grinder and part of another behind it, the prominent parts of which are placed obliquely to the horizon, forming an inclined plane, so that the anterior parts are worn before the posterior. The anterior complete one, which is employed in mastication, undergoes progressive detrition till its anterior portion is worn down to the level of the jaw. In this state the fangs of the anterior part of the tooth begin to diminish, rendering the tooth narrow before; while the crown of the poste-

Comparative Anatomy. Yrior begins to be worn, and undergoing the same detrition, the posterior fangs also begin to give way. While this process, however, is advancing, the posterior tooth, of which only the anterior part was appearing above the gum, gradually rises, with its crown forming a plane inclined from before backwards, similar to that of the anterior grinder. When this posterior tooth has been raised sufficiently to allow its anterior margin to be used in mastication, the residue of the anterior tooth drops out altogether, and the posterior one continues to rise and advance rapidly, until it is completed, when it is found to be much larger than the previous tooth, and to consist of a greater number of plates of ivory and enamel (denticuli). In no long time this new tooth, which undergoes the same process of detrition, is succeeded by another one, the anterior margin of which rises first behind the posterior one of its predecessor, and which passes through the same stages of growth, detrition, and shedding. This process is repeated at least seven or eight times, and each succeeding tooth is larger, and contains a greater number of ivory and enamel plates than its predecessor. The elephant has thus 7 or 8 grinders in each half-jaw, or 28 or 32 grinders respectively; yet there are never more than one tooth and part of another, or at most two, that is, eight teeth in the upper and lower jaws, at the same time. Though the disappearance of the fangs of the anterior tooth is ascribed to absorption—which indeed is a good general name—yet the true reason is the fact that the maxillary or dental vessels of the elephant are unable to sustain more than one tooth in each half-jaw at once; and that since these vessels gradually transfer their blood to the new tooth, while those of the old one shrink and are obliterated, the new tooth grows as the old one is actually dehematised or atrophied. The order in which the teeth of the elephant succeed each other is nearly the following. The first or milk grinder, composed of 4 eburneo-vitreous plates (denticuli), cuts the gum eight or ten days after birth, is well formed in six weeks, and completely out in three months. The second, which consists of 8 or 9 plates, is completely exposed at the age of two years; the third, consisting of 12 or 13 plates, at six years. The fourth to the eighth grinder consist of plates varying in number from 15 to 23; but the period at which these teeth appear has not yet been determined. This process has been shown to have taken place also in the gigantic fossil animal named the mastodon.

A similar process of displacement and renovation takes place in the poison-teeth of serpents, and in the teeth of the shark, diodon, and tetraodon tribes. In the wolf-fish (anarrhicus lupus) the teeth are shed along with the spongy membrane in which they are contained, exactly as the horns of the stag.

Dentition of the horse. In the horse, in which the process of dentition has been carefully observed, it is usefully employed to determine the age of the animal. The milk incisors appear at the end of 15 days; the four middle ones, or the nippers, are shed at 30 months; the four following ones at 42 months; and the four external, or the corner teeth, at 54 months. The permanent corner teeth do not grow so quickly as the other incisors; and by these especially the age of the horse is determined. At first they scarcely rise above the jaw. Their middle then presents a hollow filled with blackish tartar, the margins of which are worn down as the tooth rises from the gum, and is rubbed against the corresponding one; and it diminishes progressively from 54 months to 8 years, when it is altogether obliterated. The hollow of the other incisors is obliterated at a later period than that of the corner ones; and the age of the animal is then estimated from the length of the incisors, which continue to increase.

The first two molar teeth appear in each jaw and on

each side about the 8th day, the next at the 20th, and the complementary or small anterior grinder about the 5th or 6th month. The first posterior molar appears about the 11th month, and the second in the 20th. At the 30th or 32d month the first two milk grinders are shed, the third in the 3d year; and about the 5th or 6th year the last posterior grinder appears. The milk grinders are longer from before backwards than the permanent ones, which are themselves contracted in this direction, as they are pressed by the posterior grinders; from which it results that the dental crowns of young horses are oblong, while those of the old are quadrangular.

In the MAMMALIA the teeth are always implanted in the jaw-bones, and never, as in other animals, in the tongue, palate, &c. The only exception to this rule is the echidna.

The three kinds of teeth, incisor, tearing, and grinder, are found together only in Man, the QUADRUMANA, the ZOOPIAGA, the PACHYDERMATA except the elephant and two-horned rhinoceros, in the hornless Ruminants, and in the SOLIDUNGULA; but, of all these animals, in man only are the three forms of teeth arranged in an uninterrupted series, and in such a manner that those of the lower jaw are applied to those of the upper. In one other animal only, now extinct, the anoplotherium, is this continuity of arrangement observed.

In the QUADRUMANA and ZOOPIAGA, and all those in which the canine are larger than the other teeth, there is a gap on each side of the jaw to receive the canine of the opposite one. In the ursine genus there is a large empty space behind each canine tooth. In the hedgehog, shrew, phalanger, and tarsier, in which the canine are shorter than the other teeth, a space is left between them and those opposite. In the maki tribe, proper bat, colugo, and camel, there is a large interval between the upper incisors. Lastly, the Ruminants want the incisors of the upper jaw, and the molar teeth of the lower.

Some animals provided with the three classes of teeth lose the incisors at a certain age; for instance several of the bat tribe, and the Ethiopian hog. Other Mammalia have only two sorts of teeth, for instance incisors and grinders, separated by an interval without canine, as the wombat and all the RODENTIA, in which there are only 2 incisors in each jaw; the kangaroo, which has two below and 6 or 8 above; and the cavy or hyrax genus, which have 2 above and 4 below. The elephant has grinders and two tusks planted in the superior intermaxillary bone, but no inferior incisors or canine teeth.

Animals may possess grinders and canine teeth without incisors, as the sloth tribe and the dugong. The grinders, which are most essential, are most rarely wanting; and when others are deficient these are present, as in the armadillo tribe, the orycteropus, the ornithorhynchus, two-horned rhinoceros, and lamantin. The jaws of the dolphin are provided with uniform conical teeth all round, while the cachalot or white whale has them in the lower jaw only. In the narwal there are only two long spiral tusks implanted in the intermaxillary bone, and of these one is often wanting.

Lastly, teeth are entirely wanting in the ant-eater tribe, pangolin, and echidna, which are therefore arranged among the EDENTATA. In the whale the teeth are represented by plates of the laminated, fibrous, bluish substance distinguished by the name of whale-bone.

§ 2. ORGANS OF INSALIVATION.

Under this head ought to be noticed the modification which the salivary glands undergo in the lower animals. Our limited space, however, obliges us to proceed immediately to notice the peculiarities of the other divisions of the alimentary canal.

The muscular tissue of the oesophagus consists, in most of the Mammalia, of spiral fibres twisted in two opposite directions, the external from before backwards, the internal from behind forwards. This arrangement, which was first observed in the RUMINANTIA, was supposed to explain the process of rumination. This opinion, however, is refuted by the fact that the arrangement is not confined to this order, but is very general among the zoophagous and other animals which do not ruminate. In the kangaroo the direction of these fibres is, as in man, transverse in the internal layer, and longitudinal in the external.

The oesophageal mucous membrane is covered by epidermis, which extends to the cardiac opening of the stomach in man, the QUADRUMANA, and all the ZOOPHAGA. This membrane, as well as the mucous, is thrown, by the action of the muscular tunic, into longitudinal folds, which are effaced only when the oesophagus is distended. In the tiger, lion, and lynx, there are large transverse valvular folds, and smaller ones in the civet and cougar—an arrangement connected probably with the carnivorous habit.

The stomachs of the lower animals vary considerably in shape, in the insertion of the oesophagus, in the disposition of their muscular tunics, and in the simplicity or complication of their cavities. These characters it is impossible in such a sketch as the present to consider in detail; and we shall confine our attention to those peculiarities which are most striking in the digestive organs of the animal world.

The stomachs of the MAMMALIA may be distinguished into the simple and compound. Those of man, the QUADRUMANA, zoophagous and most of the herbivorous tribes, belong to the former order. This simple form of stomach, however, may be generally distinguished into two parts, a cardiac and a pyloric, more or less separated from its other by a central transverse contraction of its annular muscular fibres. This is particularly seen in the horse, man, murine family, and many other animals which occasionally feed both on animal and vegetable matter. In the human stomach this contraction is represented in Plate XXXVI. fig. 4. In the porcupine, however, there are three pouches. This contraction depends on a strong annular band of muscular fibres at this part of the organ. In the pure carnivorous animals, however, as the feline family, the annular fibres, which are very thick, are nearly equally so from the cardiac to the pyloric end.

The compound stomachs, or those which contain more cavities than one, are found in the sloths, and the ruminant and cetaceous animals chiefly.

In the first tribe the stomach of the Unau, or two-toed sloth, is two-fold. The first cavity is large and globular, but tapering behind into a conical appendage, separated by a semilunar fold; while a large cul de sac on the left of the cardiac opens into a canal which proceeds at first backwards, and then turning to the right, enters the second cavity by a narrow aperture. The second, which is small, tubular, and folded under the former from left to right, is distinguished by a semilunar fold into two halves, the first of which opens into a small cul de sac on the right side of the first cavity. The inner membrane of both cavities is smooth, and without villi. A similar arrangement is found in the At, or three-toed sloth, with this exception, that the appendage of the second gastric cavity is divided into three compartments by two longitudinal bands. This canal seems analogous to the arrangement of the ruminating stomachs, in so far as it may allow the alimentary matters to pass occasionally from the oesophagus directly into the second stomach.

The stomach of the hyrax, ashkoko, or Cape cavy, also

consists of two pouches, separated by a middle partition, in which there is an aperture for mutual communication. In the hippopotamus the cardiac communicates with three pouches, two of which only are cognizable without, and with a long tubular bowel, the interior of which is divided across by several valvular folds.

In the kangaroo the stomach receives the oesophagus near its left extremity, which is small and bifid (Plate XXXV. fig. 8); and forming a larger cavity on the right, passes upwards, making a turn, and crosses to the left before the oesophagus, makes another turn, and again crosses the mesial plane to the right, where it terminates in a tapering cavity at the pylorus. In this course it presents internally a longitudinal band (L, L, L), extending all round to near the pyloric end, and crossed by valvular membranous folds, which divide the cavity into cells not unlike those of the colon, especially in the horse. The mucous epidermis is continued from the oesophagus over the space marked c, c.

The stomach of the Ruminants consists of four distinct communicating cavities. The first, denominated the Paunch (ποῦσα, rumen, penula, ingluvies; la panse, l'herbier, la double), is a large bag occupying the left side of the abdomen chiefly, marked externally by two sacular appendages, and separated within into four parts. (Plate XXXV. fig. 1, A, A, A, A.) Its inner surface, upon which the epidermis is continued, is occupied by flat papillæ. By a pretty wide aperture (B, B), with rounded margins, this communicates with the second cavity named the Kingshood (χημωφαλας, reticulum, le bonnet), which is distinguished by the rhomboidal and polygonal cells, into which its inner membrane is moulded. An aperture at the further end of this (C) leads into the third or smallest cavity, termed the Maniples (χρυς, omasum, le feuillet), from the numerous concentric crescentic folds formed by its inner membrane (fig. 2, c and p). These folds amount to about 40 in the sheep and 100 in the ox. The smallest of them, between the aperture from the second into the third cavity, are puckered or collected towards their further end by a transverse membranous fold, which acts as a valve to the aperture between this and the fourth cavity. This one, generally named the Red (πυργος, abomasum, la caillette), is of an elongated pyriform shape, slightly incurvated on itself (fig. 2, d), and is marked internally by longitudinal folds (p) incurvated according to the curvature of the cavity itself, and terminating near the pyloric end in rugæ or irregular duplicatures (r). This is the structure of the gastric cavities in the ox.

In the camel, dromedary, and lama, the stomach is equally complicated, though the structure is a little different. In the first of these animals, which may be taken as an example of the others, the Paunch or first cavity is a large bag, divided into two compartments on the posterior part, by a strong band passing from the right side of the cardiac orifice longitudinally downwards (L, L, fig. 3), and forming one border of a groove leading to the orifice of the second. From the left margin of this band proceed eight muscular bands, nearly at right angles, and intersecting with others, form cellular cavities on the left side of the paunch (s); while on the right side there are similar cells, though smaller, and wholly unconnected with the longitudinal band. From the left margin of the cardiac, in like manner, proceeds a broad muscular band (M, fig. 3) to the aperture (B) of the Paunch into the Kingshood, after which it takes another direction (M, fig. 4) towards the Maniples or third cavity, within the orifice of which it terminates (c, fig. 4). The Red or fourth chamber is much the same as that of the bullock (D, P, R, fig. 5), and the only peculiarity is, that after terminating in (Π) the pylorus, it opens in a small cavity (o) which leads into the duodenum (Δ).

From this description, for the particulars of which we are indebted to the accurate account of Sir E. Home, it results that the stomach of the camel differs from that of the ox and other horned ruminants chiefly in the possession of the quadrilateral cells in the second stomach. Into these the water is conveyed by the animal when drinking, and in these it remains. By the action of the muscular band (m), the aperture between the Paunch and Kingshood is opened, and the water is directed into it so as to fill its cells. When these are filled the surplus runs off into those of the first stomach, where at least those on the left side of the long band may be regarded as part of the general cellular structure. These cells are represented of a large size in fig. 5, in which they appear like oblongs with rounded corners. They are always larger on the left side of the band, in the Paunch, than those in the second stomach.

Both in the bullock and in the camel, and in all the Ruminants, the first and second stomachs only are covered by mucous epidermis.

In the Cetaceous animals similar complication of the gastric cavities is observed. In the bottle-nose porpoise, which may be taken as an example, the œsophagus, which is large and capacious, terminates in a spheroidal or ovoidal flask-like bag (fig. 6, A, A) with an aperture a little below the cardiac, consisting of rose-like annular folds, and leading into a second cavity. This, which corresponds to the ruminant kingshood, is nearly spherical in shape (fig. 7, B), and presents valvular folds more circular than in the ruminants, and intersected by others so as to give it the honeycomb appearance characteristic of this cavity. From this another aperture leads into the smallest cavity of the three (C, fig. 7); and thence into the fourth, which is long, cylindrical, and slightly incurvated. The third cavity is remarkable for presenting in its inner membrane numerous apertures of mucous glands.

Birds are distinguished by possessing a stomach consisting of three cavities. The first is the crop, which may be regarded as a mere expansion of the œsophagus, and confined chiefly to land birds. It is filled not only with food, but with small stones; and its chief purpose seems, by mechanical comminution, to supply the place of the teeth in dividing the granular aliment, and bruising or killing the animals swallowed. It is found chiefly in the granivorous birds. It is wanting in the ostrich, in the piscivorous birds, and most of the GRALLA.

The second is the glandular crop or subsidiary stomach (ventriculus succenturiatus, bulbus glandulosus), a membrano-glandular sac, which may be also regarded as an œsophageal dilatation. It is larger when the crop is wanting; and though, when conjoined with it, it is always very glandular, and may be therefore regarded as a chemical solvent of alimentary substances, it appears to supply the want of the crop, which is certainly chiefly a mechanical apparatus. The glandular crop or subsidiary stomach is remarkable for the number and size of the glandular bodies contained between its mucous and muscular tunics. These glands, though variable in shape, are generally conical; and some consist of several glands conjoined in one common peduncle (Plate XXXVI. fig. 3). All of them are hollow, and secrete a fluid which is discharged by one or more minute apertures, and which is of essential importance in the solution of the food. In some instances, as in the American ostrich (fig. 2), they are few in number, and occupy only a small part of the posterior wall of the œsophagus.

The gizzard or proper stomach of birds may be considered as a horny mucous membrane, somewhat cartilaginous, continuous with that of the œsophagus, and covered by two strong thick muscles, the fibres of which converge to a point. (Plate XXXVI. fig. 1.) In the carnivorous and piscivorous genera of birds, especially those in which the

crop is wanting, the gizzard loses its muscular character, and is converted into a membranous pouch.

The stomach of REPTILES does not present those dilatations observed in BIRDS; and when it changes its diameter or capacity, it is only progressively and insensibly. Its general diameter, nevertheless, is proportionally larger than in the two classes already noticed. Most generally without cul de sac, its shape is spheroidal, more or less oblong; its membranous walls are thin and transparent; its muscular layer almost imperceptible; the cellular identified with the mucous tissue; the situation of the cardia indeterminate; and the pylorus, without valve, is distinguished by a simple tapering contraction of the gastric walls, and the appearance of the structure proper to the intestines.

In this class of animals, further, digestion appears to be less regulated by fixed principles than in the other two. It is evidently not confined exclusively to the stomach. The œsophagus of the turtle is provided with numerous large, firm, pointed processes, which in all probability contribute to the mechanical division of the food, so much the more requisite as the CHELONIAD REPTILES are toothless. Except in the crocodile, the SAURIAN Reptiles are destitute of large arch or proper cardiac cavity. In the OPHIDIAL or Serpentine Reptiles the stomach has the figure merely of a dilated sac between the œsophagus and intestines, and presents no curvature. It is probably in connection with this modification of structure that we find animals remain for days in the œsophagus of serpents; and this tube appears to be to a certain extent capable of digesting aliment as much as the stomach. The best mark of distinction in such circumstances is the cessation of epidermis.

In no class of animals does the stomach vary more in shape, structure, and situation, than in FISHES; and perhaps the general character of the alimentary canal in this class is most justly given by representing it as deviating from those attributes of regularity which we find in the higher classes. While in the MAMMALIA and BIRDS it is always distinguished by its spheroidal or pyriform enlargement between two tapering extremities, and by being much more dilated than any other part of the alimentary canal, in the Reptile class it begins to part with this character; and it loses it altogether in the Fishes. In most of the finny tribes it is often not more capacious than the œsophagus; and it is distinguished from this tube only by the villous character of its internal membrane. In general, also, the situation of this pyriform dilated sac is transversely across the body in the Mammiferous class. In the feathered tribes this character also is slightly set aside, and partly from the alteration in shape, partly from that of position, the stomach occupies less of the transverse diameter than of the longitudinal extent of the body. Among the Reptiles this character, though still retained in the CHELONIAD, and even in the SAURIAN, is gradually enfeebled in the OPHIDIAL; and in the FISHES it may be said to be entirely obliterated, since the organ occupies much more longitudinal extent than transverse width of the body.

The first character of the alimentary tube of fishes is the width or capacity combined with shortness of the œsophagus. The latter character is manifestly associated with the absence of lungs and consequent want of chest; so that between the throat and abdominal cavity, the interval, which corresponds only to the space occupied by the heart, is extremely abridged. The œsophagus consists, as in the other classes, of a mucous membrane surrounded by a muscular tunic; but the mucous membrane is distinguished by the firmness and whiteness of its corion, which in some genera approaches to the consistence of horn or cartilage, and by the presence of conical papillæ, sometimes of great hardness, and which appear to

Comparative Anatomy. act mechanically on the food. It is almost superfluous to notice the facility which the large capacity of this tube affords fishes for swallowing their prey. Most of them are voracious in the extreme; and it is not uncommon to find the stomach and œsophagus crammed to the throat.

The figure and position of the stomach of Fishes are so variable, that it is difficult to give a character of general application. Though in many genera, especially of the JUGULARES, it consists of a cylindrical sac with a slight dilatation immediately below the cardia, in others it is oblong ovoidal, as the ray and shark tribe; and in others, as the sole, dory, and flat fishes generally, it is orbicular. In the sturgeon it consists of a cylindrical tube incurvated twice on itself. In none is there more than one cul de sac, the depth of which varies as the part corresponding to the pylorus is more or less remote from the fundus. When the limits of the stomach are indistinct, the situation of the cardia is equally so. In the lamprey (petromyzon) and pen-fish (synnathus pelagicus) the whole tube is of a uniform size from the mouth to the anus; and much the same may be said of the carp genus. The genera in which it forms a distinct dilatation or cul de sac are chiefly the following; the eel (muraena anguilla), conger (m. conger), the bullhead genus (cottus), the scorpion horrida, lobster genus, perch, cuckoo gurnard, mackerel, herring, salmon, mormyrus genus, mullet, and stiltus Bagre. In the anablers the dilatation disappears; and in the chætodon ciliaris and some others it is a large sac incurvated in an arch-like bend.

The intestinal canal in most of the Mammalia is very similar to that of man; and the chief differences of different orders and genera are found in the difference of longitudinal extent either of the whole intestinal tube or the comparative lengths of its several parts.

Comparative length of the intestine to the body. From the time of Grew to that of Cuvier, and most modern anatomists, it has been a point of some importance to determine the length of the intestinal tube in relation to that of the body. This comparative length, which is greatest in the MAMMALIA, diminishes successively in the BIRDS, REPTILES, and FISHES. It has been occasionally stated by different anatomists, that the intestinal tube is longer, ceteris paribus, in granivorous than in carnivorous animals, and conversely. When we come, however, to compare the different lengths of this tube in the several tribes, we find that this statement demands modification. In the ape family its length varies from 5 to 8 times that of the body; in several of the lemur tribe from 4 to 6 times, the smaller length being in this case compensated by the size of the cæcum; and in others of the lemur tribe, e. g. the lori, the intestine is only three times the length of the body. Among the CHIROPTERA there are two examples of very great contrast in this respect. While the intestine of the noctula or great bat (vespertilio noctula, Lin.) is the shortest of all the MAMMALIA, and scarcely does more than exceed the length of the animal's body, that of the roussette (pteropus) or East India bat, which lives chiefly on vegetable matter, is at least 7 times longer than its body. A similar instance of the meeting of extremes is found in the Marsupial order, in which the marmoset and cayopollin have intestines only 2½ times longer than the body, while that of the phalanger is more than 11 times longer. In the plantigrade or ursine family, which occasionally live on vegetable matters, the proportional length approaches to that of the ape.

In the carnivorous animals the intestine, though generally short, varies from 3 times to 8 times the length of the body. The former is the proportion in the lion, wild cat, ocelot, cougar, and weasel; and the latter in the hyena. Some of them also vary among themselves. Thus the intestine of the wild cat is greatly less than that of the domestic animal. The proportion is very great among

the RODENTIA, in several of which it is 8, 10, 12, or 16 times longer, as in the agouti, than the body of the animal. Among the murine tribe, however, it undergoes a diminution. Among the EDENTATA, again, especially the sloth tribe, it diminishes very much, being only about 3½ times longer than the body in the Al and Unau. This brevity is so much the more extraordinary that these animals are void of cæcum, and live on vegetable matters; but, in all probability, the duplicature of the gastric cavities in some degree compensates this deficiency.

The intestinal tube attains its greatest proportional length in the Ruminant animals, being at least 11 times longer than the body, as in the goat, and 28 times longer in the ram. This immense length is supposed to compensate the absence of dilatation in the large intestines, and the small size of the cæcum. In the SOLIDUNGULA, again, which are remarkable for the large size of the cæcum, the length of the intestine diminishes much, being 8 times longer than the body in the zebra, 9 times in the ass, and 10 times in the horse.

Lastly, it is a singular circumstance, that in different species of the same genus the comparative length varies much. We have already noticed the difference between the length of the intestine in the wild and in the domestic cat. The wild and tame boar is a similar instance of the same, the intestine being only 9 times longer than the body in the former, and so much as 13½ times in the latter. It is possible that such differences may depend on the different habits of the animal in his wild and domestic condition. This explanation, however, is totally inadequate to account for the difference in the comparative length of the intestinal tube in the Asiatic and African elephant, being 10 times longer than the body in the former species, and only 7 times in the latter. The same may be observed of two animals very closely allied, if not of the same genus, the echidna and ornithorhynchus. The intestine of the former animal is 7 times longer than his body, while that of the latter is only 5 times.

In BIRDS generally the intestine is shortest among those genera which prey on animals and fish; it varies from twice to five times the length of the body. In the gallinaceous and passerine birds, which live on grains, it is always longer and more capacious than in those which live on animal substances. In the cassowary, which is granivorous with a short intestine, the intestine is divided by contractions into distinct pouches, to compensate for the brevity of the tube. It must be remarked, nevertheless, that in several birds which prey on fishes, the intestinal tube is proportionally longer than in those which feed only on grain; and the proportional length is not diminished in birds living indiscriminately on animal and vegetable aliment.

The intestinal tube of the Reptiles is still shorter than that of Birds, and often it does not exceed twice the length of the body. It is longest in the CHELONIAN, and shortest in the RANINE and SERPENT tribe. In the tadpole, however, a singular peculiarity is observed. The intestinal tube of the tadpole of a frog is nearly ten times longer than the space between the muzzle and the vent (anus); whereas, after the animal has become a frog, the intestine is only twice as long as this space.

In the class of Fishes the intestinal tube is still shorter and more direct in its course to the vent or outlet. In a few which live chiefly on marine vegetables, for instance some of the chætodon genus, it is about six times longer than the body of the animal; and in a few of the carp genus (viz. cyprinus capax) it amounts to ten or twelve times longer than the body. In others, however, of the same genus, for instance the cyprinus mursa, it is scarcely as long again, showing here once more an instance of the conjunction of extremes not easily explained.

In most vertebrated animals the intestinal tube may be distinguished by natural marks into two divisions, one extending between the pyloric end of the stomach to a part of the tube, where it becomes wider and more capacious; the other from this to the vent or outlet. In some instances, however, this distinction is very obscurely and imperfectly marked. In the MAMMALIA, in which the distinction is observed, it is indicated by one or more appendages, which, if large, are denominated cæca, and if slender and long, are termed vermiform processes. Man, the ourang, and the wombat (phascolomys), are the only animals which are possessed at once of cæcum and vermiform appendage. In the other genera of the ape tribe, in the maki of the lemur, in the colugo among the CHIROPTERA, the ichneumon, many of the carnivorous tribe, the opossum and kangaroo, the RODENTIA except the dormouse, the Cape ant-eater, the PACHYDERMATA except the hyrax, the Ruminants, SOLIDUNGULA, and AMPHIBIA, there is only a cæcum without vermiform process. The cæcum is wanting in the sloths, the bat tribe, the Ursine except the ichneumon, the marten, pine-marten, weasel, &c. the dormouse, and all the Cetaceous animals.

The presence of cæcum or vermiform process, however, is not necessary to distinguish the canal into two portions. The inner or mucous surface of the ileum is always villous and uniform; and the whole intestine, except its superior or pyloric portion, is convoluted in proportion to its length, and moves about freely in the abdominal cavity; whereas the colon is more or less fixed at different points, it is shorter and more capacious, and its inner membrane is merely mucous without long villi. A mark equally general is the semilunar duplicature of mucous membrane placed between the ileum and colon, and named the ileo-colic valve. In the sloth and armadillo tribes, which want cæcum and vermiform process, this and the slight difference of diameter are the only marks of distinction between the ileum and colon. In all the other MAMMALIA which are destitute of cæcum the whole tube is of the same calibre, occasionally diminished towards the vent; and the division into ileum and colon is no longer cognizable.

In all the MAMMALIA with one cæcum, it appears in the form of a production from the large intestine beyond the part at which it receives the ileum; and though variable in its diameter and structure, it bears a general resemblance to the colon in these respects. In herbivorous animals, and even in some that are omnivorous, as the ape and lemur tribes, it is generally large and puckered by tendinous bands. To this, however, an exception is observed in the Ruminants, in which the cæcum is moderate in size and unpuckered. It is small and unpuckered in the kangaroo-rat and wombat; while, conversely, in the colugo and brown phalanger, which are believed to be chiefly zoophagous, it is very large and puckered.

In the zoophagous animals generally both the colon and cæcum are of small calibre, little different from that of the ileum; and both the colon and cæcum are destitute of cells or compartments. In herbivorous and several omnivorous animals, on the contrary, the inner surface of the colon is divided by longitudinal and transverse bands into a number of cells or compartments. From this rule, however, there is an exception in the wombat, kangaroo, and the Ruminants. In the RODENTIA the colon is cellular at its commencement only.

In birds the canal is provided with two cæca, one on each side, not far from the vent. In the omnivorous and granivorous these cæca are generally long and capacious. While they are very large in the nocturnal predatory birds, they are either obliterated or wanting in the diurnal predatory birds, in the green woodpecker, the lark, and the cormorant. In the heron, bustard, and grebe, there is a single small one; in the cassowary two very

slender; and in the merganser, diver, &c. they are short and thick. Comparative Anatomy.

In all birds the short bowel between the insertion of the cæca and the cloaca is a little wider and more capacious than those between the pylorus and the cæca; and this is the only circumstance which indicates in this class the distinction of the tube into ileum and colon.

In the reptiles the intestinal tube is generally void of cæca or appendage; and the only distinction consists in the one part of the tube, which is long and slender, being joined to another which is short and thick, and in the presence of a semilunar membranous fold at the point of insertion. In the iguana alone has a genuine cæcum been observed.

The distinction into small and large intestine, or ileum and colon, is still less obviously observed in the class of fishes. It sometimes happens that the difference of capacity is inverted, and that the calibre of the portion which terminates at the vent is actually smaller than that of the part connected with the stomach. This arrangement is observed in the ray, shark, sturgeon, and even the bichir; in the syngnathus, trunk-fish, and balista. In other instances the diameter is the same throughout; and the only distinction is derived from the anatomical characters of the inner membrane. In the lamprey, sea-devil, rough star-gazer, radiated sole, holocentrus sago, carp tribe, mormyrus, and mullet, it is impossible to distinguish the intestine into large and small.

Fishes resemble REPTILES in being destitute of cæcum Pyloric at the junction of the small and large bowel. In many of them, however, there is attached to the intestine, somewhere below the pylorus, a variable number of small intestines terminating in blind ends, similar in size and structure to the intestine with which they communicate. These tubes, which have been not very happily named pyloric appendages (appendices pylorice), in so far as they are most frequently connected rather with the part of the bowel corresponding to the duodenum, vary in number from 2, 4, 6, or 8, to 80 or 180 in some genera, and even their number is not the same in different species of the same genus. Thus, while there are 6 in the smelt (salmo eperlanus), there are 68 in the s. lacustris, and 70 in the salmon (s. salar). In like manner, though there are 18 in the anchovy (clupea encrasicolus), there are 24 in the herring (c. harengus), and fourscore in the shad (c. alosa). In some, as the cod and pollach, they consist of several large trunks ramified into numerous small ones.

These appendages, however, are wanting in the cartilaginous fishes with free branchiæ, in most of those with fixed branchiæ, in the APODES, and in several of the thoracic and abdominal order. In the sturgeon and some others they are represented by a series of communicating cavities inclosed in the intestinal membrane, which is covered by a cellular-muscular tunic and peritoneum (Plate XXXVI. fig. 5), and which assumes the external appearance of a pancreas. Peculiar form of the appendages.

Among the cartilaginous fishes the brevity and directness of the intestinal tube is compensated by a peculiar disposition of the intestinal mucous membrane. This consists in part of the membrane projecting like a broad fold or process from the inner surface of the intestine, and winding round from the pyloric to the anal or lower extremity (fig. 7 and 8). This, which is denominated the spiral valve, may be easily understood from these figures, which represent the arrangement as it is observed in the shark. In the sturgeon, in which it is found in the last portion of intestine (fig. 6), its peculiarities have been described by the writer of this article in the Wernerian Transactions, vol. vi. The spiral valve of fishes.

In all the MAMMALIA the intestinal tube terminates in a distinct bowel denominated the rectum, the mucous membrane of which is continuous with the skin at the

anus. This rule can scarcely be said to be violated in the case of the echidna and ornithorhynchus, in which there is an aperture at the lower part for the urine and the semen of the male and the ova of the female. These anomalous and singular animals form a transition to the mode in which the intestinal tube terminates in the three genuine oviparous classes. In BIRDS, REPTILES, and most of the cartilaginous FISHES, the intestinal tube terminates in an outlet common to it with the urinary organs, denominated generally the cloaca. In the sturgeon, however, it has been shown, in the paper already mentioned, that there is a distinct urinary outlet; and that consequently this animal cannot be said to have a cloaca. In the greater part of fishes, while there is one vent for the excrement, there is another common one for the urine, the ova, and the spawn.

On the subject of the liver, spleen, and pancreas, it is impossible to enter with any interest in this sketch.

CHAP. II. SECT. I.—THE HEMATROPHIC ORGANS.

The Mammalia and Birds agree in having a heart consisting of two pairs of chambers, a venous auricle and ventricle, and an arterial auricle and ventricle. The Eustachian valve is often wanting in the Mammalia. It is wanting, for instance, in the lion, bear, and porcupine; while it is broad and muscular in the seal, and assumes a spiral direction along the upper walls of the right auricular sinus in the elephant.

It was at one time imagined that the aquatic MAMMALIA were distinguished from the terrestrial by the foramen ovale being open and forming a communication between the two auricles. This, however, is a mistake, at least in the adult animal; for neither in the otter, the seal, nor the porpoise or dolphin, did Cuvier find this aperture pervious; and it may be inferred, that when it is open, it is an abnormal remain of the fetal structure. In the ornithorhynchus, also, according to Sir Everard Home, it is impervious. In the porcupine and elephant, in which there are two anterior venæ cavae, the blood of the left anterior cava, which opens in the sinus near the auriculo-ventricular aperture, is conveyed directly into the right ventricle.

The reptile heart, the first in the cold-blooded division, varies somewhat in the several tribes. In the first three, the CHELONIAD, SAURIAL, and OPHIDIAL, it consists of two auricles and one ventricle, divided in some instances into communicating chambers. In the Batrachoid family, on the contrary, it always consists of one auricle and one ventricle, the interior of which is unilocular or undivided. In several of the Turtle tribe, among the first family, the ventricle consists of a pulmonary chamber, in which the blood is more particularly directed to the pulmonary artery, and a general or aortic chamber, which is above, and from which the blood is conveyed into the aorta. In the crocodile the ventricle is divided into three chambers, communicating by several apertures. One is inferior and to the right, and communicates with the right auricle by a large aperture provided with two valves. On the left, and before, is the second chamber, receiving the orifice of the left descending aorta. Behind is an aperture leading into the smallest chamber of the three, situate at the middle of the base of the heart, and receiving the common trunk of the pulmonary arteries. The left chamber is above. In the OPHIDIAL or SERPENTINE family the ventricle is divided into two chambers, a superior and inferior, separated by an imperfect partition, which allows the two to communicate.

The heart of the finny tribes is as simple as that of the Ranine reptiles, which indeed constitutes the preparatory step in the descending scale of organic forms. It consists, as in these animals, of two chambers only, an auricle and ventricle. The former receives the blood from the body

at large, and transmits it to the ventricle, which is almost in all cases unilocular. From this a single vessel conveys it, not to the body at large, but to the gills, from which it is again collected by several branchial veins.

Of the blood-vessels of the four vertebrated classes it is superfluous to speak in any detail.

The most remarkable circumstances are the minute subdivisions which in some classes the arteries undergo previous to final distribution. In the Ruminants, and several of the PACHYDERMATA, the branches of the carotid artery, instead of uniting by the communicating vessels, are subdivided into a great number of minute vessels, which form round the pituitary gland a communicating plexus, denominated by the ancients Rete mirabile.

In the slow lemur (lemur tardigradus) Sir A. Carlisle found the subclavian artery, after entering the axilla, divide into 23 arterial cylinders of equal size, and the iliac on the brim of the pelvis into at least 20 equal-sized tubes, which in both cases surrounded the principal artery, reduced to a small vessel, and, proceeding along the extremity, were distributed chiefly to the muscles. (Phil. Trans. 1800.) A similar arrangement, carried to a still greater extent, was found in the Al or three-toed sloth, in which the axillary and iliac arteries were divided into about 60 or 65 cylindrical parallel tubes. In the slender lemur (Lemur loris) these vessels are subdivided into 4 or 5 only.

In fishes in which the unilocular ventricle sends its blood to the gills only, the heart is pulmonary, and the arterial system is destitute of central impulsive organ. In the sturgeon, which is one of the best examples of the distribution of the arterial system in the finny tribes, the blood, which is distributed in the branchiae by the large artery, is collected in numerous vessels, which may be regarded as analogous to the pulmonary veins of the warm-blooded classes, but which have thick parietes like arteries; and these uniting, form a large vessel, which is lodged in a cartilaginous canal formed by the continuous bodies of the vertebrae. This vessel is further void of compressive or elastic tunics, and the blood moves through it as through an immovable and inelastic tube. From its sides, however, a series of arterial vessels issue, which forthwith assume the usual characters of arterial tubes. In many other fishes the parietes of the large artery adhere in part to the semi-ossous canal in which it is lodged.

SECT. II.—THE ORGANS OF AERATING CIRCULATION.

The lungs of the MAMMALIA are in all essential points perfectly similar to those of the human subject.

The lungs of Birds differ chiefly in not presenting distinct lobules, in having the air-vessels larger and more distinct, in the branchial tubes not becoming quite so small, and in terminating not alone in the pulmonic vesicles, but in perforated parts of the surface of the lungs, which lead into large air-sacs, communicating with all parts of the body, and forming an accessory lung. In the ostrich, which may be taken as a general example, there are four of these cells or acrolabous sacs. The first, which is anterior, extends from the apex of the chest to the iliac bones, between the first ribs and heart above, and between the lower ribs and a cell which surrounds the intestines. It is divided into four chambers, the first two communicating with the lungs by large apertures, while the fourth opens in the iliac bones. Behind this large sac are two small ones, between the iliac bones and the peritoneal sac. Before it is another small one occupying the lateral regions of the apex of the chest, and communicating with sacs in the axillæ and neck. Besides these, the stomach, liver, heart, and intestines are surrounded by sacs. All of these communicate by saccular processes with the cavities of the bones. By this peculiar arrangement BIRDS possess the greatest extent of respiratory surface of all classes of animals.

The lungs of Reptiles are distinguished by consisting of large sacs, subdivided by membranous partitions into polygonal cells, which again are subdivided by smaller slips into smaller cells. In these cells the bronchial tubes are not ramified, but divide abruptly in them at the surface of the lung. Some idea of this arrangement may be obtained from the lung of the ordinary land-tortoise (Testudo Græca, Plate XXXVII. fig. 1). The sacs or large cells are smaller and more numerous in the turtle (fig. 2); but the general disposition is much the same.

The young of the Batrachoid family, and several of the water-lizard tribe, are provided with fimbriated or ciliated processes attached to the neck, and which are in all respects similar to the gills of fishes. These gills disappear as the animal grows; and when it assumes the true ranine or reptile character, vesicular lungs like those of other reptiles, and which had continued in a hitherto latent and rudimentary state, are developed, and the animal breathes as others of the same tribe.

A peculiar form of respiratory organ is found in the lamprey or seven eyes, and the two species of hag-fish, (Myxine, Lin.; Gastrobranchus, Bl.; and Gastrobranchus Dombyi). The former has on each side seven apertures leading into cylindrical tubes, in which the branchiæ are contained. (Plate XXXVII. fig. 3.) In the two species of hag these tubes are dilated into ovoidal cavities, in which the water is received, and on the membrane of which the branchial vessels are distributed. In this respect, therefore, the hag-fish approaches to the mode of respiration among the cephalopodous MOLLUSCA, in which the branchiæ are inclosed in a cavity. Lastly, in the Aphrodite aculeata, which may be taken as an example of the respiration of worms, there is a series of tubes like tracheæ and bronchi, proceeding from the surface to the interior, and in which the water containing the air requisite for respiration is received. (Fig. 6.)

Under this head ought to be described the urinary organs of the four vertebrated classes. Those of the MAMMALIA agree in consisting of kidneys more or less lobulated, ureters or excretory tubes, a reservoir or urinary bladder, and a urethra opening on the same mucous surface with the organs of generation. In the three oviparous classes considerable changes are made. Though in Birds and Reptiles the glandular organs denominated kidneys are left in the shape of aggregated glands with the two excretory tubes, the bladder is withdrawn, and the ureters open in the cloaca. The only apparent exceptions are the ostrich and cassowary, in which the cloaca is so organized that it may serve as a bladder or temporary receptacle of the urinary secretion. In the Reptiles the presence of this organ is variable, being found in the CHÆLONIAD and BATRACHOID; and the iguana, tupinambis, chameleon, stellio, and dragon, among the SAURIAL tribe; but wanting in the crocodile, lizard, agamæ, gecko, and the whole OPHIDIAL tribe. In Fishes it is not less variable. While the ray and shark tribe are destitute of bladder, and the ureters terminate in a cloaca, this receptacle exists in the sea-devil, lump-fish, globe-fish, and others of the cartilaginous division.

A peculiar secreting organ, deserving notice, is the poison gland of the poisonous serpents. It is a glandular body situate on each side above the upper jaw, behind and below the eyes, with a considerable cavity, which opens into a long excretory tube, lying along the outer surface of the upper jaw, and opening in the tubular tooth, represented at fig. 17 and 18; and which is movable in an articulation, and may be erected, as in fig. 18, or depressed, as in 16, at the will of the animal. The poisonous serpents are therefore distinguished from the innocuous by the presence of

the erectile movable tubular fangs. Fig. 15 is the head of the innocuous, and 16 of the poisonous serpent.

Under this head we mention only the nipple-bag (marsupium mammillare), or secondary uterus of the Marsupial animals (Plate XXXVII. fig. 7); the nipples (fig. 8); and the manner in which the fetal animal, in a very imperfect and embryal form, becomes attached by the mouth to the nipples (fig. 10). The Marsupium, therefore, ought to be regarded, not as a mere pouch in which the young may take refuge after they are grown, but as a subsidiary uterus, combining the character of the Mammæ of the other orders.

The MAMMALIA are peculiar in possessing a uterus. In the other classes this organ is withdrawn, and the ovary (fig. 11) and oviduct alone are left. In the ovo-viviparous animals, as the ovo-viviparous shark, the oviduct (fig. 12) resembles that of the common fowl. In the lower classes the ova are hatched out of the body entirely.

In the space assigned to this article, it was impossible to treat fully of a subject so extensive as the structure of the animal world; and while the author has arranged its divisions in such a manner as to show in what order it may be most easily and advantageously studied, he has introduced only those topics which are most indispensable, and most require illustration. For more complete details, therefore, he refers the reader to the following works.

1. 1. Leçons d'Anatomie Comparée de G. Cuvier, Membre de l'Institut National, &c.; recueillies et publiées sous ses yeux par C. Duméril, chef des Travaux Anatomiques, &c. Cinq tomes. Paris, tome i. 1799;—tome v. 1805.
2. 2. Blumenbach's Manual of Comparative Anatomy; with additional Notes by William Lawrence, Esq. F. R. S. Second edition, revised and augmented by William Coulson. Lond. 1827, 8vo. The notes are derived chiefly from the work of Cuvier and the papers of Sir E. Home in the Philosophical Transactions.
3. 3. Gore's Translation of Cuvier's Introduction to the Comparative Anatomy of Animals. Lond. 1827, 2 vols. 8vo. The arrangement of this work, in which the author examines the forms of organs as they ascend, from the lowest to the highest classes, diminishes its general interest.
4. 4. Lectures on Comparative Anatomy, in which are explained the Preparations in the Hunterian Museum. By Sir Everard Home, Bart. Lond. 1823, 6 vols. 4to. This work consists of the papers read by the author at the Royal Society, and published in their Transactions. Though entitled, therefore, Lectures on Comparative Anatomy, it embraces a much more extensive field, and contains a great number of physiological and pathological papers. This renders it at once rather desultory and prolix. It contains, nevertheless, a great number of facts illustrative of peculiarities of structure in the animal world; and it is particularly valuable by the number of engravings with which it is embellished. It can scarcely be said to possess any arrangement whatever.
5. 5. Recherches sur les Ossemens Fossiles, où l'on rétablit les caractères de plusieurs animaux dont les Révolutions du Globe ont détruit les espèces. Par M. le Bar. G. Cuvier, &c. Nouvelle édition. Tome i. 1822, Osteology of the Elephant, Hippopotamus; tome ii. partie i. 1822, Osteology of the Rhinoceros, Horse, Hog, Daman, and Tapir; tome iv. 1823, Osteology of the Deer and Ox, the Bear, Hyena, Lion, Glutton, Wolf, and Dog; tome v. partie i. 1823, Osteology of the Reptiles, the Ichthyosaurus and the Plesiosaurus; partie ii. 1823, RODENTIA, EDENTATA, MONOTREMA, AMPHIBIA, and CETACEA. Paris, 1824. These papers contain much accurate osteological description.

VEGETABLE ANATOMY.

ALL the plants which collectively form the vegetable kingdom have been arranged under two great divisions. Those which possess visible organs of reproduction, as stamens and pistils, have been termed phenogamous, and constitute the first 23 classes in the sexual system of Linnaeus; while those in which the reproductive organs are either obscure or have not yet been discovered, have been termed cryptogamous, and form the 24th class of that system. Humboldt estimates the total number of known species at 44,000; of which 38,000 belong to the former division, and 6000 to the latter. M. Brogniart, jun. supposes the present known Flora to embrace 50,350 species, of which 10,200 may be considered as cryptogamous, and 40,150 species as belonging to the phenogamous division. On the present occasion we propose to describe the structure of those plants only which are included in the division of phenogamous plants.

Of the vast number of plants which cover the surface of our globe, it belongs to the botanist to describe the external forms in such manner as may serve to discriminate species, and assign to each its place in a methodical system of arrangement: it is the province of the anatomist to demonstrate, by dissection, their internal structure, so as to prepare the way for a rational explanation of their functions. In pursuing this object we may examine successively each of the parts as they severally present themselves to view; or we may, in the first place, study the elementary organs common to all plants, and then consider how their combinations form the different parts of vegetables. The former method is the analytic, and was necessarily that of the first observers; but, now that all the different parts are recognised as being formed out of the same elementary organs, we gain, both in conciseness and clearness, by adopting the latter or synthetic mode.

We shall accordingly describe, first, the elementary organs, and the primary textures which they form; and proceed afterwards to the consideration of the individual members and more complex organs of the plant. These elementary organs have been denominated vessels and cells; and they form, either singly or conjointly, what are called the vascular system and cellular tissue of plants. Their combination gives rise to certain textures, which appear in the well-known forms of skin, of bark, and of wood. On these we shall bestow the appellation of common textures, and exhibit a general view of their structure and disposition in the several parts of the vegetable body. A brief description of some minuter parts, as of hairs and glands, will terminate this division of the subject.

In the second part we shall begin with a description of the general structure of seeds, and afterwards treat more particularly of those bodies under the two great divisions of monocotyledonous and dicotyledonous seeds, tracing also the changes of form and of structure which they exhibit in their evolution and progress to the state of the mature plant. The structure of the mature plant itself will next claim our attention; and we shall accordingly exhibit the anatomy of its several members, as of the trunk, the branch, and the root, in their more remarkable varieties and forms. After this, the structure of the organs that spring from these several members, as buds and bulbs, leaves, flowers, and fruits, will be separately and distinctly examined; and having thus followed the progressive changes of form and of structure exhibited in the several stages of vege-

table existence, we shall terminate our descriptions by anatomical representations of the organs in which the seed was produced, and the series of appearances successively displayed in its formation.

Through the whole of the descriptive detail we shall adhere as closely as we can to the language of demonstration; supporting and illustrating our representations of structure by continual reference to figures, selected in great part from authors of repute, and in some instances from dissections made by ourselves. We are aware, that of many reputed anatomical facts very different representations have been given, all equally professing to rest on microscopical observation. In such circumstances, we can do no more than report concisely the statements of different observers; but shall dwell chiefly on those descriptions and representations which seem best entitled to credit, and appear most conformable to the analogies of other organized structures.

When we consider the immense number of species that compose the vegetable kingdom, and call to mind that in form, in size, and in structure, each species differs from every other through every period of its existence, it must appear altogether impracticable to describe and delineate any considerable number. Fortunately, however, these diversities arise not from differences in the elementary organs, but chiefly from their varied proportion, disposition, and texture. In numerous species the disposition of the internal organs is very similar, where the external form and texture widely differ. In other instances the arrangement and composition of the internal parts vary not less than that of the external figure. Of these varieties we shall exhibit different examples.

In describing individual parts or organs, we might have brought many concurring examples, and exhibited many similar representations, to confirm the views of structure under consideration; but, in general, we have dwelt only on one or two examples, and these we have selected from plants which are either important in themselves, or whose structure has been most satisfactorily displayed, or which seemed to afford the best illustration of the peculiarities we were engaged in describing. From one example clearly given, the reader will readily apprehend the nature of analogous structures, and escape the perplexity and fatigue which unnecessary repetitions might occasion.

Instead, also, of describing the vegetable at one or two stationary points of its existence, in some of which its size is so minute as to be scarcely capable of demonstration, we have followed it through the several stages of its growth. In this way we really study it as a living body, continually exerting its vegetative powers, and daily exhibiting the most striking variations in external form, and frequently in internal structure. We hope thus to have conferred an interest on the descriptive part, which may in some measure relieve the unavoidable dryness of anatomical details; to have exhibited, in some instances, clearer views of vegetable organization; and to have given a continuity to the subject which isolated dissections, at a few stated periods, could not alone have bestowed.

It remains only to add a few remarks on the nomenclature employed in the present article.

In the description of external parts we have adhered chiefly to the Linnaean nomenclature; but some of the terms employed by Linnaeus, in relation to anatomical structure, are exceedingly vague and inappropriate; others

are manifestly erroneous, and, however well suited to the purposes of botany, are not at all to be tolerated in any thing that aspires to correct anatomical description.

In the anatomy of seeds we have adopted many of the terms employed by Gærtner, in his excellent work De Fructibus et Seminibus Plantarum, most of which had previously been used by Malpighi and Grew.

Thus, in every instance, we have exercised our own judgment in the selection of terms, and, where it seemed necessary, have subjoined the synonyms of different writers. Though we presume not to say that we have uniformly chosen the best, we trust they will always clearly

express the idea we designed to convey; and that, in general, they have been used in one and the same sense, and in no other. Except in one or two trivial instances, we have not ventured to introduce new terms, but have studiously sought to avoid it, retaining even an inappropriate expression, sanctioned by use, if it did not, at the same time, lead to ambiguity, or convey an idea evidently false; and we have in general resisted that torrent of new and barbarous terms, founded often on fancied refinements and pretended discoveries, with which several continental writers have of late attempted to deluge this branch of science.

Before we proceed to describe the structure of the individual parts of vegetables, it may be useful to exhibit a general view of the elementary organs of which they seem to be composed. Such a view will prepare the reader for understanding more clearly the descriptive language hereafter to be employed, and will even much abridge the extent to which that description must otherwise be carried.

Every one is familiar with the natural division of plants into herbs and trees, and is aware that, how different soever they may appear in form and texture, they all possess, in common, certain parts or members which we name the root, the trunk, and the branches, from which proceed the leaves, the flowers, the fruits, and seeds. Infinitely varied as these several parts are, in figure, size, and texture, they all originate from a few constituent or elementary organs, whose situation, proportions, and combination, give rise to all the diversity that we see. "Upon the anatomical analysis of all the parts of a plant," says Grew, "I have certainly found, that in all plants there are two, and only two, organical parts essentially distinct, viz. the pithy part and the ligneous part." (Anatomy of Plants, p. 19.) "And as every part hath two, so the whole vegetable, taken together," he adds, "is a composition of two only, and no more. All properly woody parts, strings, and fibres, are one body; all simple barks, piths, parenchymas, and pulps, and, as to their substantial nature, peels and skins, are all likewise but one body; the several parts of a vegetable differing from each other only by the various proportions and mixtures, and varied pores and structure, of these two bodies." (Ibid. p. 47.) In the anatomical descriptions of Malpighi, the compound structure is resolved, in like manner, into two constituent parts, called by him the ligneous and utricular portions. To these parts may be assigned the general appellations of the Vascular System and Cellular Tissue of plants, the description of which shall form the first subjects of consideration.

By the vascular system may be understood, in a general sense, all those parts of a plant which do not exhibit the form either of membrane or of cells. It constitutes almost the entire bulk of the more solid parts of

trees, and by Grew and Malpighi was denominated the ligneous body, in contradistinction to the cellular tissue which accompanies it, and which forms by far the largest portion of many herbaceous plants. To common observation, a piece of dry wood appears to be a mass of solid fibres, that is, a series of particles arranged in a filiform figure, and destitute of any continuous canal. Thus Tournefort and others considered the ligneous parts of a plant to be a mass of minute solid filaments, placed parallel to each other, like the threads in a skein of silk, between the interstices of which the sap ascended; but the anatomical researches of Grew conducted him to a different conclusion. "If it be asked," says he, "what all that part of a plant, whether herb or tree, which is properly called the woody part—what all that is? I suppose that it is nothing else but a cluster of innumerable and most extraordinary small vessels or concave fibres." (Anat. of Plants, p. 20.) Malpighi held similar opinions concerning the vascularity of plants, which was farther attested by the microscopical observations of Hooke and Leeuwenhoek. Du Hamel, though he admits that, under maceration, the parts of plants seem capable of indefinite subdivision, yet, from many circumstances, avows his conviction of their vascularity; and Hedwig maintains that the oldest and most compact plant is but a congeries of vessels and cells, which have nothing of the character of a fibrous solid, except in the thin membranous coats by which they are formed. (De Fibra Vegetab. Ortu, p. 17.)

Few circumstances have contributed more to perplex and retard our knowledge of the structure of plants, than the vague and erroneous nomenclature that has been employed to designate their constituent organs, more particularly in relation to the vascular system. Thus the several terms filaments, fibres, strings, threads, and nerves, which, in their ordinary acceptance, are understood to express a solid substance, have been constantly made use of in describing the tubes or vessels of plants. The same organs, however, even by the same writers, are frequently called tubular bodies, ligneous tubes, concave fibres, ducts, canals, arteries, veins, and vessels. In our future descriptions we shall employ the term vessel in a generic sense, to express all the diversity of names just enumerated; and the different kinds or species of vessels we shall hereafter attempt to discriminate by appropriate appellations.

Vessels, as we have said, exist in almost every part of a plant. In the higher orders of animals the fluids contained in the vessels are conveyed to a central reservoir called the heart, from which they are again sent out to all parts of the body. Near to this reservoir the ves-

Hollowary cells are few in number and large in size; and they gradually lessen in size and increase in number as they recede from it. In plants there is no such reservoir, but the fluids which enter by innumerable mouths at the root are at once distributed equally through all parts of the vegetable that are fitted to receive them. Hence in plants there is little variation in the diameter of the vessels; and their general figure is therefore cylindrical.

This size. From the extreme minuteness of the vessels, it is scarcely possible to compute their number with accuracy. By driving off their fluids without destroying their figure, as is done in the preparation of charcoal, Hooke numbered in a line, $\backslash frac\{1\}\{100\}$th part of an inch long, not fewer than 150 vessels; therefore, in a line an inch long, there must be 2700, and in a surface of a square inch, 7,290,000 vessels; "which would seem incredible," says he, "were not every one left to believe his own eyes." These facts he verified by other observations on decayed wood, in which the vessels were empty; and also on petrifications of ligneous bodies, in which the places of the vessels were very conspicuous. In very close and dense wood, as that of guaiacum, the vessels were still more minute than in the examples just quoted. (Micrographia, p. 101, 108.) In a piece of oak of the size of about $\backslash frac\{1\}\{15\}$th part of a square inch, Leeuwenhoek reckoned 20,000 vessels; so that in an oak-tree of no more than one foot in circumference, or about four inches in diameter, there will be found, according to his computation, 200,000,000 of such vessels. (Select Works, translated by Hoole, vol. i. p. 3.) The largest vessel observed by Hedwig (De Vegetab. Ortis, p. 25) in the stem of the gourd appeared through his microscope about $\backslash frac\{1\}\{100\}$th of an inch in diameter; and as his instrument magnified 290 times, the true diameter must be reckoned the 3480th part of an inch, which would give for the square inch 12,110,400 vessels. In certain plants, however, the vessels are large enough to be discerned by the naked eye, and in some cases they acquire a large size.

Fasciculi. The vessels of plants do not, like those of animals, exist single, but are collected into fasciculi or bundles, which, however, have often the appearance of single vessels. In the stems of herbs, and in roots, Grew discovered each small fasciculus to be composed of from 30 to 100, or sometimes many hundred vessels. (Anatomy of Plants, p. 20.) The direction of these fasciculi in the trunk is generally perpendicular, but in other parts their course is often oblique, and in their smaller ramifications they produce all sorts of figures. In herbs the fasciculi are more or less numerous, and placed often at considerable distances from each other, exhibiting the appearance of small columns dispersed through the cellular tissue: in other instances they are much more numerous, but destitute of any symmetrical arrangement; while, in trees, they are disposed regularly around the axis, presenting in their transverse section the well-known appearance of concentric circles in the wood.

Ramify. In some parts, where the fasciculi stand at a distance from each other, some vessels often quit one parcel to unite with another, and return afterwards to that which they had previously left. In this manner they are said to ramify, and frequently, by their ramification, a reticulated appearance is produced, as occurs especially in the bark and leaves. In the wood of the trunk, where they stand collaterally in a perpendicular direction, they very seldom, if ever, run into one another, but keep, says Grew, like so many several vessels, all along distinct; as, by cutting, and so following any one fasciculus, may be observed. (Anatomy of Plants, p. 20.) In branches and roots, though the direction of the fasciculi be changed,

they seem only to break into smaller parcels, and run side by side, never inosculating with each other, nor being ramified, so as to be successively propagated one from another, as in the vessels of animals. Neither, adds he, are they wound together like threads in a rope, but are only contiguous or simply tangent, like the several cords in the braces of a drum. (Anatomy of Plants, p. 66.) Even in the leaf, where the vessels seem to ramify out of greater into less, as in the arteries of animals, yet, if the skin and pulp of the leaf be removed, and the vessels laid bare, it will appear that they are all of the same size everywhere in the leaf, and all continued through it, as distinct tubes, like the several threads in a skein of silk. The distribution of the vessels is not effected, therefore, by their ramifying out of greater into less, but by the division of a greater fasciculus into several smaller fasciculi, till at last they come to be single. (Anatomy of Plants, p. 155.)

It may be doubted whether, when the vessels of different fasciculi come into contact, they ever actually unite and are lost in each other, forming that kind of connection which is called inosculation or anastomosis. Grew strenuously contends against any such connection of the vessels in any part of the plant. "On a superficial view, indeed, the vessels of the leaf," says he, "seem to be inosculated, not only side to side, but the ends of some into the sides of others: but neither is this ever really done, the lesser fasciculi being only so far diducted as to stand at right angles with the greater; but they are never inosculated, except end to end, or mouth to mouth, after they are come at last to their final distribution." (Anatomy of Plants, p. 155.) Malpighi, however, from the fact of the alternate separation and conjunction of the vessels, and from analogy with what occurs in the animal system, speaks always of the anastomosis of the vessels, but he nowhere gives us any thing like proof of the fact; and Du Hamel, from actual dissection of several fasciculi, regards them in their union as resembling more the nerves than the blood-vessels of animals. When, indeed, we consider the extraordinary minuteness of the vessels, and the circumstance of their possessing nearly the same size in every part, there is no room for that continual ramification out of one into another, and consequent diminution of diameter, that occur in the vessels of animals; and the immensity of their number, together with the endless separations and re-unions which their fasciculi make, seems calculated to fulfil the purposes of less general distribution in the plant, which successive division and perpetual anastomosis effect in the animal system.

Another general circumstance in the vessels of plants is, that we do not discover in them any structure which has the true nature and use of valves, similar to what is met with in the veins and absorbent vessels of animals. Dr Hooke could never observe in their canal any thing that had the appearance of valves. (Micrographia, p. 116.) Did such a structure exist, the absorption of nutrient matter from the lobes of the seed, and its conveyance, in a backward course, to the embryo, could not, says Grew, have place; neither could the root, as it often does, grow upward and downward both at once. (Anatomy of Plants.) If the piece of a root of elm-tree be cut in autumn, the juices, says Du Hamel, are found to escape indifferently at either end, as the one or the other is alternately held downward; a circumstance, he observes, inconsistent with the opinion of Mariotte, who maintained the existence of valves. (Phys. des Arbres, tom. i. p. 56.) It is well known, also, that many plants may be made to grow in an inverted position, so as to put forth leaves and flowers from their roots; and large trees have been

Elementary nourished by juices received through the extremities of their ingrafted branches, after all connection between the earth and the roots had been cut off. In general, indeed, the extreme minuteness of the vessels seems almost to preclude the possible existence of valves in their canal; but in some instances, where the vessels in aged trees have become enlarged, membranous productions have been found to occupy their cavities, which some have alleged to perform the office of valves. They occur, however, only at an advanced period of growth, and form no necessary part of the structure of the vessel, and, instead of promoting, contribute only to obstruct the course of the fluids.

Thus far with regard to the general nature of the vessels of plants: let us next discriminate their several species or kinds.

ART. II.—Of the Common Sap-Vessels.

To ascertain the kinds and situation of the vessels of plants, various means have been employed. The plant has been dissected both in its dry and recent state; the natural qualities and movements of its fluids have been observed; and its vessels have been filled with coloured liquors, by causing it to vegetate in them. By the combined use of these several means, many important particulars have been ascertained; but it must be acknowledged that the question is still beset with doubts and difficulties, and that, with relation to it, great diversity of opinion continues to prevail. A concise statement of the facts ascertained with respect to the movements of the vegetable fluids may perhaps contribute to define the situation and kinds of the vessels that convey them.

It has been proved that, early in spring, before the leaves appear, a watery fluid rises abundantly in the woody part of the trunk of trees, and continues visibly to ascend to the very extremity of the branches, until the leaves are developed, when, to appearance, it ceases to flow, and can no longer be collected by perforating or tapping the tree. This fluid has been shown to ascend through the wood, and to rise, in general, most abundantly through its youngest or outmost circle; but in trees whose vessels have not been obstructed from age or other causes, it rises through every circle to the very pith, and, as far as can be judged, in all the vessels that compose those circles. At this early period of vegetation no fluid is found in the bark, nor between it and the wood, nor in the pith; but the vessels of the bark are perfectly dry. These facts are deducible from observations on the natural flow of the fluids by Grew, Du Hamel, Walker, and others; and are supported by various experiments of M. de la Baisse, Bonnet, Reichel, and others, made by causing plants and parts of plants to grow in coloured liquors, in which the vessels of the wood alone became filled; but no tinge of colour was communicated to those of the bark. To these vessels the several names of lymph-ducts, sap-vessels, ligneous tubes, ascending and aduent vessels, have been applied: we shall in future denominate them sap-vessels.

The vessels which thus form the mass of wood have by some writers been distributed into different kinds, and supposed to exercise very different functions. At certain periods of vegetation they appear empty; and hence Malpighi supposed two species of vessels to exist in the wood, one destined to carry sap, and the other to convey air: to these latter, from their supposed office, he gave the name of trachea, and from their structure called them spiral vessels. (Anatom. Plantar. Idea, p. 3.) Grew also believed these empty tubes to be air-vessels, but ad-

mitted that at certain seasons they carried sap. At an early period, however, Ray maintained that the vessels thus supposed to convey air were truly sap-vessels; and Du Hamel, in common with Grew, admitted that they carried sap in spring. Hill considered them altogether as sap-vessels; and Reichel, Hedwig, and others, by filling them with coloured fluids, proved that such was their true office. On the other hand, no experiments, says Ludwig, have yet shown that there exist in vegetables vessels destined to convey only air; and in this opinion subsequent writers, with few exceptions, have acquiesced. It will, however, be convenient to treat of their general nature and form under the distinct appellations of sap-vessels and spiral vessels, by which they are commonly known.

By Grew and Malpighi the common sap-vessels were regarded as entire tubes, having no apertures but in the direction of their length. Grew represents a single vessel as having the appearance exhibited in fig. 1, Plate XXXVIII., the aperture or canal of which is not visible unless highly magnified, as in fig. 2. According to Malpighi, these vessels send off numerous capillary filaments to the cellular tissue, so that the cells are surrounded by a plexus of vessels, as is particularly seen in the pith of elder and some others; and these ramifications, he adds, spring probably from the perpendicular vessels both in the bark and wood. (Anatom. Plantar. p. 29. Lugd. Bat. 1688.) These lateral ramifications were observed also by Lecuvenhoeck in a piece of fir-wood newly felled. Of this wood he procured a longitudinal section, so extremely thin that he could see distinctly the particles of fluid moving in the vessels, as represented in the upper portion of fig. 3, Plate XXXVIII.; while lower down, on many parts of these vessels, small points or dots were visible, which he at length discovered to be round apertures; and as he did not see these apertures in any other parts than those in which he had separated the horizontal cellular tissue from the perpendicular vessels, he concluded that, at these points, these two organs were connected. He farther separated two of the vessels from the remainder, and through the microscope they appeared as in fig. 4; but the "engraver," he adds, "said that he could not possibly draw all the jagged parts that he saw; and we both of us perceived, in the broken membrane or coat of the tube, many excessively minute vessels, which, by reason of their smallness, he was unable to express in the drawing." (Select Works of Lecuvenhoeck, by Hoole, vol. i. p. 12.)

It may however be said, that this communication between the vessels and cells is maintained not by ramifications from the vessels, but through apertures or pores in their sides; and, accordingly, many appearances have been remarked as existing on the sides of the vessels, of which different authors have given very different representations, and which some have regarded as pores. Thus, on the vessels of the fir Malpighi observed certain dotted appearances, which he describes as roundish tubercles, and which were so numerous that they appeared to cover the vessels. On the vessels of the elm, the beech, and the willow, Lecuvenhoeck saw similar particles, which resembled small globules. (Select Works, vol. ii. p. 1.) Dr Hill describes the vessels of the alburnum, or newly formed wood of the willow, as connected with each other by a flocculent interstitial matter. When, by long maceration, this matter is detached, the vessels then exhibit a dotted appearance; and, if examined by a highly magnifying power, these dots, according to him, are so many oval swellings, and each has, as it were, a mouth. Through these mouths, which he represents as innumerable, and

Elementary existing on all parts of the vessels, he conceives the fluids to be discharged into the cellular tissue. (On the Construction of Timber, p. 18.) In fig. 5, Plate XXXVIII., is a representation of these vessels connected by flocculent matter, with their extremities collapsed from the escape of their juices, and their sides sprinkled with the little mouths which he mentions. These mouths, if they exist at all, are probably not pores in the sides of the vessels, but the little apertures seen by Leeuwenhoek, and produced by the separation of the cellular tissue while the parts are still young and tender. The same author, speaking of the vessels of the mature wood of the pear, states them to be close canals, as in fig. 6, with no lateral apertures in them.

A still later writer, M. Mirbel, declares, that not fewer than five species of vessels are to be found in the woody part of plants. These he denominates porous tubes, cleft tubes, tracheæ, mixed tubes, and vessels en chapelet, from their supposed resemblance to a string of beads. Of these several species we have given representations in figures 7, 8, 9, 10, 11, and 12, Plate XXXVIII. The first species, or porous tubes, according to this writer, exist in every part of the plant where the sap moves with freedom. Their sides are covered with small eminences or projections, in the centre of which is to be found a small pore. (Exposition de la Théorie de l'Organisation Végétale, p. 107.) Improving a little on Hill, he represents these pores not as promiscuously placed, but as ranged in transverse lines (fig. 7); and through them he conceives the fluids of the plant to percolate, not, however, into the cells only, but out of one layer of tubes into another, in a lateral direction. In this manner he conducts the fluids from the centre to the circumference of the wood, and at length, by a route not so easily followed, contrives to get them into the vessels of the bark, the sides of which he declares to be perfectly entire, and alike destitute of pores and clefts. (Ibid. p. 297.)

Several writers have sought to discover these alleged pores in the sides of the vessels. With Malpighi, some regard the appearances observed not as pores, but as elevations on the surface of the vessels; others, as particles contained within them. Under a very highly magnifying power, Kieser thinks he has, in some vessels, detected the existence of pores; while M. Dutrochet pronounces the same objects to be corpuscles, filled with nervous matter; and De Candolle suggests the probability of their being small glands, destined in some way to contribute to the nutrition of the plant. These different opinions, formed on viewing the same objects, sufficiently manifest the difficulty and uncertainty of microscopical observations made with highly magnifying powers, as was long since pointed out by Hooke, who first applied the microscope to the examination of the structure of plants. "Of such minute objects," says he, "there is much more difficulty to discover the true shape by an instrument than by the naked eye; the same object quite differing in one position to the light, from what it really is and may be shown to be in another. In some objects," he adds, "it is exceedingly difficult to distinguish between a prominence and a depression, between a shadow and a black stain, or a reflection and a whiteness in the colour; and the transparency of most objects renders them yet more difficult than if they were opaque." (Micrographia, Preface.)

ART. III.—Of the Spiral Vessels.

Various as have been the opinions of writers respecting the common sap-vessels of plants, they have differed yet more in their views concerning the position,

number, size, structure, and uses of those which have been denominated spiral. The common fathers of Vegetable Anatomy, Grew and Malpighi, who at the same time, but in different countries, prosecuted their inquiries without any knowledge of or communication with each other, are nearly of one opinion on all the more important points in relation to these vessels. Later writers have differed alike from them, and from each other, on almost every point. As the subject is of fundamental importance in the economy of vegetables, we shall endeavour to set before the reader the leading facts and opinions concerning it; to canvass their relative merits; and deduce from the whole such conclusions as seem most nearly to approach the truth.

Malpighi describes spiral vessels as existing in the ligneous parts of all plants. He called them spiral tubes, because, when extended, they were resolved, not into separate rings, but into a single zone, which might be drawn out to a great length. In general they form continuous tubes, but are sometimes contracted at regular distances, so as to resemble somewhat a series of oblong cells. One of these contracted spiral vessels, as delineated by Malpighi, is represented in fig. 17, Plate XXXVIII., at the extremity of which the spiral filament is in part drawn out; and similar appearances of the spiral structure are exhibited in figures 9 and 10, by Mirbel. In herbs, according to Malpighi, these spiral vessels constantly accompany the common sap-vessels, and are ensheathed by them; in shrubs, they occur in every part of the wood, single or in clusters; and in trees, an intermixture of spiral vessels with the common sap-vessels is observed in every part of the wood. In the fir, they are found immediately beneath the bark, and are so numerous as to constitute the chief bulk of the wood. They exist with the sap-vessels in the petioles and ribs of the leaf, and likewise in the petals of the flower. In the roots they are also met with, and in some roots are so abundant, as to exceed in bulk all the other parts. They exist, he adds, in every part of the plant except the bark; and are annually formed with the albuminous vessels of trees. (Anatom. Plantar. passim.)

From the writings of Grew we collect also that of Grew, spiral vessels exist in every part of a plant except the bark. In the root they are numerous, of very various size, and their bore is generally larger than that of the common sap-vessels. In the trunks both of herbs and trees they are equally visible, and in position, size, and number, subject to great variation. Sometimes they are collected into fasciculi, at other times they are disposed in rays, and in other instances they are arranged in a circular form: they stand sometimes next to the pith, in other instances next to the bark, and in other cases again they alternate with the common sap-vessels in every part of the wood. They have a more ample bore than the common vessels, and vary in size to at least twenty different degrees. In the leaf they always accompany the sap-vessels, and both in its petiole and ribs are constantly surrounded or ensheathed by them. They have a similar position in the petals of the flower, and in the vascular parts of the fruit. (Anatomy of Plants, passim.) Hence, then, it appears, from the dissections of Malpighi and Grew, that, in every plant in which vessels are to be seen, and almost in every part of it, spiral vessels abound; they exist, however, only in the ligneous portion, or that part in which the sap ascends, and are never to be found in the bark.

By Du Hamel, the spiral vessels, supposed to convey air, are described as existing in the leaves and the flowers, the petals of which are almost wholly composed of them. In the herbaceous portion of young branches

Elementary they are also well seen; which portion afterwards becomes ligneous; so that it cannot be doubted that they exist in the wood, though he could never discover them but in young branches. (Phys. des Arbres, tom. i. p. 42.) If, however, all the empty vessels seen in a transverse section of the wood be deemed air-vessels, and all air-vessels have a spiral structure, then, says this author, they would, in many plants, form a great part of the ligneous body. From these large vessels, however, he has seen, in autumn, fluids to escape; so that they are not properly air-vessels, or, as Grew observed, they sometimes carry sap. They are not to be found in the bark.

Opinion of Hill, Under the common name of sap-vessels, Hill delineates all the varieties of vessels that constitute the wood; they are largest in the outer circles, smaller in the others; they contain, says he, in spring and at midsummer, a limpid liquor; but at all other seasons they appear empty, and have therefore been erroneously deemed air-vessels. He says nothing of their spiral construction; but describes the vessels which form the chief mass of the wood as possessing solid and firm coats, forming an arrangement of plain and simple tubes, as in fig. 6, Plate XXXVIII., resembling those of the albuminum, except that they have no mouths in their sides. (On the Construction of Timber, p. 8 and 19.)

of Reichel, M. Reichel maintains the existence of spiral vessels in almost all parts of the plant. By causing plants to vegetate in coloured fluids, he traced them from the roots, through the trunk and branches, to the extremity of the leaves, and into all parts of the flower,—as the calyx, the petals, the style, the stamens, and the anthers. In fruits and in seeds, and in the radicle and plume of the latter, they were equally apparent. The coloured liquors, as they rose, communicated a tinge to the cellular tissue of the wood, as was previously observed by De la Baisse; but no trace of colour was ever observed either in the bark or pith, which therefore contain no spiral vessels. He considered the spiral vessels as the organs everywhere conveying nutrient matter to the plant, and as having no title to the appellation of air-vessels. (Encyclop. Méthodique, article Physiol. Végét. p. 288.) Similar experiments of Hedwig and others confirm these facts as to the general distribution of the spiral vessels, and their bearing to every part the fluids absorbed by the roots.

of Mirbel, We have already enumerated the five different species of vessels which Mirbel regards as constituting the woody part of plants. According to him, true spiral vessels (which he continues to denominate tracheæ) are not to be found in the root, but only in the trunk and the parts which are produced from it. Even in the trunk they are to be found only around the pith, and never in the exterior ligneous layers. He admits that they exist in all soft and succulent parts, and that coloured fluids rise in them as well as in the other varieties of vessels; but they never are to be found either in the bark or pith. (Exposition de la Théorie de l'Organisation Végétale, p. 74, 78.)

of Kieser, According to M. Kieser, spiral vessels are found in all the more perfect plants, from the earliest period of their existence; and in all parts of them, except the bark and pith. They vary much in size, and in aged plants are frequently obstructed by a species of vesicle, which originates from the interior sides of the vessels. According to him, nothing but air is found in their cavity, though, in the wood of guaiaecum, he has seen all the spiral vessels, in common with the cellular tissue, completely filled with resinous matter. He does not know the anatomical relation subsisting between the spiral vessels and the other

organs; but thinks it proved that no direct communication exists between them and the cells. (Mém. sur l'Organisation des Plantes, 1814, p. 116, 117.) Lastly, M. de Candolle admits the existence of spiral vessels in the wood of the trunk, but not of the roots; and, with Kieser, regards them as conveying only air. (Organographie Végétale, tome i. p. 39.)

From the combined observations of all the preceding writers, with the exception of MM. Mirbel and de Candolle, we learn that spiral vessels exist not only in every part of the trunk, but in the root, and in every other part of the plant except the bark and the pith. From the actual presence of sap in these vessels early in spring, and again in autumn, and from the entrance of coloured fluids at all seasons, we also learn, that their true office is not to convey air, as Malpighi and others formerly, and Kieser and others more recently, maintain; but to carry sap. Are we therefore to regard them as a species distinct from the common sap-vessels, or are both to be held merely as varieties of one common kind? This question may receive some illustration from a little farther inquiry into the formation and structure of the sap and spiral vessels.

ART. IV.—Of the Structure of the Sap and Spiral Vessels.

Malpighi seems to have regarded the common sap-vessels as being formed of a series of small vesicles or cells, mutually opening into each other. (Anat. Plantar. p. 28.) In this instance he appears either to have mistaken a series of elongated cells for vessels, or to have conceived that the contractions, occasionally formed in the vessels themselves, afforded evidence of their having been constructed originally by cells. Grew, on the other hand, considered these vessels to be composed of straight fibres, placed parallel to each other, so as to form a cylindrical tube. (Anat. of Plants, p. 112.) Leeuwenhoek held them to be composed of two fine transparent coats, formed of other vessels excessively minute. (Select Works, vol. i. p. 11.) And Du Hamel, by submitting thin layers of wood to long maceration, obtained bundles of longitudinal fibres of extreme minuteness, by which he considered the lymphatic vessels of the wood to be formed. (Phys. des Arbres, tome i. p. 32.) Hill describes the albuminous vessels as being formed of the same material as the cells, and to collapse when emptied of their fluids. In one instance, where a strong light was made to penetrate the vessel, it appeared as if composed of numerous cells; but on farther examination, these seeming divisions altered their places, and were found to proceed from portions of watery sap still retained in the vessel. This appearance, as he properly observes, may be a very necessary lesson against hasty judgments. (Construction of Timber, p. 33.)

According to M. Mirbel, "the entire mass of the plant is nothing but cellular tissue, the cells of which differ only in form and dimensions." The cells and vessels are thus considered to be formed out of one and the same membranous tissue. In forming cells, this membrane is supposed to dilate in every direction; in producing vessels, it increases only in length. (Exposition de la Théorie de l'Organisation Végétale.) The manner in which he supposes the vessels to be formed out of the primitive membrane is very fanciful; and he is probably right in thinking that "no one before himself had formed a similar conception of vegetable organization."

Concerning the structure of the spiral vessels, opinions have varied still more than in relation to those just of spirals considered. Malpighi describes them as composed of vessels.

Elementary a thin pellucid plate, of a silvery colour, somewhat broad, and which, being placed spirally, and united at its edges, constructs a tube. When this tube is drawn out, it does not separate into distinct rings, but is resolved into a continuous spiral zone. At particular places these vessels are sometimes contracted, so as to exhibit the appearance of oblong cells opening into each other, as in fig. 17, Plate XXXVIII. In size they greatly surpass the ordinary saps-vessels; and their canal is then frequently occupied by membranous vesicles, which nearly fill their cavity. (Anat. Plantar. p. 3-26.) These vesicles were also observed by Leeuwenhoek, and their appearance, in the transverse section of an enlarged vessel, is represented in fig. 25.

According to Grew, the thin plate that forms the spiral vessel is not flat, and, instead of being single, consists of two or more round filaments or threads, placed collaterally, but perfectly distinct. These component filaments he regards as united by other smaller transverse filaments, and the thin plates which they form by their connection constitute the spiral vessel,—"as if we should imagine," he adds, "a piece of fine narrow riband to be wound spirally, and edge to edge, round a stick, and the stick being drawn out, the riband to be left in the figure of a tube, answerable to a spiral vessel." As, however, the riband is composed of numerous threads, placed parallel to each other, so is the plate that forms the spiral vessel; and it is according to the greater or less delicacy of the vessel examined, and the manner of its dissection, that it appears to be constituted either of a flat plate or a round filament. The spiration of the filaments he considered to be made in the root from right to left, and in the trunk from left to right. (Anat. of Plants, p. 73 and 117.)

The opinion of Du Hamel with respect to the construction of these vessels was very similar to that of Grew, and he employed the same analogy of a riband twisted round a stick to illustrate it. Hedwig gave a very different account of them. He considered the spiral vessel to be composed of two distinct parts; one a membranous canal conveying air, and the other a spiral tube rolled round it, by which the fluids were conveyed. The spires of the tube, in some instances, are represented as close; in others they are separated, and the intervening portions of the membranous canal exhibit a dotted appearance. He considered all the sap-vessels, from their first formation, to possess this compound structure, and, by a series of changes, which he professes to describe, to be ultimately transformed into the solid fibre of the wood. (De Fibra Vegetab. Ortu, p. 25.) Others, also, have believed the spiral vessel to be formed of a membranous tube, but have denied that it conveyed air; and the spiral tube of Hedwig they have regarded as a solid filament. From the different appearances which they exhibit, MM. Treviranus and Bernhardt distinguish several varieties of spiral vessels; as does also M. Kieser, whose account of these vessels is not only the most elaborate, but that probably which approaches nearest to truth. We shall therefore subjoin a brief abstract of his observations on the structure of these vessels, and more particularly on the series of transformations which they seem to undergo.

The construction of these vessels M. Kieser professes to have studied with the greatest care, and to have established incontestably the following points. Sometimes, says he, only one fibre, sometimes many, go to form the spire of a vessel. These fibres are commonly round, sometimes a little flattened, and are twisted spirally about an empty space, so as to form a tube. The spiral fibres in young plants, and sometimes in mature ones, are in close contact; in other instances they are separated, and the interstices are then occupied by a dotted or punctuated

membrane, as is very frequent in trees; or sometimes they are connected by ramifications which proceed from the spiral fibres themselves. From the minuteness of the spiral fibre, it is difficult to decide whether it is a solid or tubular body, and often, from the same cause, to pronounce whether it is round or flattened. It is transparent, has considerable consistence and tenacity, and in some plants appears to possess elasticity.

The number of fibres that form the spire of a vessel is very various; sometimes, as before remarked, there is only one, but more often several, which are twisted in the same plane and the same direction. He has seen nine fibres thus united; and when unrolled, the spires of the vessels then seem to form a kind of riband. When many fibres are employed to form a spire, they always run parallel, and do not cross; so that the side of the vessel has never more than the thickness of a single fibre. At their first formation the fibres are not united; it is only in more advanced age that they become united, either by ramifications from their edges, or by a peculiar membrane. The direction of the fibre in twisting is sometimes from right to left, and sometimes the contrary. The size also is very various. In young plants it is very small, so that a millimetre comprises the diameter of 2600 of these fibres; and in some instances they have little more than half that size. In older vessels they have a larger size. Between the knots of the trunk they are simple, and do not ramify; but in the knots they undergo great changes of form, and are variously ramified and combined.

ART. V.—Of the Transformations of the Spiral Vessels.

It was the opinion of Grew, that the "common sap-vessels begin to be formed in spring, but the spiral vessels not till the end of summer, at least not till about that time do they appear." (Anat. of Plants, p. 131.) Malpighi also describes these vessels as gradually appearing in the alburnum, and augmenting in size in every successive ring of wood; an observation made also by Grew, who describes the "spiral vessels as being amplified in each annual ring, so as to form a vessel of a wider bore." These facts indicate that the spiral vessels, after their formation, undergo considerable changes in the progress of vegetation, or, to use the language of Grew, that "they are postnate, and seem produced by some alteration in the quality, position, and texture of their fibres."

In treating of the transformations or metamorphoses of the spiral vessels, M. Kieser illustrates his ideas by a series of dissections of the stem and root of the gourd (cucurbita pepo), made apparently with great care. These designs occupy several plates of his work; and being taken from different parts of the same plant, at different stages of its growth, and viewed with the same magnifying power, they show the progressive changes in form and appearance of these vessels as the plant increases in age. M. Kieser exhibits corresponding vertical and horizontal sections of the same parts; but our object will be answered by copying only certain portions of the vertical sections.

The ligneous portion of the stem of the gourd possesses ten bundles of vessels, symmetrically disposed around the pith, and standing at unequal distances between it and the bark. In fig. 13, Plate XXXVIII., we have copied the representation of a vertical section of the vessels in one of these bundles, taken from a portion of the plant near the summit of its stem, and made with the aid of a microscope magnifying 130 times. The number of vessels in each bundle at this period is said to vary from three to five. They are constructed of one or more fibres, placed

Elementary contiguous to each other, and twisted into a spiral form, producing a round cavity within. These are called simple spirals. They are found in every young plant, and in the newly formed parts of old ones. Their size is smaller than that of the other varieties to which they give origin. The number of fibres in each spire varies from one to nine; and they are the only variety met with in some of the inferior vegetables.

The next figure (fig. 14, Plate XXXVIII.) exhibits a similar section of the vessels of the same plant at the internodial space below. Of the four vessels contained in this bundle, the outer one, a, is larger than the rest, but its spires are contiguous, as in those of the former figure; the inner one, i, is formed of a series of rings, disposed in a perpendicular line. These rings are very like the spires of the preceding variety, and often combined with them in the same vessel. They are sometimes separated from each other only by a space equal to their own diameter; in other instances to eight or ten times that distance, forming then the substratum of the next variety. In their present form they constitute the annular spiral; and in herbaceous plants occupy frequently the same position as the simple spiral, that is, next to the pith.

From the two simple varieties of form just described, other more complex forms are produced, as the plant advances in age. In the next section, made still lower down on the same plant, the vessels exhibit the forms represented in fig. 15 of the same plate, which brings the third variety of spiral vessel into view. The exterior vessel, b, of this fasciculus, or that nearest to the bark, is of much larger size: its spires, represented by the white lines in the figure, are separated to nearly equal distances from each other, and the intervening spaces are filled up with a membrane, sprinkled over with small obscure points or dots, constituting the punctuated spiral vessel. In this variety the rings are never contiguous. Such vessels occur only in the more advanced age of herbaceous plants, but are said to be original formations in the albumen of trees: they acquire a much greater size than the two former varieties, especially in the stem; but do not become so large in the root. In herbs they occupy a position exterior to that of the two preceding varieties; but in trees the largest vessels of the annual layers are nearer to the pith. The three other vessels in this figure are simple spirals, in part unrolled.

The spiral fibre in these vessels is of various size, being largest in the greatest vessels: it is sometimes so small as to be scarcely visible. It is often difficult to determine whether its form is spiral or annular: for the most part it is spiral in herbs, and annular in trees. The membrane which connects the spires or rings in this variety is not visible in young vessels, but appears in more mature age. It is at first transparent, but becomes opaque from age. In this punctuated variety of vessel the spires cannot be unrolled without tearing the connecting membrane. As to the points or dots observed on this membrane, they have been taken by some for pores on its surface; by others for clefts produced by the joining of the spires. They are clearly not clefts, says M. Kieser, and it is against the supposition of their being pores that their position has a relation always to the spires, and not to the adjoining cells. In the vessels of some plants, as in those of the French bean, when magnified 400 times, these points, fig. 21, a, seem as if pierced by a hole exceedingly small; but in other instances they are quite dark, and similar to those found on the membranous productions that fill the cavity of aged vessels, where there is not the smallest reason for considering them as pores. There seems a great analogy between these con-

nnecting membranes and those which thus fill the cavity of aged vessels. In some instances the size of the dots is large, and they appear transparent at the centre; in others the dots are extremely small. Their form is oblong, and they are ranged generally in determined lines, parallel to the direction of the spires.

In the next vertical section of a fasciculus of vessels, taken from about the middle of the stem of the same plant, all the peculiarities of size and form are still more distinctly seen. Of the twenty-three vessels which appeared in the horizontal section of a fasciculus at this period of growth, six only were apparent in the vertical section, and are exhibited together in fig. 16, Plate XXXVIII. Of these, the first two, or outer ones, are by much the largest, and belong to the punctuated variety. A portion of the anterior wall of these two vessels has been removed by the sections at l, l, so that the inner surface of the posterior wall is brought into view; and in the former vessel, d, a portion of the posterior wall itself has also been carried away. In both vessels the spiral fibre, indicated by the white line, is prominent in the section; and in the second, e, it is unrolled at m, n,—where, at m, it is seen single, transparent, and round, while at n it is united with the punctuated membrane that connects the spires, and has a flat or riband-like form. The third vessel, f, of this fasciculus is also of the punctuated class, and the three last, g, h, i, are simple spirals; the last of all, i, being in the progress of formation, and its spires not yet brought into a state of contiguity.

There is a fourth variety of vessel that is said to have the same origin as the last, being formed in part by spires or rings, but the separations between which are filled up, not by membrane, but by small productions or ramifications from the spires or rings themselves; and these ramifications are often so implicated as to form a net-work; whence the names of ramified and reticulated spiral vessels. Like the former variety, they do not exist in the young plant, but are formed gradually in a more mature age, and by the same series of actions: that is, the spires or rings, which were at first contiguous, separate at a later period, and the intervals are filled up by ramifications from the spires themselves. At first these ramifications are few, as in the vessel, a, fig. 21, when they are termed ramified spirals; and as these ramifications increase, the vessel becomes reticulated, as exhibited at the letter Reticulated, fig. 20. They are found but in few plants, and appear to differ from the punctuated variety chiefly in the position of the rings, which are more or less obliquely placed, and at different points send off one or more branches.

In his ninth plate, M. Kieser has given similar representations of the vessels of the same plant, as they appear from sections made near the root. The fasciculus at this part of the stem contained twenty-nine vessels, as seen in the horizontal section, of which only eight were visible in the vertical section. These retained nearly the characters of those given in the preceding section, fig. 16; but exceeded them in number and size. The large ones, situated next the bark, were of the punctuated class, while the smaller ones, next the pith, were simple spirals. In fig. 18 we have copied the three smaller vessels. In one of these, o, which belongs to the punctuated variety, the interior of the vessel is nearly filled up with small portions of punctuated membrane, produced from the inner wall of the vessel itself, as is still more clearly seen in horizontal sections of similar vessels in fig. 25. The two vessels p, q, of fig. 18, are the youngest, and consist only of the simple spiral fibre; the last, it is said, being composed of two such fibres, which, in consequence of the transpa-

Elementary necessity of the posterior wall of the vessel, appear to cross each other.

But M. Kieser has not confined his examination to the stem of this plant. In his tenth plate he has given similar sections of a portion of the root. The piece of root he examined presented, in the horizontal section, four fasciculi, each of which contained about thirty-seven vessels, varying in size: the larger ones being placed exteriorly, and the smaller ones towards the centre of the root. The cells of this part were small, and nearly of the same size and form throughout, differing much from the varying size and form of the cells in the stem. Of the thirty-seven vessels before mentioned, only ten were retained in the vertical section, all of which belonged to the punctuated variety. Of these, two had united so as to form but one cavity; and two others were filled up with membrane. In fig. 19 we have copied the appearance of three of these vessels. One of these, indicated by the letter r, is from the middle of the figure, and its size is intermediate between those on each side of it. A portion of its anterior wall has been carried away in three places, which brings the inner side of the posterior wall into view. The vessels s t of the same figure are the smallest, and situated at the centre of the root: all the three belong to the punctuated variety, no simple or annular spirals being found at this period in the root.

In the knots of plants the vessels undergo great changes of form, the larger vessels being contracted in different parts, and giving off productions or ramifications which form new vessels, and give to the whole a very irregular appearance, as may be seen in the upper part of fig. 20, taken from the appearance of the vessels in the knot of the balsam (Impatiens Balsamina L.). In this figure M. Kieser has also given an ideal representation of the various metamorphoses already described, viz. of the annular vessel into the simple spiral, and of that into the other varieties: thus the letter a represents the rings of the annular vessel, which at b begin to form the spiral: at c the simple spiral passes into the ramified variety, and this again, at d, into the reticulated vessel: at e the vessel becomes contracted at different parts, as previously observed by Malpighi, fig. 17, and by Mirbel, fig. 11 and 12. This latter author called this the vessel en chapelet, from its resemblance to a string of beads. From these vessels go off the branches f f f, to form the ramified spirals of the knot. In fig. 12 of the same plate M. Mirbel has given a similar ideal representation of the five species delineated by him in figs. 7, 8, 9, 10, and 11; but they do not carry with them the same evidence of accurate observation as the several varieties indicated in the alleged transformations of Kieser.

Such is the account given by M. Kieser, of the development and appearances of the vessels in herbaceous plants, and of the successive changes in character which, in the progress of growth, they undergo. All the figures except the last are taken from dissections of the same plant, beginning with that part of the stem which was of latest growth, and proceeding in succession to the parts of earlier growth, or older formation. At first the vessels are comparatively few in number, of nearly uniform size, and belong to the varieties of annular and simple spiral: at a later period those first formed become augmented in size, and assume the characters of punctuated and ramified vessels; they are now also removed farther from the centre, and their place is supplied by new vessels of simpler forms. With the age of the plant, they all continue to increase in size and number, and to acquire the new characters which distinguish the several varieties in the manner already described.

In the early growth of trees, the formation and development of vessels go on at the centre, adjacent to the pith, in the same manner as in herbs; but at a later period, when the additional growth is made at the circumference of the tree, the first formation of vessels must occur in the album, and the subsequent changes of character be effected in the several layers of wood; which accords with the fact that spiral vessels are rarely to be detected in the album, but are found with all their different characters, and of all sizes, in the several layers of wood, as may be seen in all the transverse sections of wood delineated by Grew, Malpighi, and others. In truth, the descriptions of Kieser with respect to the structure and transformations of the spiral vessels accord, in most respects, with the opinions of Grew, as briefly stated, page 47 of this article, and as may be further seen by referring to pages 73 and 177 of Grew's Anatomy of Plants.

It is further to be remarked that, in these representations of Kieser, which exhibit all the vessels visible in all parts of the stem and root of this plant, only one kind of vessel is described, which at different periods of growth assumes different forms; that no vessel answering to the common or lymphatic vessel of authors is mentioned; and consequently, that all the vessels described, however differing in size and external character, are in truth but varieties of one common kind. Calling to mind also that, in the earliest production of leaves and of flowers, and even of stamens and pistils, vessels either possessing the spiral character, or capable of assuming it, are everywhere met with,—that they occur also in the stem adjacent to the pith in succulent plants, and in the album of trees, and every successive ring of wood produced from it—in short, in every part and portion of the plant, except the bark and pith,—and believing, as we do, in those changes of form suggested by Grew, and subsequently demonstrated by Kieser,—we are led to the general conclusion that only one kind of vessel exists in the ligneous portion of other plants, as well as in that of the gourd; and that in different plants, and in different circumstances and conditions of vegetation, this vessel is capable of assuming the different forms, sizes, and characters, which have led many writers to describe these organs as constituting different species, and serving different uses.

M. de Candolle, in his late work on the Organography of Plants, seems also to reject the separate existence of the ordinary sap-vessels or lymph-ducts of authors, and to consider, with Kieser, all the vessels collectively to possess a spiral character, which appears under five different forms in the vessels severally named trachea or simple spirals, annular spirals, punctuated and reticulated spirals, and the vessels en chapelet. (Organographie Végétale, tome i. p. 32.) Without deciding positively in favour of Kieser's views, he admits that the different orders of vessels have among them very intimate relations; and in favour of the opinion that they are really modifications of each other, he alleges that almost all these orders exist together in certain classes of plants, and are wanting altogether in others. Still, though he has no theoretical objection to make to these alleged transitions of form, he has never been able to verify them by actual observation, and thinks the facts require further investigation. (Ibid. p. 48, 51.) With respect to the existence of pores in the spiral vessels, as alleged generally by Mirbel, and even occasionally by Kieser, but denied by a great number of observers, M. de Candolle has been led by his own observation to entertain doubts, and to believe that what has been regarded as a pore is a luminous point, such as is exhibited by bubbles of air in water, when placed under the microscope. (Ibid. p. 55.)

The late opinions of M. Dutrochet on the structure of plants seem to us to have drawn attention more from their novelty than their value. He criticises the opinions of Mirbel in relation to the structure of the vessels, and says that those denominated by that author porous tubes, are altogether destitute of pores, but have corpuscles adhering to, or rather implanted in their sides, on which account he calls them corpusculiferous tubes. (Recherches Anatom. sur la Structure des Végétaux, &c. p. 23.) In like manner the tracheæ are said to have similar corpuscles attached to their sides, but which adhere only weakly, and form no necessary part of the organization. (Ibid.) With respect to the uses of these two kinds of vessels in the wood, he considers the corpusculiferous tubes to be the canals by which the sap ascends from the roots; while the spiral vessels, or tracheæ, convey not air, but a diaphanous liquid, prepared in the leaves, and which descends, by the spiral vessels, to impart its vivifying influence to all parts of the plant. (Ibid. p. 29-31.) No direct communication, it is said, exists between the cells and vessels which compose the vegetable texture; they are only contiguous. Fluids, nevertheless, are transmitted from one to the other; not however through holes or passages expressly designed for that purpose, as Mirbel contends, but through intermolecular spaces, left between the integrant globular molecules, of which the organic solids are composed. (Ibid. p. 47, 48.)

Such is an outline of the different opinions that have been held regarding the structure of the sap-vessels. Differences not less great still continue to exist with regard to their uses. Our reasons have been already given for coinciding in opinion with those who hold the vessels to be the proper channels in which the sap is conveyed. M. Kieser, however, contends that the spiral vessels of the wood are always empty, or contain only air, and considers them, with Malpighi, to be organs by which air is conveyed. To this opinion M. de Candolle also inclines, though he admits that in some cases they may convey lymph. (Organog. Végét. tome i. p. 61.)

Those physiologists who thus oppose the opinion that the vessels carry fluids, suppose the sap to be conveyed by certain minute channels, said to exist between the cells. The place of these alleged channels is marked out in a section of cellular tissue (fig. 29, Plate XXXVIII.) taken from M. Kieser, in which the black points at the angular junctions of the cells are said to denote them: from their position they have obtained the name of intercellular canals. By many, however, the existence of such spaces at the angles of the cells is denied; and, indeed, when we contemplate the various size and form of the cells, the different degrees in which they are filled with fluid or solid matters, and consequently the greater or less degree of compression which these alleged intercellular spaces must undergo, we can hardly admit the sufficiency of such minute, tortuous, and precarious channels to convey the ascending sap with that force and velocity which the experiments of Hales and others have proved it to possess. On the other hand, no direct communication between the spiral vessels of plants and the external atmosphere, similar to that which occurs between the tracheæ of insects and the air, has been shown to exist; and the total absence of spiral vessels in the bark, where they would be near to the atmosphere, and their presence in the root, where they are almost beyond its influence, are alike repugnant to the notion that they exercise a respiratory function. Lastly, the absence of sap in these vessels at certain seasons, on which alone the notion of their being tracheæ or air-vessels was founded, depends, as we have seen, on the period at which the examination is made;

for at certain periods they have been found filled with sap; and at all periods, when in active vegetation, they will absorb and convey coloured fluids. It should also be borne in mind that the respiratory function in vegetables is performed, not by the trunk, but by the leaves of the plant.

Dismissing then these notions concerning the uses of the spiral vessels, and adhering to the opinion, that in all their variety they serve the office of conveying sap through the plant, we shall conclude this branch of the subject with a brief notice of their modes of termination.

If we begin at the root, we may consider these vessels as terminating at that part, by continuation of canal, in the capillary absorbents of the rootlet. As, further, the vessels of plants, like those of animals, are the only organs which convey the materials out of which not only the fluid, but the solid parts of the plant are formed, it may, in a general sense, be said that their modes of termination are as numerous as the kinds of distinct parts or organs which the vegetable system contains. Hence, therefore, as the cellular tissue of plants contains various matters often precisely similar to those which exist in the vessels, it may be inferred that the vessels have a termination in those organs; and this inference may perhaps carry more weight with many than the attempts before related of Malpighi and Leeuwenhoek to demonstrate it anatomically. Another termination of these vessels must be in certain minute and ill-defined organs called glands, which separate from the mass of fluids peculiar secretions; and a fourth mode may, in the leaves, be in other vessels which carry back the juices from those organs. The last and fifth mode of termination is into transpiring or exhalant organs, by which a certain portion of the contents of the vessels is discharged: so that in plants the sap-vessels terminate externally at one extremity in absorbents, by which fluids are received; and at the other in exhalants, by which these fluids are in part discharged.

ART. VI.—Of the Proper Vessels.

By observation of the natural flow of the sap, combined with the results of experiments made with coloured liquors, we have endeavoured to determine the situation and kinds of vessels by which it is conveyed in the wood. The same method will best assist us in ascertaining the nature and place of the vessels met with in the bark.

It was before stated, that, early in spring, the sap of plants rises through the wood alone, and that no fluid whatever is then to be found in the bark. At a later period, however, the case is completely reversed; for the vessels of the wood no longer appear to carry sap, and those of the bark then become abundantly supplied with it. This difference is very clearly and concisely stated by Grew. "The sap," says he, "in many plants, as the vine, ascends visibly through the wood for a month, in March and April, and rises through every ring of wood to the very centre, yet at the same time there ariseth no sap at all out of the bark, nor between it and the wood." "But late in spring," he continues, "and in summer, the sap is no longer visible in the wood, but is abundant in the bark, in the inner margin adjacent to the wood." Du Hamel, too, remarks, that when the lymph rises abundantly through the wood in spring, the bark is in the dry, and adheres to the wood, and no sap then issues from it, nor from between it and the wood; but, later in the season, he adds, the bark yields abundance of sap. These statements have been verified by the multiplied observations of various subsequent authors.

But why is the sap thus present only in the wood at its first rising in spring? Why at a later period does

Elementary it cease to be visible in that part? And how and why does it afterwards find a passage into the bark? Some observations of Du Hamel, Hales, and Walker, point, we think, to the true cause. In spring, says Du Hamel, when the sap rises vigorously, the buds have not appeared; when they begin to open, the sap then flows less freely; and when the leaves are fully developed, then the flow of sap entirely ceases. Dr Hales also remarks, that, towards the end of April, when the young shoots come forth, and the surface of the vine is greatly increased by the expansion of the leaves, the sap then ceases to flow in a visible manner till the return of the next spring. All bleeding trees, he adds, cease to bleed as soon as the young leaves begin to expand enough to perspire plentifully, and draw off the redundant sap. The bark of oak, too, separates easily when lubricated with sap; but before the leaves appear and perspire, the bark will no longer run (as they term it), but adheres most firmly to the wood. (Vegetable Statics, 3d edit. p. 126.) In like manner, in an experiment of Dr Walker, a birch-tree bled from every perforation in its trunk, and from every cut extremity of its branches, until vernal or budding began; then the bleeding was almost immediately checked; and when the young leaves had pushed beyond the hybernaculum, the bleeding entirely ceased. (Edin. Phil. Trans. vol. i. p. 31.)

It is however certain, that though the sap was no longer visible in the wood after the leaves were developed, it continued nevertheless to rise through it; for in no other way could the leaves obtain the large portion of fluid which it is known that they constantly discharge by transpiration; and coloured fluids manifest their presence in the vessels of the wood, as well after as before the development of the leaves. The leaves, therefore, must be regarded as the organs which, by their perspiration, draw off, as Dr Hales observes, the redundant sap; and hence, in an experiment where a notch was cut two or three feet above the lower end of a stem, though a great quantity of sap passed by the notch, yet was it perfectly dry; because, says he, "the attraction of the perspiring leaves was greater than the force of trusion from the column of water." (Vegetable Statics, p. 111.)

Not only, however, does the development of the leaves render the sap no longer visible in the wood, but they also appear to be the organs by or through which it finds its way to the bark. In all the experiments just recited, the bark continued dry until the sap disappeared from the wood; in other words, until it was drawn off by the leaves; and then, and not till then, the bark became moist, and continued laden with sap through the rest of the summer. Now, as it has been before shown that no sap enters the bark by the roots, nor gets into it directly from the wood, there is no other known channel by which it can be conveyed, except through the leaves; and these, therefore, necessarily appear to be the organs by which it is apparently carried off from the wood, and by or through which it at the same time finds its way to the bark.

This inference appears to follow, not only from the fact of the bark continuing dry until the leaves are developed, but from the circumstance that it is again rendered dry after having become moist, if these same leaves be removed. If, says Du Hamel, we remove the leaves of a young tree when in full sap, and whose bark is easily detached, in a few days after the same bark will adhere as closely to the wood as it commonly does during winter. This direct connection between the leaves and bark is also well illustrated in an experiment of Hales, employed by him for a very different purpose. From two thriving shoots of a pear-tree he cut, in several places, half an inch of the bark off all round.

All the ringlets of bark between these incisions had a leaf-bud upon them except one, and all but this one ringlet grew and swelled at their bottoms till August; and the larger and more thriving the leaf-bud was, so much the more did the adjoining bark swell. (Veg. Statics, p. 149.) Mr Knight also found the bark of the vine to become shrivelled and dry when the leaves were stripped off; but in those parts in which it communicated directly with the leaves, it continued moist and flourishing. (Phil. Trans. 1801, p. 335.)

By connecting, therefore, the circumstances attending the flow of sap with the development of the leaves, we gain satisfactory reasons for all the apparent anomalies observed to attend its course. Early in spring, before the appearance of the leaves, no natural outlet for the escape of the rising sap exists, and therefore, when the vessels are cut or perforated, they readily pour out their sap, or bleed; but late in spring, and in summer, when the leaves are developed, the more watery parts of the sap are thrown off by transpiration; and while this process proceeds, the fluids do not accumulate in such quantity in the minute vessels of the trunk as to be effused or bleed through their cut or perforated sides. A cold day, however, or a moist and still atmosphere, by checking transpiration from the leaves, restores more or less the propensity to bleeding from the trunk; and in autumn, when fructification begins, and vegetation makes a pause, the same disposition to bleeding recurs in the trunk, from the check imposed on the more active powers of growth.

It is further evident, that, when the vessels of the bark become supplied with fluid, they cannot have de-sap in rived it immediately from those of the wood, since, in these different parts, the fluids have frequently no sort of agreement in properties. Thus, Grew remarks, that almost all plants, late in spring, and in summer, bleed from their bark; and the sap has either a sour, sweet, hot, bitter, or other taste. At this period the bark of the vine yields a sour sap; but, "what is vulgarly called bleeding in the vine is," he adds, "quite another thing, both as to the liquor which issueth, and the place whence it issues; that is, it is neither a sweet nor a sour, but a tasteless sap, issuing, not from any vessels in the bark, but from the air-vessels of the wood." (Anatomy of Plants, p. 125.) Malpighi, also, was well aware of the difference in the qualities of the sap, and thought every plant possessed its peculiar sap; but he has not so accurately defined the situation of the vessels that contain it. Du Hamel, however, points out distinctly the difference of quality in the sap of the bark and wood. In the bark of some plants it is white, in others red, and in others yellow. It is in some instances milky, in others resinous, and in others gummy. In many plants it has a sweet taste, in some it is caustic, and in others insipid. It has sometimes much odour as well as flavour, and frequently it is destitute of both. (Phys. des Arbres, tome i. p. 68.)

But if the qualities of the sap in the bark be thus acquired in different from those of the sap in the wood; if these peculiar qualities are detected in it only after the development of the leaves, and the leaves be the organs by which alone the sap can be conveyed from the one part to the other; then it seems to follow that the sap must acquire these new properties in the leaves during its transmission through them. Malpighi remarked the existence of this altered sap in the leaves, and held them to be the organs which prepared nutrient matter for the plant; and Dr Darwin, by immersing plants of spurge in coloured liquors, not only saw, as others had previously done, the red fluid ascend through the leaf, but another fluid, of a

Elementary white colour, returning at the same time from the extremities of the leaf, and descending into the petiole.

Organ. (Botanic Garden, vol. i. notes, p. 450-53.) This same returning fluid Mr Knight observed in similar experiments on branches of the apple and horse-chestnut trees, and even traced it through the petiole into the inner bark, by the vessels of which it seemed to be conveyed to the extremities of the roots. (Phil. Trans. 1801.) The motion, therefore, of the sap in the bark is not that of ascent, as Grew and Malpighi and many others have believed, but of descent, as the observations and experiments of M. de Sarrabat, Du Hamel, Knight, and others, abundantly prove. To these vessels of the bark Grew assigned particular names, according to the apparent quality of the fluid they conveyed. Malpighi gave them the general appellation of vasa peculiaria, from their containing a fluid different from the common sap. By others they have been called cortical vessels, a term, however, not applicable to many tribes of vegetables which are entirely destitute of bark. Lastly, Du Hamel and others denominated them proper vessels, which term differs but slightly from the appellation of Malpighi, and, though not very precise, is that we shall continue to employ.

Description of the proper vessels;

From regarding the newly formed vessels of the wood as part of the bark, Grew uniformly represents the bark as possessing two distinct species of vessels, in which he has been since followed by several other writers; but the vessels which he calls lymphoducts belong rather to the wood, so that we may regard him as describing only one species of proper vessels. In herbs, these vessels stand sometimes in distinct parcels or columns; sometimes they are disposed in a ring; sometimes they have a radiated position; and sometimes they are more intermixed with the sap-vessels, and seem to alternate with them. In trees, the vessels of the bark are more distinct, and have a much more regular appearance. They are commonly postured near the inner margin of the bark, and, when viewed in a longitudinal direction, seem collected into fasciculi, which are more or less numerous, and the component vessels of which continually diverge and join with others, so as to form a reticulated appearance, as in fig. 24, Plate XXXVIII. Of these reticulated fasciculi, many layers exist in an old tree; and to these layers the thickness of the bark is chiefly owing. As they proceed inward, the direction of the fasciculi is less oblique, so that near to the wood they are almost straight. Hence the spaces formed by the reticulations are very unequal, often large in the exterior part of the bark, and diminishing in size towards the wood: they are everywhere filled with cellular tissue. Such is a general description of the situation of the proper vessels of the bark, as given by Grew, Malpighi, and Du Hamel.

their structure; The vessels which thus form the vascular portion of the bark appear to differ but little in structure from the more simple vessels of the wood. Grew considered them to possess a similar structure, from believing them to be formed by the inner bark at the same time with the vessels of the wood; and therefore, he adds, "they may be reasonably thought similar in the bark and wood." (Anat. of Plants, p. 112.) Malpighi regarded them as simple tubes, containing sometimes peculiar juices; but advances nothing particular respecting their structure. (Anat. Plantar, p. 3.) M. Mirbel's opinion deserves some notice, inasmuch as he declares the structure of these vessels to differ entirely from all those of the wood. Their sides, says he, are perfectly entire; they have neither pores nor clefts, and may therefore be deemed simple tubes. (Exposit. de la Théorie de l'Organisat. Végét. p. 109.) A single vessel of this kind is represented in Plate XXXVIII,

fig. 22; and in fig. 23 a fasciculus of the same vessels is magnified as delineated by M. Mirbel. A remark also of Hill, if it be deemed to rest on correct observation, is entitled to great attention. "The vessels of the bark that form the fasciculi are not," says he, "united to each other, but are connected with the cellular tissue at numerous places; and, when separated from it, there appear on the sides of the vessels small oval depressions, dotted as it were with pin-holes." (On the Construction of Timber, p. 28.) These appearances he regards as of a glandular nature, but their description corresponds better with that of Leeuwenhoek respecting the lateral ramifications from the vessels of the wood; and they may probably be the points at which communication is effected between the vessels and cells.

The cause of the reticulated appearance which the vessels of the bark exhibit in trees, is doubtless to be attributed, as we think Du Hamel somewhere remarks, to their peculiar mode of growth. A new layer of cortical vessels is every year added to the inner surface of the bark, as well as a new layer of ligneous vessels to the outer surface of the wood; so that, as Grew observes, the new matter of the tree is every year distributed two contrary ways,—one part falling outward towards the bark, and the other part retaining its situation inward to constitute the wood. At first the newly formed cortical vessels are straight, and stand parallel, like those of the wood; but, by the continual growth of the new parts formed between the bark and the wood, the older vessels of the bark are gradually forced outward; and being thus every year disposed around a large cylinder, are necessarily more and more separated from each other, and produce at length that net-like form which we observe them to possess. The newly formed vessels of the wood, on the other hand, retain their original position; and therefore preserve their parallelism, seldom or never exhibiting those flexures and reticulations so common to the vessels of the bark.

In the bark, as well as in the wood, the vessels are found to possess different sizes. In the pine, vessels containing turpentine are represented by Grew, which are very much larger than the common sap-vessels, and are surrounded by smaller ones, exactly as the large spiral tubes of the wood are said to be ensheathed by the common sap-vessels. The milky juice of a species of sumach is contained in very large vessels, disposed so as to form a ring, and each large vessel is surrounded by many smaller ones. (Anat. of Plants, tab. 20.) The appearance of these large vessels in one species of pine is well represented by Hill, and the account he gives of their formation is probably correct. In this tree (pinus orientalis), some of these vessels form oval openings, large enough to admit a straw; these openings occupy the centre of the bark, and are surrounded by a ring of smaller vessels. As their contents are soluble in alcohol, it is easy to obtain them empty. In fig. 26, Plate XXXVIII., the vessels of this tree, as they appear in the bark, are displayed; the woody portion of the tree having been scooped away, so that the longitudinal aspect, as well as the transverse sections of them, is exhibited. From a strict inquiry into their nature, Dr Hill concluded that these larger vessels were originally the same as those of the smaller fasciculi in the bark of the same tree; "so that if we conceive one of these smaller fasciculi opened in its centre, and the vessels driven every way outward, till they are stopped by the substance of the bark, we shall have an idea of the structure of this large vessel, which is nothing more than a great cylindrical hollow formed in the centre of such a fasciculus." (Hill on the Construction of Timber,

Elementary Organs. p. 29.) It is in trees that have copious and viscid juices that these enlarged vessels are formed; and where the juices do not concretize, it is probable that, as the vessels annually recede from the centre, they suffer a reduction in size from the continued effects of desiccation and compression to which they are exposed.

The foregoing varieties appear alone entitled to the appellation of the proper vessels of the bark. In many tribes of vegetables, however, as will afterwards be shown, no distinction of bark and wood exists; but one uniform distribution of vessels extends from the centre to the circumference of the plant. Such plants have also their proper vessels, but the place and disposition of these vessels are not so precisely ascertained. From the mode in which their growth is accomplished, as well as from observation of their structure, it may be inferred that their proper vessels are distributed through the whole substance of the plant, accompanying in every part the sap-vessels. In some plants possessing this arrangement of parts, such as different species of wheat, Malpighi (Anat. Plant. tab. 4, fig. 15) describes and delineates a vasa proprium as forming a part of each fasciculus of vessels; and a similar intermixture of the two kinds must, we conceive, exist in all similar structures.

Even in many plants possessing a distinct bark, vessels containing proper juices are found in the wood. In every circle of wood, from the inmost that surrounds the pith, to the outmost in contact with the bark, vessels containing a gum, turpentine, or some other concrete or coloured juice, may frequently be found. Malpighi conceived them to exist in all plants, though, from the nature of their fluids, they could not always be distinguished; and he believed them to afford a highly perfect juice for the nutrition of the plant. According to Grew, "the turpentine vessels that are scattered up and down the wood of the pine and fir are the self-same which did once appertain to the bark; but being pinched up by the wood, they are become much smaller pipes." (Anat. of Plants, p. 115.) Du Hamel also regarded them as similar to those of the bark, but rendered much smaller by compression. In the pine and fir they are disposed circularly around the axis, much like the sap-vessels, and alternate with them. (Phys. des Arbres, tome i. p. 41.)

In Piscidia erythrina, the proper juices are of a scarlet colour, and the vessels that contain them are therefore readily discerned wherever they exist. This plant has been selected by Hill to demonstrate the position of these vessels. In the bark they are collected into fasciculi, all the vessels of which contain coloured juices, and are disposed in a ring on the inner margin of the bark. Within this ring stands the albuminum, through the substance of which many smaller red vessels are distributed; and similar red vessels are more sparingly seen in every layer of wood, particularly in that which envelopes the pith. (Construction of Timber.)

Now the red vessels thus observed in the wood of the above-mentioned plant must either have been formed in the situations they occupy, or transported from some other place. The latter supposition is inadmissible, inasmuch as the wood of trees is formed by layers of new vessels superimposed on one another; and no removal of the old vessels, nor reproduction of new vessels within the old layers, ever takes place; consequently no actual transposition of vessels could occur, nor could new vessels be developed in the wood after it had been completely formed. If, however, an alternate deposition and absorption of matter go on into and from the cells, it is possible that the vessels might in this way become filled with a matter different from that which they originally possessed;

but in the case before us, a readier explanation presents itself. The new matter of the wood is formed at the same time, and in the same place, as that of the bark; and through this new woody matter the red vessels were dispersed as well as in the bark. Consequently every addition made to the ligneous layers would furnish some vessels that contained these proper juices; and this, being annually repeated, would exhibit that intermixture of proper vessels with sap-vessels which is observed in all the ligneous layers of this and many other trees. Hence these proper vessels of the wood must be held to retain the position in which they were originally produced, and cannot be said to have approached nearer to the centre, but only, by the addition of new layers exterior to them, to be placed farther from the circumference of the tree. As Grew, therefore, held the albuminum to be a part of the bark, he might correctly say that these vessels "did once appertain to it."

ART. VII.—Of Collections of the Proper Juices in the Cellular Tissue.

Besides this accumulation of the proper juices in certain vessels of the wood, it frequently happens that deposits of similar matter occur in all parts of the cellular tissue of plants. In the bark of the oak and poplar, and of other trees, resinous concretions are often found in the cells; they are situated irregularly, and, according to Malpighi, are observed even in very young bark. Sometimes the cells containing milky and resinous juices are so postured in the bark, says Grew, as to form cylindrical channels, which are neither parallel nor anywhere inosculated, but run, with some little obliquities, distinct from one another. They appear to be formed out of the cells, and are not bounded by any walls or sides proper to themselves, but only by those of the cells. (Anat. of Plants, p. 112.) They are often short and tortuous, always isolated; and are sometimes placed irregularly, at others disposed in a circular form. They occur sometimes in the pith, and possess very different sizes and figures. They have frequently been deemed a species of proper vessels, from the mere circumstance of their containing similar juices, and from possessing sometimes an elongated form; but they are organs which, neither in form nor in function, bear any resemblance to vessels. Mirbel proposes to name them secretory canals, and M. Link cellular reservoirs—the term certainly most generally applicable, and involving no hypothesis respecting either their formation or functions.

The manner in which these cellular reservoirs may be produced in the bark or pith is readily explained, on the supposition that a communication everywhere subsists between the vessels and the cells. One set of vessels has been shown to receive and carry sap to the leaves, and another to bring it back from them to the bark; and these two sets of vessels are everywhere in their course surrounded by cellular tissue. Hence the cells in every part may receive a portion of the fluid which the vessels are employed to convey. Thus in herbs, the cells both of the bark and pith are filled with fluid, which led Grew to believe that the sap was actually transmitted through those organs; but at the same time he delivers facts which, even in his own opinion, prove that it is derived directly from the vessels. Not only in herbs, but "in every annual growth, whether of a sprout from a seed, of a sucker from a root, or of a scion from a branch, the pith is always found the first year full of sap; but in the second year the same individual pith always becomes dry, and so it continues ever after. One cause whereof is, that the

Elementary lymphæducts of the bark being the first year adjacent to the pith, they do all that time transfuse part of their sap into it, and so keep it always succulent. But the same lymphæducts the following year are turned into wood, and the vessels which are then generated and carry the sap stand beyond them in the bark; so that the sap, being now more remote from the pith, and intercepted by the new wood, cannot be transfused with that sufficient force and plenty as before into the pith, which, therefore, from the first year always continues dry." (Anat. of Plants, p. 124.)

All that is here said respecting the transfusion of the common sap from the vessels into the cells as it ascends through the wood, is equally applicable to the proper juices as they descend through the bark. During the first year of growth, both the sap and proper vessels are adjacent to the pith as well as to the bark, and each order may therefore transfuse its fluids into the cells of either. The common sap, from retaining its fluidity, is frequently removed by absorption, and the cells that contained it appear empty and dry; but where the proper juices are transfused, and become viscid or concrete, they are retained, and appear in different quantities and forms both in the bark and pith, according to the nature and properties of the fluids from which they are derived, and the texture and situation of the tissue into which they are poured.

In inter-cellular canals. To these collections of the proper juices in the cells of the cellular tissue may be referred the opinions of those who describe them as existing sometimes in vacuities between the cells. Grew had remarked the occasional presence of these juices in cylindrical channels formed by these vacuities, to which M. Treviranus has given the name of intercellular canals. According to him, these canals not only contain the proper juices, but convey them to all the cellular parts of plants, and are the true proper vessels of the bark. We before saw that M. Kieser, after making the wood to consist almost entirely of spiral vessels, destined only to convey air, supposed the sap to be carried by these same intercellular canals. (Kieser, Mém. sur l'Organisat. des Plantes, p. 36.) With M. Treviranus, he also considers them as the real proper vessels of the bark, by which alone the juices are conveyed. (Ibid. p. 86.)

Opinion of De Candolle. M. de Candolle considers the proper juices to possess no special organ for their conveyance, but to nestle in the cavities that surround them, and form sacs or reservoirs of various size and form. He distinguishes several kinds of these reservoirs, which vary with the figure of the containing sac. Thus, what others call "proper vessels," he designates by the names of tubular and fascicular reservoirs. In this, however, he appears to confound the "proper juices," destined immediately to the functions of nutrition and growth, with various fluid or solid matters, either simply deposited in the cellular tissue, or that have undergone subsequent alteration by the agency of a secreting organ. The reservoirs of the proper juices, and the juices themselves, are, he says, readily seen in many families of dicotyledons; but have not been observed with certainty, either in monocotyledonous or acotyledonous tribes. (Organog. Vég. tome i. chap. 11.) The reason of this seems to be, that in dicotyledons the bark, in which the proper juices are seen, is distinct from the wood, in which the common sap flows; and the colour, consistence, and qualities of the two fluids, therefore, at once point out their respective places; but in plants belonging to the two latter tribes there is no proper distinction of bark and wood; and the two orders of vessels, as shown by Malpighi in the instance of wheat, are associated together through the

whole substance of the plant, which renders the discrimination of their common and proper juices more difficult.

Besides two orders of vessels existing in the ligneous part of trees, M. Dutrochet describes a particular form of cell, which, in the wood, serves only as a receptacle for fluids; but to which, in the bark, he assigns a vascular function. This is the elongated cell observed by Malpighi, and represented by him in fig. 18, Plate XXXIX. M. Dutrochet, from the resemblance which its figure is said to bear to a spindle, has given it the name of clostre; and his representation of it is given in fig. 17, Plate XXXIX. These elongated cells form what Kieser and others have called the ligneous fibres. By M. Dutrochet they are regarded as reservoirs of the nutrient juices, and as containing secretions which impart solidity to the wood. In the bark, however, these same clostres are declared to be the organs in which the elaborated sap is contained, and by which it is transmitted to the roots to furnish the materials of their growth. The clostres, which thus convey the nutrient matter, ought not, he adds, to be confounded with the "proper vessels," which are the true secretory organs. These vessels are tubes of greater diameter than the clostres, and contain various excrementitious substances, as resin, &c. in different plants. (Recherches Anatomiques, &c. p. 36, 37.) According to M. Dutrochet, therefore, the elongated cells and vessels which exist in the bark and wood exercise functions precisely opposite; for whilst in the wood the corpusculiferous vessels convey the sap, and the cells are reservoirs for containing it, in the bark the cells are the organs which transmit the elaborated sap, and the vessels those in which the excrementitious matters are contained.

In our view, these notions are the reverse of fact, since we consider the vessels, both in the wood and the bark, as the organs by which the fluids are conveyed, and the various forms of the cellular tissue as the general receptacles into which they are in part deposited: they may, however, in some instances be retained in the vessels themselves, as well as in the cells.

SECTION II.

Of the Absorbent and Exhalent Systems.

ART. I.—Of the Absorbent System.

Connected with the vessels that distribute the fluids through organized bodies, is another system of vessels, by which extraneous matter is taken up to support the growth of parts, and supply the waste occasioned by the exercise of the various functions. To these vessels anatomists have given the name of Absorbents. The function which they perform is carried on either from the external surface, or from some internal part of the body; and its exercise, in animal bodies, may be distinguished into three kinds or stages. The first is that in which new or extraneous matter is taken up and added to the system, as in the absorption of substances from the skin, or of chyle from the intestines; the second is that in which substances previously separated from the fluids by secretion, but without becoming organized, are again taken up by the absorbents, and reconveyed into the blood-vessels, as in the absorption of milk, bile, and fat; the third kind is that in which the secreting organs themselves, and successively all the solid parts of the body, are removed by the action of the absorbent vessels. By plants this function is exercised to less extent, and seems to comprise only the first two kinds or degrees of it, viz. the primary absorption of extraneous matter, and the re-absorption of certain secreted matters; but in them there

does not appear any power of re-absorbing what has once formed an organized part of the system. Hence the formation of organized parts in plants is not accompanied, as in animals, by the unceasing removal of old particles; but the particles which have once become organized continue permanent until removed by some cause or process foreign to the living powers of the plant. When once, therefore, the organs of plants have become mature, they are exposed to decay only from the operation of foreign causes, and cannot be mined from within by that gradual loss of balance between the secreting and absorbing functions which the advances of age bring on the animal system. Neither, in the vegetable system, can the removal of organized parts under disease, any more than in health, have place; and, consequently, the several modes or stages of ulcerative absorption, so finely illustrated by the researches of Mr Hunter, belong not to the economy of plants.

As the function of absorption is thus of more limited extent in plants than in animals, so may we expect to find the arrangement of organs destined to its exercise. In the more perfect animals, the absorbent vessels are quite distinct from those which carry blood, are provided everywhere with glands, and, like the veins, are furnished with valves. From their beginnings, in all parts of the body, to their termination, they continually unite with one another, until, after a long course, they form at length two trunks, which deliver their contents into the large veins near the heart. As no common reservoir exists in plants, there is no such single point to which their absorbed fluids require to be carried. Hence their absorbents seem everywhere to have a very short course, to form no union with one another, but to deliver their contents at once into the sap-vessels adjacent to them. They appear to be destitute alike of glands and valves, and indeed, in an anatomical view, they can scarcely be considered distinct from the sap-vessels, but may rather be deemed ramifications from them; so that although we grant, with Grew, that the ordinary sap-vessels do not ramify one out of another, yet they certainly send off those fine ramifications which, from their office, we denominate the absorbents of plants.

In the root, the absorbents are capable of being demonstrated. When a plant is immersed in coloured fluids, many of its capillary absorbents become tinged through their whole course to their termination in the sap-vessels, proving them to be simple ramifications from these vessels themselves. A further proof of the identity of these two systems is derived from the fact, that any part of a sap-vessel is not only capable of emitting capillary absorbents from its sides, but of exercising itself an absorbent function, whenever its cut extremity is brought into contact with a fluid. These absorbents are formed very speedily and in great multitudes on the roots of annual plants; and even in perennial plants they appear, like the leaves, to suffer an annual decay, and to be reproduced with the return of vegetation.

M. de Candolle describes a structure as existing at the extremities of the roots and of the pistil, and also on the outer coat of seeds, which possesses a peculiar aptitude for absorption. To this structure he gives the name of spongiolæ: it is composed of small round cells, and in roots is situated at all their fibrous extremities. The entire body of each radical fibril is indeed made up of an analogous cellular tissue; but it is the extremity alone which actually absorbs, and which therefore possesses a peculiar absorbent or hygroscopic power. The pistillary spongiolæ are in like manner situated at the extremity of that organ, and absorb the fecundating liquor. The third

variety, or seminal spongiolæ (spongiolæ seminales), are placed on the surface of the grain, apparently with great regularity; and by them the absorption of moisture is effected. Not only water, but coloured liquors, are absorbed by these spongiolæ, though the molecules of such liquids never enter the ordinary pores of plants, which pores must be infinitely greater than the apertures with which the spongiolæ may be furnished. This is the more remarkable, says M. de Candolle, when we call to mind that these colouring molecules traverse the compact and almost stony coats of seeds, and yet do not enter leaves whose tissue is so lax, and which are visibly furnished with pores that have the power of absorbing water when in contact with it. (Organog. Végétale, tome i. chap. 7.)

But plants absorb fluids by other parts of their surface as well as by the roots. This absorbing power extends, in some of the lower tribes, over the entire surface of the vegetable, which is destitute of any organ analogous to a root; in other instances, where the roots are small and the soil arid, the plant derives almost all its moisture by the absorption of dews through the leaves. The experiments of Bonnet show that all leaves, both those of herbs and of trees, when brought into contact with water, are capable of absorbing it, and that the moisture thus absorbed is communicated through the vascular system of the leaf. The leaves of herbs he found to absorb nearly alike from either surface; but those of trees absorbed best by the lower surface. The petiole and larger riblets appeared to absorb much less than the other parts of the leaf. So great and general is this absorbing power, that vegetables, says he, may be said to be planted in the air, nearly as they are in the earth, the leaves being to the branches what the capillary rootlets are to the roots. (Recherches sur l'Usage des Feuilles, p. 22, 47.)

What then are the organs by which this function is carried on in the leaf? M. Bonnet imagined the vessels of the leaves to receive their fluids through the pores adjacent to them; and that the leaves, which had only few pores, possessed but little absorbent power. (Ibid. p. 20, 22.) He thought, also, that the hairs frequently distributed over the leaf attracted moisture, and might even act as absorbents; though he admits that many leaves which have only slight inequalities on their surface, without hairs, exercise an absorbent function. (Ibid. p. 47.)

This subject has been since investigated by M. de Candolle, whose researches appear to confirm the account of Bonnet as to the absorbing power of the pores. According to him, these pores are found on all parts of the leaf except the riblets, which have none, but are covered with hairs. At the mouth of the pore a vascular net-work is always to be found, which he regards as a production from the larger vessels of the leaf. He asserts that pores are found only in those parts where vessels go to terminate, and not in others; and that in trees this structure occurs chiefly on the lower surface of the leaf, while in herbs it is equally seen on both surfaces. The stem in general has few or no pores, except where it is soft and herbaceous; and even then the pores occur only in the deeper green furrows, not on the prominent lines which bound them, and are usually covered with hairs. No pores are to be found on roots or bulbs or fleshy fruits, but most of the organs above ground are more or less furnished with them. Exposure to the air seems necessary to their formation; for plants, or parts of plants, that live beneath water or earth, are destitute of pores, but acquire them if brought into the free air. Exclusion of light prevents also the formation of pores; and hence etiolated plants are not furnished with them. (Mém. de l'Institut. Nat. tome i. p. 351.) In general, his anatomical researches respecting the ex-

Elementary istence of pores in leaves agreed with the results of Bonnet's experiments on the absorbing power of their surfaces; and when we consider that all plants and parts of plants secluded from the air are at the same time destitute of pores, and of the power of absorbing by their surfaces, it may be inferred that the organs of absorption on the external surfaces of plants are the minute vascular productions which, in tender and succulent parts exposed to the air and light, everywhere perforate the cuticle, and form in it those innumerable orifices which we denominate pores.

But the function of absorption in plants is not confined to the taking up of extraneous matters. Many facts prove that, in every part where active vegetation exists, internal absorption is continually going on. The organs by which this absorption is immediately performed cannot, perhaps, from their extreme delicacy and minuteness, be rendered capable of anatomical demonstration; but certain facts which occur in plants, coupled with certain analogies derived from other organized textures, must, we think, carry complete conviction of their existence. In almost every part, and on every surface of animal bodies, the vessels which exercise absorption may be traced; but the mouths or orifices by which they actually absorb are scarcely ever to be seen. In one instance only, viz. in the intestines, have they been followed to their beginnings, and discovered in the act of exercising their appropriate function. (Gordon's System of Human Anat. vol. i. p. 70.) But though their orifices remain in other parts undiscovered, no anatomist hesitates to admit their existence when he sees the canals of the vessels themselves laden with blood, or milk, or bile; and seeing them thus to convey fluids whose colour manifests their presence, he equally believes them capable of absorbing and conveying other fluids, though they may not be visible to the eye. Believing further, that no solid part of the body, nor even fluid part that has been deposited in close cavities, can be removed in a natural manner from its place, but by the agency of these vessels, he comes to regard the simple fact of the disappearance of such solid or fluid part as sufficient evidence of its absorption.

Now the facts and analogies on which internal absorption rests in the vegetable system are precisely of the same nature, and the evidence of its existence is scarcely less complete. In every part of the cellular tissue of plants various substances have been found which must have been primarily derived from the vessels, the only organs which furnish new materials to the plant. These substances, however, often disappear from the cells, and are again to be detected in the vascular system. Thus, in the seed, as will afterwards be shown, the cells of the cotyledon contain a solid unorganized matter, which could have been originally deposited in them only by means of the vessels. During germination, this solid matter is rendered fluid, disappears from the cells, and is again to be traced in the vessels on its way to afford nutriment to the radicle and plume. We say, then, that this unorganized matter must have been taken up from the cells of the cotyledon, and conveyed into the vascular system, by the agency of absorbent vessels, which, it is probable, are distributed everywhere on the inner surface of these cells; just as, in the animal system, absorbent vessels are considered to take up the fat from the inner surface of the cells in which it is deposited, and convey it into the vessels of the animal.

This alternate absorption and deposition of the nutrient matter of seeds is sometimes strikingly displayed in the growth of potatoes. It frequently happens that potatoes lying in a damp cellar put forth shoots which grow to a considerable size, without the access of any

foreign agents except heat, water, and air. On these shoots young bulbs, as large as the eggs of pigeons, are sometimes to be found, and the substance of the old bulb has in great part disappeared. In such cases, the matter from the cells of the old bulb must be considered as removed by absorption, and conveyed into the vessels of the shoot, where it was in part employed in forming the new organs of the young bulb, and in part deposited, to experience, perhaps, in some future growth, similar successions of removal and deposition. In the living parts of perennial plants, also, nutrient matter appears to be alternately deposited and absorbed from the cells during the active periods of vegetation; and in the cellular tissue of herbaceous plants a similar deposition and absorption of fluids seems to be frequently taking place; so that, in all the vegetating parts of plants, these alternate functions of secretion and absorption are more or less constantly exercised.

A good illustration of the manner in which these functions are alternately exercised is afforded by an experiment of M. de Candolle. The parasitic plant called misletoe draws its nourishment, as is well known, from the tree on which it grows. M. de Candolle placed a branch of an apple-tree, bearing a stalk of misletoe, in an infusion of cochineal for five days. He then dissected it, and observed the coloured liquor to have risen through the wood and albuminum of the apple-branch, and reached the place of junction between it and the misletoe, which it strongly reddened; and from thence it penetrated into the woody part of the misletoe. There did not, however, appear to be a true anastomosis between the vessels of these different plants; but, at the base of the misletoe, where the parts were so deeply reddened, a minute cellular structure was observed. Into these cells the vascular system of the apple appeared to deposit its sap, and from them the capillary absorbents of the misletoe, distributed upon the cells, seemed, like the ordinary absorbents of roots, to take it up. (Mém. de l'Institut. Nat. tome i. p. 370.) From these and many similar facts it may be inferred, that absorbents, communicating with the vessels of the plant, exist in every part, and that the removal of all secreted matters from the cells and other close cavities of the vegetable, when effected by the living powers of the plant, is accomplished, as in animal bodies, by the exercise of an absorbent function.

ART. II.—Of the Exhalent System.

But from their external surface, and from the same organs as we have seen to exercise an absorbent function, plants, in certain circumstances, give off a large quantity of fluid by transpiration; and the organs by which this function is performed seem, from many considerations, to be the same as those by which, under other circumstances, absorption is accomplished. This function of transpiration is common to all terrestrial plants, and, with few exceptions, all are more or less furnished with pores; but it does not occur in aquatic plants, according to De Candolle, which are destitute of pores. Fleshy leaves and fruits, and the petals of flowers, which have but few pores, transpire little; and etiolated plants, which are destitute of pores, do not transpire at all. On the contrary, herbaceous stems and plants, which have numerous pores, throw off most fluid by transpiration. This general agreement between the existence of pores and the exercise of the transpiratory function leads to the presumption that they are the orifices through which the fluids are discharged; and if it be admitted that these pores are situated at the extremity of the fine ramifications that come

elementary off from the vessels, their fitness for such an office cannot be denied. Comparing these facts regarding transpiration with those previously stated concerning absorption, M. de Candolle is led to conclude that the pores on the surfaces of plants are the organs by which these functions are alternately carried on, according to the existing condition of dryness or humidity in the surrounding atmosphere.

Repulsive as this conclusion may at first seem to our ordinary conceptions of organized bodies, yet there are many circumstances in the structure and habits of plants that give to it great probability; so much so, that we ourselves had long since reached the same point by a route different from that pursued by M. de Candolle. It is highly probable that the exhalents of the leaves are simple ramifications from the larger vessels, like the capillary rootlets; and as they have no valves in their canal, there is no mechanical impediment to their exercising an inverted action. The sap-vessels themselves readily absorb even coloured fluids when inverted; and though their exhalent terminations are too fine to receive such fluids, yet why may they not, like the trunks from which they spring, be capable of taking up ordinary fluids in that manner? The fluids absorbed through the leaves must at once enter the sap-vessels, for there is no common reservoir to which they can first be carried; and it is extremely improbable that, from the same parts of the same vessels, exhalents and absorbents, capable of exercising only opposite functions, should at the same time arise. In the animal system the exhalents spring from arteries, and the absorbents terminate in veins; but in the less complex structure of plants it seems demonstrable that both orders of vessels must at once communicate with the same sap-vessels. It is therefore more probable to suppose that, instead of two distinct orders of vessels, as in animals, one only should be provided, capable, under different circumstances, of exercising different functions. This vicarious office of the organs under consideration leads to no confusion in its exercise; for the condition of the atmosphere, which favours transpiration, is that which removes from the leaves the power of absorption; and, on the contrary, absorption occurs only in a humid atmosphere, when, as Hales has shown, little or no transpiration takes place.

Besides exhalation carried on through visible pores, plants, like other bodies, lose their humidity by simple evaporation. This evaporation seems to go on in a moderate degree under favourable circumstances by day and by night, and is quite distinct from the copious transpiration that is carried on by the agency of solar light through the pores. That the pores are exhalent organs, was an opinion first advanced by Hedwig, and subsequently illustrated by De Candolle, Sprengel, Link, and Rudolphi. (Organog. Végétale, tome i. p. 86.) In like manner, though absorption may be carried on in certain circumstances through the pores of leaves, yet there are plants which largely imbibe water, but in which few or no pores have been detected. Thus, fleshy leaves which have lost their moisture quickly regain it either when plunged in water or exposed to rain, though they possess but few pores; and the aquatic algae, says M. de Candolle, evidently imbibe water over their whole surface, though they are wholly destitute of pores. He supposes that, beside the ordinary pores which may be readily seen, plants may be furnished with others that are invisible from their extreme minuteness. By such pores we might account for the loss of weight which plants destitute of visible pores gradually undergo when exposed to the free air, and also for the increase of weight which mosses, and

other plants equally destitute of pores, rapidly acquire when plunged in water. (Organog. Végét. tome i. p. 88.)

The view thus presented of the external absorbent and exhalent vessels may probably be extended to the minute vascular productions which seem everywhere to spring from the vessels internally. If secreting and absorbing vessels be held to exist in every part of the plant, they must everywhere communicate with the vascular system; for it is from the vessels of this system that the matter of their secretions is primarily derived, and it is into the same vessels that, in many cases, these secretions are subsequently returned. Nor does the exercise of the two functions of secretion and absorption in plants present any apparent obstacle to the supposition of their being performed at different times by the same organs. Thus, when nutrient matter is deposited in the cellular tissue of the seed, it is destined only for a future use; and the purpose of nature would be defeated were an absorbent function to be at the same time employed for its removal. On the other hand, when this matter is again taken up during the germination of the seed, no secreting function seems then to be exercised in that part; for the organ itself, in many seeds, gradually wastes, and no fresh matter is deposited in it. Even when the cotyledon to a certain extent augments in size, its nutrient matter is continually drawn off for the support of the radicle and plumule, and no fresh matter of the same kind seems to be then deposited; so that the same vessels which formerly exercised the function of secretion may, without disturbing the economy of the plant, be now employed in the exercise of absorption. As thus the two functions do not require to be performed in the same part at the same time, they may, if nothing else forbid, be exercised at different times by the same organs. In the animal system, where the organs themselves are removed, secreting and absorbing vessels must necessarily co-exist; and to maintain the integrity of parts, their functions must proceed at the same time, and to a certain extent balance each other; but as no similar operations appear to be carried on in the vegetable system, no such complex organization is required to sustain them.

SECTION III.

Of the Cellular Tissue.

ART. I.—Description and Structure of the Cells.

The elementary organ denominated cellular tissue may be said to consist of a membranous substance, disposed into a great number of small circumscribed cavities, connected with each other, and arranged in rows or suites, generally in a direction opposed to that of the tubes, which represent the vascular system. From Grew it received the appellation of parenchyma, a term still often used in describing different parts of this tissue; by Malpighi, it was called the utricular substance; and it owes, we believe, its present name to M. du Hamel.

The cavities which distinguish its construction were called indifferently bags or bladders, pores and cells, by Grew; by Malpighi, utricles; by others, vesicles; and more commonly cells. The form of these cells varies so much, not only in different plants, but in different parts of the same plant, as to have authorized, in some degree, these different appellations. The tissue which they constitute enters into the composition of every organ in the more perfect plants. Of many herbaceous plants it forms the chief portion, and some of the lower tribes of vegetables, as the algae, lichens, mushrooms, &c. are said to be wholly composed of it; in other words, no vessels can be

Elementary actually demonstrated in them. In most cases it contributes greatly to modify the form of organs, and adds always to their bulk and strength. Nothing can exceed the diversity of appearance in figure, bulk, and texture, which it exhibits in the several parts, circumstances, and conditions in which it is placed. It represents sometimes a lax cellular substance, all the parts of which are succulent and transparent; in other instances it is compressed into a solid, opaque body, retaining but faint traces of its former cellularity; and in others, again, it is spread out into a most thin and delicate membrane, in which the cellular character is wholly lost. It everywhere envelopes and holds together the vascular system, and seems to be the general receptacle of almost all the vegetable secretions.

Figure of the cells; The figure of the component cells of this tissue is exceedingly various. Sometimes they have nearly a globular or spheroidal shape; in other instances they are angular, and exhibit in their section a greater or lesser number of sides and angles, being in a few examples triangular; in others square, but more commonly hexagonal, the figure which collections of soft cells, mutually impressing each other, seem naturally disposed to assume. This form is represented in the transverse section of cells, fig. 28, Plate XXXVIII.; they seem in this figure, and in most of those given in different works, to possess double sides; but, as M. Kieser has remarked (Mém. sur l'Organisat. des Plantes, p. 91), this appearance is produced by the borders of the subjacent cells being seen through the transparent sides of the superior layer. In Plate XXXIX. fig. 16, is a representation of a transverse slice of the cellular part of sugar-cane, drawn from nature, and so thin as to exhibit only one layer of cells, in which the sides appear distinctly single; but, in a thicker slice of the same plant, fig. 21, comprehending more than one layer, the double appearance becomes very evident.

their size; The size of the cells varies not less than their figure in different plants and in different parts of the same plant. In one of the plates of Grew they are represented as possessing twenty different sizes, from that of a minute pore, to the size of a common pea. Hooke examined them in cork, and in the pith of many plants. In cork he reckoned several lines of these cells or pores, as he calls them, and found there were about sixty placed endwise in one eighteenth part of an inch, or somewhat more than a thousand in the length of an inch; and, therefore, in a square inch above a million, and in a cubic inch above 1200 millions. (Micrographia, p. 114.) In this substance the cells are not visible by the naked eye, but become very distinct when highly magnified. In most plants, however, they are readily visible, and their appearance is familiar to every one.

When viewed in a longitudinal section, their hexagonal form is much less distinct, and is sometimes wholly lost. In fig. 22, Plate XXXIX., is a series of single columns of the cells of sugar-cane, in which each cell is, to appearance, bounded only by four sides. Similar representations are given by Hooke of the cells in cork, and by Kieser in most of the figures which exhibit longitudinal sections of the cellular tissue; but in some instances the hexagonal form is visible even in these sections. In fig. 23 of the same plate we have given, in outline, the appearance of two series of columns of these transparent cells, in which one series is seen behind the other, and gives somewhat of the confused double appearance exhibited in the transverse section, fig. 21.

contents; The nature of the matter contained in the cells of this tissue varies according to the part in which it exists, and the peculiar powers of the plant. Both Hooke and

Grew remarked, that in the pith and bark of succulent plants the cells were often filled with aqueous juices, and in the same plant, at other periods, they appeared empty, or filled only with air. In the seed, the cells of the cotyledons contain minute unorganized particles, which, at a future period, serve as nutriment for the young plantule. Other particles of still smaller size, of a resinous nature and a green colour, exist in other parts of this tissue, and bestow on the plant its verdure. In every part of the plant these cells are also the occasional receptacles of the peculiar fluids which both the sap-vessels and the proper vessels convey; and hence various gummy and resinous substances, corresponding in quality to the fluids previously existing in the vascular system, are frequently detected in them. In the pulp of fruits, the various acid, saccharine, and austere juices that we meet with, are contained in different modifications of this tissue; and it is into its cells that the osseous secretions, which constitute their shells and stones, are made. These facts prove not only the great importance of this tissue in the construction of the vegetable organs, but the active share it bears in the economy of their functions; and demonstrate likewise an universal communication between the vessels and the cells.

The sides of these cells, when emptied of their contents, and viewed through the microscope, appear to be formed by a very fine transparent membrane, which some maintain to be everywhere entire, and others to be perforated with pores. The same sources of error exist here as before noticed in similar microscopical observations on the vascular system; and, accordingly, the respective disputants maintain, with equal confidence, the same opinions with regard to the porosity or non-porosity of the cells, as they had previously held concerning the vessels. We must therefore call in the aid of other means besides those of the microscope, for determining the important fact whether the cells have or have not any direct communication with each other.

Dr Hooke examined the films or sides of the cells of cork, of the pith of elder, and of many other plants, for the very purpose of discovering whether any direct communication existed between them; but "each cavern or cell," says he, "is distinctly separate from the rest, without any kind of hole in the encompassing films;" nor could he, with his microscope, nor by his breath, nor by any other way that he tried, "discover a passage out of one of those cavities into another." (Micrographia, p. 116.) Dr Grew describes the little cells or bladders that compose the bark of roots as possessing a spheroidal shape in most plants. When viewed with the microscope, their sides are as transparent as water; and "none of them," he adds, "are visibly pervious from one into another, but each is bounded within itself." (Anat. of Plants, p. 61.) Both Hooke and Grew, however, believed a communication to exist between the cells, from the fact of their containing liquor; and Malpighi held the same opinion from similar considerations; but they nowhere describe the mode or structure by which they conceived it to be accomplished.

Later writers have not only adopted this opinion, but professed to demonstrate the structure by which the communication is maintained. M. Mirbel describes the sides of the cells as composed of an extremely thin, colourless, and transparent membrane, which is commonly perforated with pores, the diameter of whose aperture is not perhaps the 300th part of a millimetre. These pores are ranged generally in transverse series, and through them, it is said, the cellular tissue both receives fluids from the vessels, and transmits them very slowly

through its cells. (Exposition de la Théorie, &c. p. 105.) M. Sprengel and some others adopt this view of the porosity of the cells; but it is denied by Link, Treviranus, and Kieser. The latter author declares, that notwithstanding all that has been said concerning the pores in the sides of the cells, his observations, made with the greatest care and exactness, have not enabled him to discover the slightest trace of them. The sides of the cells, he adds, are always formed by a membrane extremely thin, but altogether smooth and uniform; and the cells themselves have never an open communication with each other. (Kieser sur l'Organisat. des Plantes, p. 94.)

If, then, no pores exist in the sides of the cells for the reception and transmission of the fluids they contain, some other means must be provided for the accomplishment of these objects. M. Link, accordingly, supposes the juices to pass from one cell to another by transudation. M. Rudolphi thinks that a decomposition of the fluid is effected by the cells themselves, during which it is transmitted through their sides; and M. Dutrochet calls in the aid of electrical agency. To us there occur no probable means of accomplishing these operations, consistently with the integrity of the cellular structure, but the exercise of those alternate functions of secretion and absorption which, from so many other considerations, we have supposed to be carried on in every living part of the vegetable system.

Another question of importance in relation to the sides of the cells is, whether they are single or double; that is, whether each cell has a side of its own, or whether one side is in every position common to two cells. Mirbel asserts the former, and Kieser maintains the latter opinion. In fig. 29, Plate XXXVIII., is an outline representation of these double sides as given by Kieser. From the extreme thinness of the membrane, it is very difficult, he says, to distinguish this double structure; but where the cells are large, and a glass that magnifies highly is employed, each partition that separates two cells is distinctly seen to be composed of two membranes, which are sometimes separated about the middle of the partition, and united towards the angular points. (Mém. sur l'Organisat. des Plantes, p. 91.) MM. Amici, Dutrochet, De Candolle, and others, support this opinion. (See Organog. Veg. tome i. p. 21.) The existence of this double structure receives some countenance from the fact observed in the construction of the honeycomb by the late Dr Barclay, who says that each side of every cell in the comb is composed of two plates, or is double. (Wernerian Transactions, vol. ii.) It may still, however, be more properly said, that each side of every cell is truly single, and is rendered double only by coming into contact with the corresponding side of an adjacent cell.

When the cells have a regular hexagonal figure, and are equally distended with their appropriate juices, there is no reason to suppose that any vacuities are left between their sides or angles. Mathematicians have long since demonstrated a regular hexagon to be one of those figures that completely fill up a given space; and that no vacuities can exist either about its sides or its angles. Where, however, the cells deviate from this regular figure, and more or less approach to a spherical form, vacuities or interstices may readily be conceived to occur. These vacuities are said to have been first noticed by Grew and Leeuwenhoek, and afterwards by M. Treviranus, who describes them as interstices left between the cells in their mutual approximations towards each other. He gave them the name of intercellular canals, as already stated.

On the other hand, MM. Mirbel and Rudolphi alto-

gether deny their existence; but M. Kieser contends strenuously for it. He describes them as small interstices situated at the angles of the hexagonal cells, and formed not by any sides of their own, but by the mutual approach of three contiguous cells, and possessing, therefore, a prismatic form. These interstices he conceives to exist at every angle, and thus every cell to be surrounded by them. In fig. 29, Plate XXXVIII., the black angular points denote their place. By their conjunction with each other they form a canal, which, when the hexagonal figure is perfect, and the cells are ranged horizontally, extends both in a longitudinal and transverse direction; and when the cells are placed obliquely, the canals have a similar direction. Their size varies according to that of the cells, by the sides of which they are constructed; they contain and convey the proper juices in the bark, but in the pith are often dry; and their course is said to terminate only with that of the cells themselves, at the surface of the plant. (Mém. sur l'Organisat. des Plantes, p. 104.) Such are the organs which, as we have seen, M. Kieser considers to convey both the sap and the proper juices in plants. That in some circumstances they may exist, and become reservoirs of the sap or other juices, seems highly probable; but of the impropriety of ascribing to such casual productions the performance of the primary functions of the vegetable system, we have already spoken.

The cellular tissue, as described above, is that form of it which must be regarded as the most perfect. From various causes, however, it is subject to great alterations. In herbs, and in the pith and succulent parts of trees, the cells preserve their original form and appearance for a considerable time; but by the growth of the other parts, and consequent extension and compression they experience, they acquire in the bark and wood an elongated figure, and this both in a transverse and longitudinal direction. In the latter case they surround and connect the layers of vessels with each other, constituting what has been named the parenchyma of the bark and wood. In this form their size is often greatly reduced, their cavities sometimes obliterated, and their cellular character altogether effaced. In other instances traces of a cellular structure are occasionally visible, appearing in detached portions among the perpendicular vessels. In some plants the cells become elongated in a longitudinal direction, and yet preserve their capacity nearly unchanged. It was probably this form of cell that led Malpighi to regard sometimes as a vessel, what, in reality, appears to be only a series of elongated cells. To his representation of these cells, as given fig. 18, Plate XXXIX., we have before referred. M. Kieser considers these elongated cells as forming the ligneous fibres of the wood, and that the bark is almost entirely formed of them. (Mém. sur l'Organisat. des Plantes, p. 99, 101.) These cells, he says, were originally round; but in trees they become so much lengthened as to exhibit the form of a tube. It is easy, however, he adds, to detect transverse partitions in these seeming tubes, which, having often a diagonal direction, give to these cells the form of a double pyramid with sharp points. The membrane which forms these cells is, in all trees except the Conifera and some others, smooth, without pores, and transparent. (Ibid. p. 299.) To these organs Treviranus gave the name of fibrous utricles; by De Candolle they are called tubulated cells; and Dutrochet, from their spindle shape, denominates them, as we have seen, clostres. His representation of them, fig. 17, Plate XXXIX., corresponds with the foregoing description of Kieser.

From suffering compression in a transverse direction, the cells have frequently their longer diameter thrown

Elementary into that position, and thus extend from the centre to the circumference of the plant. This position, as will afterwards be shown, was fully noticed by Grew and Malpighi. Leeuwenhoek also observed it, but mistook the cells thus elongated for vessels, and considered their partitions as valves,—errors which M. Kieser, as well as others, duly points out. In this transverse direction the tissue forms partitions more or less large between the vessels, as will afterwards be shown; and by the obliteration of its cells it is frequently reduced to the condition of a solid membrane.

Ruptures of the cells. Besides these more constant and necessary changes in the figure and character of the cellular tissue, it often suffers others of a more casual and accidental nature. In the pith, as the plant grows up, divers ruptures, says Grew, occur, oftentimes very regularly, and observed constantly in the same species of plant. These ruptures are sometimes prolonged, so as to form a tube of considerable length. (Anatomy of Plants, p. 120.) Others have observed similar canals in the pith, formed not by sides of their own, but by those of the adjacent cells, and very various in size and form. They have been called lacunæ, or reservoirs, contain a variety of substances, and sometimes, especially in aquatic plants, only air. As we have seen the cavities of the larger spiral vessels to be filled with vesicles, so the larger cells of the pith, according to Grew, frequently contain smaller ones, or are divided by cross membranes. A similar observation is made by Kieser, who likewise remarks that, in the empty cells of Calla Æthiopica, he has sometimes seen small round-headed bodies, supported on little peduncles, which spring from the sides, and point towards the centre of the cells. Small crystallized bodies are also occasionally found in the cavities of the cells, and within the intercellular canals. (Mém. sur l'Organisat. des Plantes, p. 94.) Of those changes in the character of the cellular tissue, by which its cells are converted into receptacles and reservoirs of the proper juices of the plant, we before discoursed when treating of the proper vessels. To such an extent does this change sometimes proceed, that, in aged oaks, and, according to Kieser, in guaiacum, and probably in many other plants, the whole cellular tissue becomes filled with these secreted matters, and the distinctive characters of the cells, and almost of the vessels themselves, are obliterated and lost.

CHAP. II.

THE ANATOMY IN GENERAL OF THE COMMON TEXTURES OF VEGETABLES.

Preliminary Observations.

Nature of the common textures. The elementary organs, whose description has so long occupied our attention, form, either singly or by their combination, all the other parts of plants. Some of the lower tribes of vegetables consist entirely of cellular tissue, in which no vessels are at any period to be seen; and, even in the higher orders, many parts exhibit no appearance of a vascular structure. There can be little doubt, however, of the existence of such a structure, since, physiologically speaking, we can form no just conception of the growth of an organized body, without associating with it the existence of a vascular system. In all plants the pith consists of a cellular tissue alone. In herbaceous plants this tissue forms their greater portion; but in trees the number of vessels is so great as to constitute the chief bulk of the plant. To certain forms of these elementary organs, whether existing singly or in combina-

tion, we have given the name of common textures, because they are very generally to be found in all plants, and in almost all parts of them, howsoever varied in quantity, proportion, and arrangement. These textures are familiarly known under the names of cuticle or skin, of bark, of wood, and of pith; to which may be assigned the general appellations of the cuticular, the cortical, the ligneous, and medullary textures.

All the several textures just enumerated are readily distinguished by their different places and characters in the section of most arborescent plants, in which they commonly appear well defined, and perfectly distinct from each other. In many plants, however, both herbs and trees, this distinction of parts is not preserved; but, with the exception of the cuticle, all the other textures are blended together through the entire substance of the plant, as was long since noticed both by Malpighi and Grew. "In the stalk of maize or Indian wheat," says Grew, "the work of nature appears less diversified; in which, although there are the same parenchymous and ligneous parts as in all other plants, yet is there neither bark nor pith, the vessels being dispersed and mixed with the parenchyma, from the circumference to the centre of the stalk." "The like structure," he adds, "may also be seen in the sugar-cane and some other plants." (Anat. of Plants, p. 104.) Similar observations were made by Malpighi, not only on different species of wheat and sugar-cane, but on ferns and palms. "In ferns," says he, "the vascular fasciculi are numerous, but placed without order, and are everywhere sustained by the intervening cellular tissue, the cells of which are sometimes much smaller than the orifices of the vessels themselves." (Anat. Plantar. p. 24, 25.) This structure is represented in the transverse section of the sugar-cane, Plate XXXIX. fig. 25; and in a similar section of the palm, fig. 26 of the same plate.

This variety of structure, thus clearly described, and distinctly delineated in the works of Malpighi and Grew, has likewise been noticed by M. Desfontaines. Vegetables, according to him, may be distinguished into two divisions: 1st, Those which have no distinct concentric layers, whose solidity decreases from the circumference towards the centre, and whose pith is interposed among the vessels, and does not extend in divergent rays: 2d, Vegetables which have distinct concentric layers, whose solidity decreases from the centre towards the circumference, and whose pith is contained in a longitudinal canal, and extends in divergent rays. The former structure he considers as peculiar to plants whose seeds are monocotyledonous, and the latter as belonging to those which have dicotyledonous seeds. (Mém. de l'Institut. Nat. tome i. p. 478.)

This opinion, though partly correct, is not universally applicable. That many plants which spring from monocotyledonous seeds are destitute of concentric layers, and have no distinct bark or pith, is certain; but it is not less certain that many herbaceous plants, which are produced from dicotyledonous seeds, are pretty much in the same condition, being equally destitute of concentric layers and of divergent rays, and in which the bark and the pith must be regarded as one continuous structure. On the other hand, some monocotyledonous plants, as M. Desfontaines admits, may deviate a little from the prescribed conditions. In a paper on the organization of such plants, M. Mirbel, who regards this doctrine as the most important step made of late years in Vegetable Anatomy, says, nevertheless, it would be erroneous to assert that they have never a bark. In several species of plants he produces examples to the contrary; and adds, that in some instances their diametral growth goes on at the cir-

cumference, which would seem to approximate them to dicotyledons. As, however, there is no appearance of divergent rays, or of concentric layers, these examples are considered by him rather to confirm than overturn the theory of M. Desfontaines. (Annales du Mus. d'Hist. Nat. tome xiii. p. 67.) But if in this theory its second division embrace only those plants in which the concentric layers are perfect, and divergent rays exist, then it excludes a great number of herbaceous plants, whose seeds have two cotyledons; and if the absence of these regular layers, and of divergent rays, serve as a passport to the first division, then many of these same plants must be admitted among those whose seeds have but one cotyledon. The theory of M. Desfontaines, therefore, rests on too partial an observation of the structure of plants.

The pith (medulla) of plants, when present, occupies the centre of the stem, where it is commonly surrounded by a circle of vessels which construct for it an appropriate canal. In the succulent shoots of trees, its proportion to the other parts is generally large; but it diminishes as the tree advances in age, and is frequently entirely obliterated. Where the vessels of the wood are few in number, as in herbs, only a few fasciculi are seen to surround the pith, and the intervening spaces are occupied by a boundary of thickened cellular tissue. In some plants, again, no pith whatever exists, but the stem is hollow or tubular. In other instances, and especially in roots, the centre of the stem is occupied by vessels; and in others, both cells and vessels, promiscuously blended together, constitute the centre of the stem.

In those plants where the pith is present, and possesses its most perfect form, it is seen to be composed entirely of cellular tissue, possessing often very different shades of colour, but, in its anatomical characters, resembling exactly the description already given of that tissue. Its bulk in different plants is exceedingly different, as are also the form and size of its cells. It is frequently entirely insulated by the surrounding vessels, but is often continuous with the cellular tissue of the bark. Its cells contain, especially in the early age of the plant, aqueous fluids, which afterwards disappear, and then the cells become filled only with air. The "proper juices" of the plant may also be sometimes detected in the cells of the pith. Of the ruptures produced in it by desiccation and other causes, we have already spoken in discoursing of the cellular tissue: they occur particularly in succulent plants, where the cells are large, and their sides thin; so that as the plant advances to maturity, the pith breaks and shrinks up, making the trunk a pipe. (Grew's Anatomy of Plants, p. 129.) We have also noticed the fact, that, within the cavities of the larger cells of the pith, new vesicular productions are sometimes found.

Grew speaks of the existence of vessels in the pith of certain plants, as in that of the fig and the pine; but he adds, that they are usually so postured as to form a ring about its margin. (Anatomy of Plants, p. 119.) They are doubtless to be considered as enlarged proper vessels, which made a part of the first ligneous circle, and have retained therefore nearly the situation in which they were originally formed. Hence, as he observes, they are of divers kinds, answerable to those of the bark, containing in the fig a milky juice, and in the pine a resinous substance. Similar vessels, containing a "proper juice," were observed also in the pith of elder by Malpighi, who seems to regard such appearances as common where the

contained juice concretes or possesses a dark colour. (Anat. Plantar. p. 4.) It is probable, however, that the organs here considered to be vessels may in some cases be cells, into which these juices have been poured; but where real vessels of this kind are found, they are not to be considered as a part of the original structure of the pith, but occurring only in consequence of the changes which the vegetable body undergoes in the progress of its growth.

The general nature of the pith is thus clearly announced by Grew. "Although," says he, "it have a different name from the parenchyma in the bark and the insertions in the wood, yet, as to its substance, it is the very same with them both; whereof there is a double evidence, viz. their continuity and the sameness of their texture; so that all these parts are "one entire piece of work, being only filled up in divers manners with the vessels." (Anatomy of Plants, p. 119.) This continuity of the pith with the cellular tissue of the bark, by means of the insertions or transverse ranges of utriculi, as he calls them, is also adduced by Malpighi as evidence of the similarity of their nature, and of the pith being, as it were, an intercepted portion of the bark (Anat. Plantar. p. 4, 30); an opinion which seems abundantly confirmed by the intermixture of the medullary and cortical textures in many plants, in which, as already remarked, the distinctive characters of bark and pith are alike lost, and the entire stem exhibits only one uniform appearance of structure.

The term medulla, employed by the ancients to note this texture, derived its origin, no doubt, from the resemblance which the pith in the centre of trees bore to the marrow in the bones of animals; and as the same term, in animal anatomy, was incorrectly employed to express alike the marrow in the bones and the nervous substance in the vertebral column, so the same latitude of signification has been extended to the vegetable system. Hence, as Malpighi remarks, the medulla in vegetables was regarded as analogous in its nature to the brain of animals,—a doctrine which even later writers have continued to espouse. It is not our present intention to describe the uses of the pith, but only to remove erroneous opinions concerning its nature, and restore to it that just anatomical character long since assigned it by Malpighi and Grew, and which some writers have of late put forth as a considerable novelty.

Immediately surrounding and enveloping the pith is the part called the wood (lignum vel lignea portio of Malpighi). It is essentially composed of vessels and of cellular tissue, but combined in such an infinite variety of proportions, and exhibiting such a boundless diversity of forms, that it is difficult to seize even its more general features, without the risk of extending our description beyond the limits which our plan necessarily prescribes.

Except in those vegetables in which no vessels have been hitherto demonstrated, but in which they must nevertheless be presumed to exist, this texture may be considered to form a part not only of every plant, but of all its organs; for into whatever part fluids are conveyed, vessels must be supposed to extend; and wherever vessels are present, cellular tissue is to be found: hence, in its distribution, it may be considered the most universal of all the textures. In trees, the vessels, as we have frequently remarked, are very numerous, and, when viewed in a transverse section, are seen to be disposed in layers or concentric circles around the axis, and to stand also in

lines or radii, diverging from the centre of the tree. (See fig. 4, Plate XL.) Between each line or ray of vessels a thin partition of cellular tissue is interposed, which extends in the direction of the ray, through the entire substance of the wood. At certain distances, varying in different trees, thicker transverse portions of the same substance are placed, and are readily distinguishable in almost every species of wood. Between each layer that is annually added to the wood, and each of the smaller layers that go to the formation of the larger one, cellular tissue seems also in some trees to be longitudinally interposed; so that it is probable that in both directions each fasciculus of vessels is intercepted by cellular tissue, and that in such trees no two fasciculi are on any side in immediate contact with each other. It is even probable that the individual vessels which contribute to form the fasciculi are themselves connected by intervening cellular tissue, which acts like the neurilema that holds together the filaments of the fasciculi in the nerves, or the cellular substance that connects the primary filaments in the muscular fibres of animals. In this manner, the whole vascular system of the plant is everywhere connected and held together by cellular tissue. Of this tissue, and the different figures its cells acquire, from the different modes and degrees of compression to which they are exposed, we have already spoken.

In many trees, however, as palms, the vascular fasciculi, though numerous, are much less abundant than in the examples just referred to. They are consequently placed at a greater distance from each other, and, not being disposed in regular lines, do not constitute that radiant appearance so common in ordinary trees, but are promiscuously dispersed through the cellular tissue. (See fig. 26, Plate XXXIX.) As this tissue itself is not, from the same causes, compressed either in the direction or to the extent before described, the smaller membranous partitions that divide the vascular radii from each other are not produced; neither, for similar reasons, is there any distinct appearance of the larger partitions that, at certain distances, intersect the diameter of other trees. The cellular tissue, therefore, in such plants, retains more of its primitive character, and appears everywhere to surround the vascular fasciculi, but nowhere to be so compressed as to form solid partitions between them. In some plants which possess this structure, as the sugar-cane (fig. 25, Plate XXXIX.), the cells indeed retain their perfect forms; and even the fasciculi of vessels, though standing at considerable distances from each other, have towards the centre of the plant a symmetrical arrangement. This latter circumstance is observable in many other plants, which have even fewer vessels than the sugar-cane; so that it is probable that, in the first instance, it takes place in all; and that the irregular position of the vessels in the palm and similar trees, particularly towards their circumference, proceeds from the peculiar modes of their growth, and are not a primary condition of their structure. This intermixture of the vessels and cells in the plants now under consideration extends from the circumference to the centre, so as to constitute their entire bulk, to the exclusion of bark and pith; unless we choose rather to say that in such plants the medullary, ligneous, and cortical textures, are all blended together.

In other examples the vessels form a still smaller portion of the ligneous texture, consisting only of a few fasciculi, which stand at considerable distances from each other, the intervening spaces being occupied by cellular tissue, which forms the chief bulk of the plant. (Fig. 15, Plate XXXIX.) Though few in number, the vessels, however, are symmetrically disposed, and in the same species

preserve always the same position, the fasciculi being placed at the same relative distance from each other and from the common centre of the pith. Sometimes, instead of a few solitary fasciculi, they consist of several ranges, forming an imperfect sort of concentric layers; and in such examples the ligneous texture is commonly separated by distinct but irregular marks from the two other textures. In these plants the cellular tissue preserves its characters, and exhibits no appearance of divergent rays.

The three modes of arrangement above described appear to constitute the chief varieties of structure in the ligneous texture; but in each variety, and through every stage by which they graduate into one another, the greatest diversity of forms prevails. Each species of plant has its peculiar internal structure, as well as its external form; and this seems to be in a great measure determined by the number of vessels that enter into its composition, and their peculiar mode of arrangement. If the vessels are few, the cellular tissue is large in proportion, and its characters are distinct and well preserved;—if they are numerous, and disposed in rays, the tissue is compressed in various directions, loses more or less completely its cellular character, and forms alike those divergent rays or transverse partitions already so often noticed, and those thin membranous expansions or fascia which, both in the bark and wood, are seen sometimes to cover the vessels in a longitudinal direction. Though the whole of the ligneous texture is thus made up of vessels and cells, variously formed and blended together, yet many writers speak of ligneous fibres, which some describe as consisting of vessels, others of cells, and others of an assemblage both of cells and vessels.

The manner, too, or rather the place, in which the vessels are developed, in perennial plants, will greatly contribute to vary the appearance of the ligneous texture. In those trees whose diameter is annually increased by the formation of new vessels around the cylinder of older wood, the new parts must necessarily present in their longitudinal section the appearance of annual layers superimposed on one another, and, in their transverse section, that of concentric circles; but in palms and similar trees, where the development of new parts seems to be accomplished in a different manner, their appearance, under similar sections, may be expected to be different.

In a longitudinal section of the palm, says M. Desfontaines, we discover an assemblage of large ligneous fibres of palms (that is, vascular fasciculi), solid, smooth, and flexible; and these are composed of others still smaller, which are firmly united together: they mostly run parallel to the axis of the trunk, from the base to the summit, without interruption; but some proceed obliquely, and cross the others at angles more or less sharp. (See fig. 24, Plate XXXIX.) In a transverse section of the same stem, continues the author, we remark neither concentric circles nor transverse partitions; but the fasciculi of vessels, placed without order by the side of each other, are enveloped by the cellular tissue, which fills up all the intervals: they sensibly approach each other, harden and diminish in size in proceeding from the centre to the circumference (see fig. 26, Plate XXXIX.); so that the stem has much more strength and solidity near the surface than in the interior,—an organization quite distinct from that of ordinary trees.

The cause of this diversity of structure seems to be amply accounted for by the different modes in which the growth of these trees is accomplished. When the seed of a palm is sown, the leaves, says M. Desfontaines, successively develop and augment in number for four or five years; the neck of the root augments in the same proportion; the bulbous part, formed by the re-union of

the petioles of the leaves, increases insensibly; its solidity augments, and at length the stem rises above the earth with all the size it ever acquires. Its figure is cylindrical from the base to the summit; and if the diameter be measured at different epochs, it will be seen, as Kämpfer had already remarked, not to increase. The palm, therefore, is a regular column, whose summit is crowned with leaves, disposed above each other circularly: those which grow in spring shoot always from the top; the older ones, placed below, dry, and when they fall leave circular impressions, which furrow the surface of the stem, and mark its years until it has ceased to grow.

If next we examine the interior, we discover, as M. Desfontaines thinks, the true reason why the stem rises in a column, and does not, like other trees, yearly augment in size. This was done by M. Daubenton, who states that every leaf of the date-palm, in proceeding from the bud, is formed by a prolongation of the vascular fasciculi and cellular tissue which exist in the trunk of the tree, as is apparent in the petiole of the recent leaves, and of the dried ones that adhere to the trunk. The elongation of the trunk is produced, therefore, by the leaves which annually proceed from it; and as the parts which form these leaves spring from the centre, they always, as they shoot, force the older leaves outwards. Hence, therefore, as the augmentation of these trees originates at the centre, all the parts capable of displacement must be pushed outwards, just as the new layers of bark and wood, formed annually in ordinary trees, force outward the older layers of bark exterior to them.

In these latter trees, continues M. Daubenton, the recession of the bark has no limits so long as new parts continue to be formed beneath it, because the new cortical layers are flexible, and the older ones readily break and give way; but in the palm the substance of the trunk has more compactness as we approach the circumference, and, at a certain point of density, it no longer yields to the central force of the interior parts; so that when this point is reached, no further enlargement takes place; and hence the date-palm scarcely exceeds ten inches diameter. It is for similar reasons that the trunk of the date-palm is of the same size through its entire length; for, in proportion as the tree rises, the exterior parts of the trunk lose successively their flexibility; and when they have acquired a certain degree of density they no longer yield to the force from within; and as this is equally the case in all parts, the trunk is necessarily of the same size throughout. (Mém. de l'Institut. Nat. tome i. p. 482, &c.)

It is further evident that, in this mode of growth, no appearance of concentric circles, similar to those of ordinary trees, can have place; for, by the growth at the centre, the exterior vessels are continually displaced from their original positions, are more and more compressed as they are forced towards the circumference, and present in their transverse section that irregular distribution which they have been described to possess. Hence the cylindrical figure and the absence of concentric layers are as necessary consequences of the mode of growth in these trees, as the presence of those layers and the conical figure are of the mode of growth in ordinary trees. The greater solidity of the parts at the circumference is clearly to be ascribed to the same cause; and even the want of regular transverse partitions must in part also have a similar origin, and be ascribed, perhaps, in part to the smaller number of vessels which these plants possess, as well as to their irregular distribution. From this difference in their mode of growth, these two great divisions have received different names: those

which, like our ordinary trees, augment in size by the deposition of new matter near the surface, are named Exogenous; and those which, like palms, grow from within, Endogenous.

This texture, in its component parts, resembles that of the bark, the wood, being made up, like it, of vessels and cellular tissue intimately connected with each other. Its structure, as a distinct texture, is best characterized in the bark of ordinary trees, as it is there separated, in a great measure, from the ligneous texture. As in these plants a new layer of vessels is annually made to the wood, so a similar but much thinner layer is yearly added to the bark, to which the name of liber has commonly been applied. These vessels are at first straight, and run parallel to the axis of the trunk; but, by the successive formation of new layers beneath them, they are gradually forced outward, become separated more and more from each other, and, touching in a few points only, exhibit at length a reticulated figure (see fig. 24, Plate XXXVIII.), the meshes of which yearly augment in size, from the greater space over which they are continually spread.

Between the vessels thus annually formed, a considerable portion of cellular tissue is interposed, which, in the young and succulent state of parts, contributes chiefly to the thickness of the bark. This tissue is variously compressed by the vessels, so as to form transverse partitions between them, which, in the vine, the oak, and many other trees, as both Grew and Malpighi remarked, are seen to be continuous with those of the wood; and in this way the two textures are united together. In the expressive language of Grew, the bark, therefore, does not "merely surround the wood as a scabbard does a sword, or a glove the hand, but is truly continuous with it, as the skin of the body is with the flesh." In the willow and other trees, when full of sap, the bark is so easily separated that it seems to have no connection with the wood; but this is supposed by Grew to arise merely from the extreme fineness and tenderness of the vessels that are annually formed in that part, and which on that account oppose no obstacle to the separation. (Anat. of Plants, p. 129.) It is probable, however, that the cellular tissue forms the only direct connection between the cortical and ligneous textures; and that, if a vascular communication exist, it is only, as in all other cases, through the medium of that tissue.

Besides the transverse compression which the cellular tissue experiences from the vessels, it is compressed in the opposite direction by the formation of the new layers of bark and wood beneath the older bark. These in the progress of their growth exert an expansive force outwards, so that the cells of the tissue are made to assume an oblong or flattened form in the direction of the vessels of the trunk, or sometimes to form a thin fascia upon the vessels, in which the cellular character is nearly or entirely obliterated. It is by the continued exertion of this force acting on the exterior and desiccated layers that these latter ultimately crack, producing figures of different sizes, which have frequently the shape of rhombs, the fissures of which represent, according to Grew, the position and track of the vessels in their reticulations. (Anat. of Plants, p. 129.) The spaces or meshes formed by these reticulations are always filled up by cellular tissue, which, in the opinion of Malpighi, originates from the vessels themselves.

Both in the vessels and cells of this texture, collections

of the "proper juices" frequently occur; especially in plants in which these juices are of a viscid nature, and disposed to concrete; as was before remarked, when treating of the proper vessels. Similar collections of matter, originally existing in the bark, are likewise met with in every layer of wood, and even in the pith of certain trees. It is only in plants in which the "proper juices" are coloured, or disposed to concrete, that this intermixture of the cortical and ligneous textures has been traced through the whole substance of the tree; but it is probable, from the conjunction of the vessels of the bark and wood at the period of their formation, that it is common to other plants, the nature and properties of whose juices afford no clew to its detection.

In most herbaceous plants, the cortical texture is not so clearly distinguished from the ligneous, since in them the greatest variety obtains both in regard to the number and relative position of the sap and "proper vessels;" and very frequently the cellular tissue is quite continuous and of uniform appearance through the entire substance of the plant. In general, however, the sap-vessels occupy the inner place, and are surrounded by the "proper vessels," disposed either in rings or distinct fasciculi, more or fewer in number. Sometimes the sap-vessels seem to be placed exterior to the others; and hence it is difficult to discover the true place of the "proper vessels" in such plants, unless the nature of their juices conduct to it. From the position of the vessels, therefore, it is often difficult to define the boundaries of the cortical and ligneous textures; and, for the reason already assigned, the cellular tissue rarely affords much assistance. According, however, to the number and disposition of the vessels, this tissue, even in these plants, suffers a certain degree of compression, so as to form thickened boundaries around the fasciculi, or sometimes large transverse partitions, which communicate with those of the wood, in which, though the figure of the cells is more or less altered, the cellular character is usually preserved.

In other plants, the intermixture of "proper vessels" with those that carry sap seems to be general through the whole plant; and consequently no distinction can be made between the ligneous and cortical textures. In such plants the "proper juices" must be considered to exist in every part; and accordingly Malpighi, as we before remarked, points out a vas proprium, or "proper vessel," as accompanying every fasciculus of vessels in different species of wheat. (Anat. Plantar. p. 24.) From the similarity of structure in palms, and especially from their mode of growth, there can be little doubt that a similar intermixture of the two kinds of vessels prevails everywhere in them; and with respect to these plants, what has already been said of the construction of the ligneous texture, is equally applicable to that of the bark.

We have thus given a very brief and general view of the principal textures that enter into the construction of plants, and pointed out the more prominent diversities of character and appearance which they exhibit, as well in their simple as in their more complex forms, and as they exist either separately or variously intermingled together. Our descriptions have been confined entirely to the trunk or stem; but, with slight variation, they are applicable equally to the root and branch, in which a similar combination of the elementary organs obtains. In the root, however, they commonly exist in a more compressed and compacted form, so that the ligneous texture is seen chiefly to predominate, frequently to the entire exclusion of the pith, and often in great part also to that of the cellular portion of the bark. This, however, is not universal, especially in annual plants, some of which, as the carrot and others, particularly in a cultivated state, are distinguished

by the very large portion of cellular tissue which enters into the construction of the cortical texture. Of the modifications of these several textures, as they exist in leaves and other organs, we shall have occasion to speak when we come to treat of them individually.

SECTION IV.

Of the Connection subsisting between the Vessels and Cells in the several Textures.

In our discussion of these several textures, we have noticed only in a general way the direct means by which the vessels and cells that construct them are connected with each other; but, when treating of the sap-vessels, of the absorbent vessels, and of the cellular tissue, we endeavoured to show that an universal communication obtains between these elementary organs, and consequently inferred that some mode of connection, by which it can be accomplished, must have place. Grew considered the vascular and cellular parts to be connected with each other, not only by the transverse partitions of cellular substance that intercept the vessels, but "per minimas partes organicas;" that is to say, the parenchymous fibres are wrapped round about the vessels, or at least interwoven with them, and with every fibre of every vessel, as in very white ash or fir wood may be observed. (Anat. of Plants, p. 121.) These fibres Grew elsewhere considers as vessels, and consequently must be regarded as maintaining a vascular connection between these organs. In a description of the young branches of chesnut and of oak, Malpighi delineates minute filaments springing on every side from the vessels of the wood, and continued into the adjacent cellular tissue (Anat. Plantar. p. 27); but he does not state whether they are to be regarded as vessels or simple fibres. He elsewhere describes minute vessels as surrounding the cells of the trunk like a network, and thinks it probable that the cells of the transverse septa are similarly furnished with vessels. In the cells of the pith of elder (sambucus), and of some other plants, the vessels are very abundant, and in every case are probably derived from the straight ligneous fibres (vessels) both of the bark and wood. (Ibid. p. 29.) Indeed he held it probable that the nutrient fluids moving through the vessels were in all parts poured into the cells, and there undergoing a certain preparation, were afterwards mixed with more recent juices, and with them taken up and applied to the support of the young buds and leaves. (Ibid. p. 30.) This doctrine has since been held by Darwin and Knight, and it necessarily supposes a vascular connection between the vessels and the cells, by which the functions both of secretion and absorption can be performed. The microscopical observations of Leeuwenhoek, already noticed, supply further evidence in support of this opinion.

In the hypothesis of Mirbel both cells and vessels are considered as formed out of one and the same membrane. He rejects, therefore, the aid of all intermediate organs as necessary to connect them together, and supposes a communication to be everywhere maintained between the vessels and cells, by the medium of pores in their sides. As, however, these pores are nowhere proved to have existence but in the imagination of the author, we may altogether reject their agency in maintaining a communication between the vessels and cells of plants. In the opinion of M. Kieser, the conjunction of the cells with the vessels is extremely simple, the sides of the cells, says he, being contiguous to the sides of the vessels. (Mém. sur l'Organisat. des Plantes, p. 94.) But mere contiguity of parts does not amount to connection, much less does it afford any information concerning the actual communica-

tion that exists between these organs. In addition, therefore, to connection by cellular substance, it seems absolutely necessary to suppose also the existence of a vascular structure, which shall at once serve as a medium both of connection and communication.

Before concluding this branch of the subject, we may observe that the structure of the cellular tissue, and its relation to the vascular system in plants, appear, in many points, to resemble that of the adipose cells, and their relation to the vascular system of animals. These cells are described as minute close cavities, possessing no apparent communication with each other; and within them adipose matter is alternately deposited and removed. Now the deposition of this matter could only be accomplished by secreting vessels which terminated in the cells, and its removal be effected by absorbents which originated from them; and accordingly both blood-vessels and absorbents are found to be present in this texture; but neither the secreting nor absorbing orifices have ever been actually observed. Within the cells of the cellular tissue of plants the alternate deposition and removal of various matters are not less certain; and in the germinating seed the matter that actually existed in the cells is found afterwards in the vessels. We are led, therefore, or rather we are driven, not only by the direct exclusion of all other alleged means of communication, but by a close analogy in the exercise of these animal and vegetable functions, to conclude, that secreting and absorbing vessels must be employed to deposit and remove the secreted matter from the cells of plants, in the same way as they are considered to effect similar depositions and removals of adipose matter from the cells of animals: and as this alternate function seems to go on in every part of the plant capable of active vegetation, it may further be inferred that a vascular communication exists between the vessels and cells in all parts of the vegetable system.

By means of this general communication between the vessels and the cells, we are enabled to assign satisfactory reasons for some puzzling phenomena, which have occurred in relation to the movements of the sap. It is by this alternate action of secretion and absorption that in young plants we must suppose the cells of the pith, during the first year, to be filled with fluid, and to be rendered dry for the most part ever after. In like manner, the surface of the bark in contact with the wood appears in some trees, as the birch, to be rendered moist during the rise of the sap in spring; which led Dr Walker and others to suppose that the sap rose in part between the bark and the wood,—an opinion not at all probable in itself, and certainly not supported by what is observed in most other trees. The fact, however, is easily explicable, on the supposition that the sap was transfused from the albuminous vessels of the wood, in the same manner as, at a later period, it is secreted in the same part, but in a different form, by the vessels of the bark, to form the new matter that is annually added to the tree.

That the sap of plants was capable of moving in a lateral direction, was inferred by Malpighi, from the fact that parts lived and grew when the perpendicular vessels that supplied them with nutriment had been destroyed. (Anat. Plantar. p. 13.) The experiments of Hales afford more decisive evidence regarding this lateral movement of the sap. He cut two large gaps in the opposite sides of an oak-branch, at four inches distance from each other, carrying the incisions down to the pith: the branch nevertheless absorbed and perspired water, but only in half the quantity that another similar but uncut branch did. In a branch of cherry-tree he made four similar cuts down to the pith, at four inches distance from each

other, and opposed to the four points of the compass: the branch notwithstanding absorbed, in forty-eight hours, twenty-four ounces of water. (Veg. Statics, p. 128, 3d edit.) And when similar incisions were made on branches while still attached to the tree, their leaves continued green nearly as long as those of other branches in a natural state; whence he justly inferred that, at these gaps made in the branch, a lateral movement of the sap must have taken place. Experiments of a similar nature have been made, and like results obtained, by Mr Knight (Phil. Trans. 1808); so that it seems clear that in certain circumstances a lateral movement of the sap must have place.

In what manner, then, must we suppose this movement to be accomplished? Grew supposed the cellular tissue, that stretches from the circumference to the centre of the plant, to be the organ by which such a communication could be maintained; but the impermeability of this tissue to fluids opposes such an opinion. Malpighi thought the lateral communication to be made by an anastomosis of vessels; but in the vessels of plants no such mode of communication appears to exist. From the ascent of the sap in branches in which the vessels had been thus previously cut through, Mr Knight infers that this fluid does not rise in the vessels at all, but is conveyed through the cellular tissue. This opinion necessarily implies the permeability of this tissue by fluids, which, as we have shown, is contradicted by direct experiment, as well as by microscopical observation. Since, therefore, this lateral movement of the sap cannot be accomplished, either by simple percolation through the cells or vessels, or by direct anastomosis of the vessels with one another, no other known means of effecting it remain, but those of alternate deposition and absorption by the vessels into and from the cells. And if, as we have seen, the sap-vessels of plants deposit coloured fluids in the cells, which the capillary absorbents of parasitic plants are able to take up, there seems no reason for denying to the vascular productions, which have been supposed everywhere to spring from the perpendicular vessels, a like capacity of absorbing fluids from the adjacent cells. These fluids must, however, in all cases, have been deposited before they could be absorbed; and, by the alternate exercise of these functions, there is no difficulty in conceiving how a lateral movement of the sap might be accomplished in parts where, by the incision of the vessels, a stop was necessarily put to its perpendicular ascent.

SECTION V.

Of the Skin or Cuticular Texture, and its Appendages.

ART. I.—Description and Structure of the Skin.

The skin, rind, cuticle, or epidermis, as it has been variously named, is the last of the common textures that remains to be described. It is the general envelope which invests all parts of the plant and all its productions, being equally common to the trunk and branches, the root, the leaves, the flowers, and the fruit; but in these different parts, and even in similar parts of different plants, it exhibits the greatest diversity of appearance and form.

In herbaceous plants, and in the young shoots of those which are arborescent, it resembles a thin membrane, but is generally thicker on the stem than on the roots or leaves. In some leaves, however, it is thick and dense, in young as is the case also in several fruits, and is thereby fitted to resist the effects of too rapid desiccation. On the upper surface of some leaves, on many fruits, and on roots, it is an entire membrane, destitute of any apertures

Common or pores; but on many stems on the under surfaces of
Textures. leaves, and sometimes on the upper, it is frequently furnished with numerous pores, often visible to the naked eye, and with other luminous points of smaller dimensions, which Du Hamel also regards as apertures. It is readily separable from the bark in recent and succulent parts, or after maceration in water; and in certain leaves it is very completely separated by a species of caterpillar, named by Reaumur the miner. It appears then to be a thin transparent membrane, often destitute of colour, and deriving, therefore, its appearance from the colour of the parts beneath; but both in leaves and flowers it is often itself coloured. It is frequently seen to extend in all its dimensions, in common with the parts it covers. Very often, too, as will be noticed hereafter, its surface is covered with hairs; and sometimes small follicles or utricles are met with, which exercise a glandular function.

In old The characters above enumerated belong chiefly to the
trees. cuticle in its young and succulent state. In perennial plants it commonly possesses others that are quite dissimilar. It is of a different colour not only on different trees, but on different parts of the same tree. It is white and shining on the trunk of the birch, and browner on the branches; greyish on the plum-tree; red and silvery on the cherry; green on the young branches of the peach; and ashi-coloured on the larger branches. In these and many other instances it does not, says Du Hamel, merely participate in the colour of the body it covers, but contributes itself to give colour to the exterior bark; for when it is stripped off, the substance below has frequently a different colour. (Phys. des Arbres, tome i. p. 10.) By the gradual enlargement of the trunk it is stretched and dried, and at length loses its vitality, and, as well as the bark beneath, is variously cracked and broken. Before this happens, however, it often undergoes considerable extension in all its dimensions, enlarging in breadth, and stretching longitudinally over the young shoots. This expanded state is particularly remarkable in certain fruits, in which, when they enlarge slowly, the cuticle is extended without rupture to a very large size; but if the expansion be very rapid, as after considerable rains, the cuticle then gives way. In certain trees the cuticle is more susceptible of expansion than in others; and in very vigorous trees it breaks more slowly than in those whose growth is languishing, although these latter push forward more slowly than the former. (Ibid. p. 11.) In some vigorous trees of this description it altogether resists rupture; and in this state the tree is often said to be hide-bound or bark-bound.

Composed of layers. In most instances the cuticle, when taken from young branches, appears to consist of a single layer; but on the branches of many species, says Du Hamel, after one plate or layer has been removed, another may be seen beneath, which resembles the former in its texture, but is much thinner and more green and succulent. From the birch-tree he has removed more than six layers, very thin and very distinct from each other, and is of opinion that more might still be separated. Sometimes the original cuticle seems to be entirely thrown off, and the exterior covering is formed by a portion of the cellular tissue of the bark. Grew thinks that this substitution takes place annually, the older skin being cast off, like the skin of an adder, by the generation of a new one beneath. (Anat. of Plants, p. 114.) Du Hamel describes also the existence of small leaflets or scales, which are continually detached from the cuticle of some trees; and these he considers to be as constantly replaced, by the formation of new ones beneath

Is regenerated. Concerning the regeneration of the cuticle on parts from

which it has been removed, Du Hamel observes, that when the wound is covered with waxed cloth, a new cuticle is promptly formed without any separation of a portion of the bark beneath. When the exterior portion of the bark is removed with the skin, the inner part of it is equally capable of regenerating a cuticle; but if the wound be not protected from the air, a certain degree of exfoliation first occurs, and, under the decayed parts, a new skin forms. Even where the bark of a cherry-tree was entirely removed from the trunk, he found that the wood was capable of regenerating a new bark and cuticle, if the parts were properly protected from the air. This cuticle did not originate from that which remained on the roots and branches, but was reproduced in isolated portions on different parts of the trunk; it continued, however, after the lapse of fifteen years, always different from that of the natural growth. In other instances, he adds, the cuticle does not seem to be regenerated at all. He remarks certain analogies to exist between the cuticle in some plants and in animals. In both, he adds, it seems in certain circumstances capable of great extension; in both it is easily regenerated, and that too in isolated portions, and not by continuity of organs, as is common in other instances; and in both, lastly, it is perpetually obliterated, and continually and imperceptibly renewed. (Phys. des Arbres, tome i. p. 11.)

With respect to the nature of the cuticle, very different opinions have been advanced, and still continue to prevail. Are we to regard it as a peculiar organ, formed immediately by the proper exercise of the vegetative functions, or is it produced in a sort of secondary manner, by some changes induced on some previously constructed organ? Grew asserted it to be sometimes original, and in some instances produced out of the exterior layer of the cortical texture beneath it; and this first view of its origin seems to be generally supported by the descriptive character which has been assigned to it. Its co-existence with the first traces of vegetable organization, its continued growth and expansion, and its subsequent regeneration after removal, all seem to favour its primary and independent origin, which is also supported by investigations into its minute structure. Mirbel, however, and some other writers, after Hill, have regarded it, in all cases, not as an original membrane, but formed by the exterior sides of the common tissue of the plant; and where there is no separation of these sides in the form of a membrane, such plants are held to be destitute of a cuticle. (Exposit. de l'Organisat. Végét. p. 103.)

Another question relating to this organ is, whether it must be considered a simple membrane of uniform structure, or a compound of two distinct parts, like the true skin and the cuticle in animals. Grew seems to have regarded it as a simple body, but constructed both of vessels and cells, the cells being continuous with those of the bark. (Anat. of Plants, p. 62.) Such, too, seems to have been nearly the opinion of Malpighi, who describes it as constructed of horizontal ranges of cells, but often delineates reticulations of vessels as constituting a part of its structure. (Anat. Plantar. p. 2, 19.) In the birch, the plum, the cherry-tree, and others, Du Hamel declares the component fibres of the cuticle to possess a direction transverse to that of the trunk; but this is not general. In the birch-tree the fibres seemed to be placed parallel to each other, and to be connected together by lateral fibres; but he could see nothing of the vesicular structure of Malpighi and Grew, and therefore regards the structure of this texture to be altogether fibrous. (Phys. des Arbres, tome i. p. 8 and 9.) M. Desfontaines, on the other hand, describes it as a membrane resembling in ap-

pearance a thin plate of parchment, and perforated by imperceptible pores, which give issue to the insensible transpiration. Its structure he regards as unknown, but considers it capable of regeneration. (Mém. de l'Institut. Nat. tome i. p. 481.)

M. Kieser, who professes to have studied this texture with great attention, adopts nearly the opinion of Grew, pronouncing the cuticle to be constructed of a very fine cellular tissue, and of extremely minute vessels which run through its whole extent. These vessels form an exceedingly delicate and subtle net-work, the meshes of which possess very different forms, and their vessels terminate at the orifice of a pore. His observations were made on the cuticle of leaves. On the inferior surface of the leaf of Amaryllis formosissima (fig. 14, A, Plate XLI.), magnified 260 times, these vascular meshes of the cuticle have an elongated hexagonal form; and four of their vessels proceed always to terminate at the orifice of the little oblong aperture or pore situated at their junction. In Canna Indica, the vessels of the meshes on the lower part of the leaf, which thus terminate in the pores, are said to originate from a fasciculus of the spiral vessels that ramify through the leaf, as is represented at d' in fig. 14, B, Plate XLI.; and within the areas of the larger meshes a still finer net-work of vessels is seen. On the inferior surface of the leaf of a species of fern, the vessels of the cuticle, instead of forming meshes of different figures, exhibit the appearance of sinuous lines, which run in every direction through the cuticle. See fig. 15, Plate XLI. which represents the central part of the little adjoining leaf, magnified 130 times. These sinuous vessels often join, and, after making a half-circle, terminate by one extremity in the minute pores everywhere spread over the leaf, and by the other in the larger vascular fasciculi that ramify through it. At the letter b, in this figure, the hexagonal cells that construct the parenchyma of the leaf are distinctly visible through the vascular sinuities of the cuticle. It was by the examination of this leaf that M. Kieser was first enabled to discover the origin and termination of the vessels that construct the cuticle, having in all his previous investigations examined the cuticle in its separated state, after it was detached from its connection with the other organs; but the researches made on this leaf rendered every thing clear. (Mém. sur l'Organisat. des Plantes, p. 141-2.)

The vascular net-work of the cuticle, thus described by M. Kieser and others, had been regarded as a deception by M. Krocker, who considered these reticulated figures as no part of the real structure of the cuticle, but merely as the sides of the subjacent cells; in which opinion Sprengel, Link, Jurine, and Mirbel, concurred; but M. Kieser, in opposition to this opinion, maintains that, in the fern and other leaves, the real cellular structure of the parenchyma is seen entire through the vascular reticulations of the cuticle, with the meshes of which the sides of the subjacent cells do not anywhere coincide. He observes that these cells are commonly much smaller than the vascular meshes which cover them; and that the vessels of these meshes may be traced, as before remarked, to the larger fasciculi that construct the leaf. In the fern the vascular structure of the cuticle is the same on both sides of the leaf, but the superior side does not possess pores.

A very different view of the structure of the cuticle was taken by the late celebrated M. de Saussure. He regarded it not as a simple, but a compound texture, consisting of a very delicate external pellicle or membrane, beneath which was placed a net-work of very fine vessels. The external membrane he describes as perforated by

pores of unequal figure, between which he observed some opaque and tortuous filaments, disposed in a reticulated manner, each mesh being formed by six filaments, four of which terminated at each pore. To this arrangement of filaments he gave the name of cortical net-work, and regarded it as quite distinct from the cuticle that covered it. The meshes of this net-work differ much in size and figure in different leaves; and, when minutely examined, they are often seen to form junctions, but never to cross each other; whence he was led to regard them as vessels derived from those of the expanded petiole, and thus constituting a very fine vascular net-work. A similar structure was observed in the petals of the flower. (Encyclop. Méthod. tome i. p. 67.)

M. de Candolle considers the cuticle in certain parts to be a simple and proper membrane; in other parts to be De Candolle. formed by layers of cellular tissue. The cuticle of leaves, and probably of all annual shoots, appears to be formed of a layer of flattened cells, differing in form and in other conditions from those of the parenchyma of the leaf. It may be regarded in this state as a proper membrane. When stripped off and examined by the microscope, it exhibits small spaces, bounded by lines or rays, or a sort of net-work, the meshes of which have different figures in different plants. The lines, says he, appear in the form of single or double filaments, and may be considered hollow, and as forming a system of cuticular vessels which terminate in the pores. In more advanced age, and especially in the trunks of old trees, this primitive cuticle gradually changes its character, and finally disappears. It is then succeeded by a thicker membrane, which no longer exhibits the net-like figure of the former, and appears evidently to be formed by the exterior cells of the cellular tissue, which, by the combined effects of distention and desiccation, assume a membranous appearance, to which the term epidermis is properly applicable. This epidermis may be considered single when composed of one layer of cells, and double, triple, or multiple, when formed of two or more successive layers. It is thus that Ulloa, in describing a tree in Peru, speaks of having detached more than 150 epidermoidal layers, when he lost all patience in counting them, seeing that he had not then reached to half the thickness of the bark. A similar appearance is exhibited in the numerous layers that form the epidermis of the white birch, amounting sometimes to 15 or 18 in advanced age, and which ends by breaking into that cracked condition of the bark that presents only discontinuous portions of white epidermis on the remnants of its cellular envelope. At an early period the cuticle is most easily torn in a longitudinal direction, which is that of its growth; but at a later period, when the diametral growth has increased, the cells of the tissue are drawn out more in that direction, and therefore more readily break transversely than in length. (Organog. Végét. tome i. chap. 5.)

ART. II.—Of the Pores of the Skin.

The pores of the cuticle, called sometimes cortical or Descriptive. sometimes exhalant pores, sometimes cortical pores, and more lately, by Link and De Candolle, pores by stomata, were first noticed by Grew, who describes many orifices as existing on the leaves of different plants, which vary in size, number, shape, and position. In the white lily they are of an oval shape, of a white colour; and each is surrounded by a slender border. When viewed through a good glass they appear as if standing about one sixth or one eighth of an inch apart all over the leaf, but not arranged in any regular order. In the pine, also,

they have an oval shape, but have no rising border, and are arranged in lines from one end of the leaf to the other. (Anat. of Plants, p. 153.) Hedwig considered the borders mentioned by Grew as produced by a ring of one or more vessels, which terminated in the pore. The number of pores he represents as exceedingly great. In the square of a line of the cuticle of a bulbous lily he reckoned 577 pores.

The characters and position of these pores or stomata have been further examined by M. de Candolle on more than 600 plants. They occur most frequently on the leaves, occupying both surfaces in some herbs, and in trees chiefly the inferior surface. Stems in general have no pores, except, as in the Gramineæ, where they are succulent, and have the character of leaves; or where the plant is altogether destitute of leaves, as the Cactus. On the prominent lines or vessels of the leaves and stems no pores are to be seen, but only in the grooves or depressed surfaces of the parenchyma. They are never observed on the root, not even on bulbous roots, where the scales of the bulb are true leaves. The small leaflets called stipulae and bractea sometimes have and sometimes have not pores. The calices of the flower in general have pores, but the petals have not. Pericarps of a foliaceous consistence have pores: when fleshy, they are destitute of them. The envelopes of the seed are destitute of pores, but they are found on all seminal leaves that rise above the ground. The lower tribes of vegetables, as the fici, musc, hepaticæ, fungi, &c. are destitute of pores.

The occurrence of pores in those plants where they are found seems to be much influenced by external circumstances. They are never met with but on vegetables, and those parts of vegetables that are exposed to the air; and therefore the internal surfaces of many leaves that embrace the stem are without pores, though on the external surfaces of the same leaves they are abundant. No plant that is completely aquatic, nor any part of it that is habitually under water, is provided with these organs; but the parts which rise above the water are furnished with them. In ranunculus aquaticus the leaves that are constantly under water are destitute of pores, while those that float on the surface are provided with them, but only on their superior face. Even leaves which do not naturally possess pores when under water, acquire them if they are made to grow in air; and land plants, on the contrary, when made to grow under water, may, by such treatment, be deprived of their pores. Thus, the leaf of green mint, when growing in air, possesses not fewer than 1800 pores on its lower surface; but if kept for a month under water, its leaves fall, and the new ones that succeed are destitute of pores.

Light seems to be necessary also to the production of pores, for etiolated plants do not possess them. When grown by the light of lamps, the leaves possess a few pores; and, in all cases, the parts secluded from light and air are destitute of these organs, but acquire them if they are duly exposed. (Mém. de l'Institut Nat. tome i. p. 351.)

This general account of the pores of plants is confirmed by the researches of M. Rudolphi. In most herbaceous plants he found the pores to occupy both sides of the leaf, but in trees only the inferior surface. They were not often met with on the parts of the flowers, or on fruits; they were never seen on roots, nor on the trunks of trees; nor ever on aquatic plants, except on such parts as were raised above the water. The lower tribes of vegetables seemed to be universally destitute of them; the leaves also of those plants that were covered thickly with hairs on both sides had no pores. The form of the pores was commonly oval or elliptical, but in a few instances square

or rhomboidal. In size they very much varied in different plants, but in the same plant the size was uniform. The largest pores were seen on the leaf of the white lily, the smallest on that of the French bean. (Kieser's Mém. sur l'Organisat. des Plantes, p. 144.)

Although the pores are generally dispersed over the parenchyma of leaves at nearly equal distances, yet where the vessels run parallel, they are disposed in one or more lines between them. Sometimes, again, they are collected into little clusters, as in the leaves of crassula cordata, where the roundish dots or points, visible by the naked eye, are clusters of pores or stomata. This appearance suggested to M. de Candolle the idea that the stomata might be the orifices of vessels, as each of the little points mentioned above is the termination of a fibre, which is itself a bundle of vessels. This idea is further strengthened by the fact that these pores are not found in plants that are destitute of vessels; and though he could never trace the continuity of a vessel with a pore, yet M. Comperetti assures us he has seen the vessels terminate in them. Mirbel, on the other hand, considers the pores to be the orifices of cells; and Kieser regards them as connected with the intercellular canals. (Organog. Végét. tome i. chap. 6.)

Some have considered the pores as organs, by which the resinous or waxy matter found on certain leaves is exhaled; others, as organs for the absorption and transpiration of air or gas; and others, with more probability, as those by which moisture is thrown off by the leaf. In support of this last opinion, it may be stated that pores exist in all leafy parts that transpire, and are more numerous in membranous leaves that transpire most, than in fleshy leaves which transpire little. They are wanting altogether in aquatic leaves and in etiolated leaves; also in fleshy fruits, in roots, and in the petals of flowers, which do not transpire in any degree analogous to leaves. In darkness too, when transpiration ceases, the pores are closed; and they again open and transpire when light reappears, and especially in bright sunshine. It is always necessary to distinguish the evaporation, more or less great, that goes on through the tissue in all organs by day and by night, from the active transpiration which occurs in sunshine, but only in organs furnished with pores, and which seems to be executed by them.

Besides transpiration, M. de Candolle believes that, in certain cases, the pores may also serve for absorption; but by this he considers to be rarely the case, and out of the natural course of vegetation; and that the experiments which favour this opinion may be explained by the hygroscopic powers of the vegetable tissue. He thinks the results obtained by Bonnet, who kept leaves a long time in life by laying their porous surfaces on water, may be explained by supposing this treatment to prevent decay by checking transpiration, rather than by promoting absorption. (Organog. Végét. tome i. p. 87.) This suggestion will not, however, explain the phenomena of an experiment continued for many years in the conservatories of the Royal Botanic Garden here, under the care of Mr Macnab. He has kept a species of fig (ficus elastica) for several years fastened against the wall, with its roots, stem, and branches entirely exposed to the air; and in that situation it has not only lived, but actually grown in every part, so as to have increased greatly in size, only by being thus kept in a warm and moist atmosphere, and occasionally sprinkled with water.

In addition to these visible pores, it is probable, says M. de Candolle, that the surface of vegetables is furnished with others that are invisible, but so small that with the strongest microscopes we are unable to recognise them:

their existence, therefore, can only be presumed from physiological phenomena. Thus, a part of a vegetable destitute of pores loses weight when exposed to the air, by the escape of the fluid it contained; or a portion of moss equally destitute of pores, if placed in water, speedily acquires additional weight. Whether these results are obtained through the medium of minute pores, or of unorganized apertures, such as are admitted to exist between the molecules of all matters,—and whether through the same pores or apertures a passage is afforded to gases or other fluids, or to the oily or waxy secretions that cover certain surfaces,—are questions, adds M. de Candolle, to which no answer can at present be given. (Organog. Vég. tome i. p. 88.)

From the surface of the cuticle, in many parts of herbaceous plants, and in the succulent parts of arborescent ones, hairs (pili) are seen to spring. They possess very different forms, and vary likewise greatly in texture. In a strict sense, they may be defined small filaments possessing considerable stiffness, which project from the surface, and stand out pretty erect. When they are very numerous, a little soft, and less erect, they take the name of villi; when still softer and less numerous they are termed down (pubes). Sometimes this down is composed of long hairs nearly resembling wool, at other times it approaches more to the character of cotton. When the hairs are stiff and ranged along the edge of a surface, like the lashes of the eye, they are named cilia; and if, with these characters, they are produced to a greater length, as in the beard or awn of wheat, they acquire the name of barba or arista. Sometimes they resemble the bristles of the hog, and are then called setæ. Many other varieties are enumerated by botanists, who further distinguish them by various names, according as they terminate in a single point, or are hooked, or forked, or branched, or feathered, &c. In some instances, instead of appearing like one continuous substance, they are composed of many joints, or are said to be articulated. In fig. 16 and 17, Plate XLI., we have copied from Du Hamel a few of the varieties, both of single and jointed hairs; but the forms they exhibit are so numerous and diversified, that we must refer to the writers on botany for minutier information. In some examples the point of the hair is terminated by a small rounded globulet, and sometimes by a fine filament, that seems to proceed out of the hair.

Some writers distinguish hairs into two general classes,—the glandular and lymphatic. In the former class, the hair or filament is sometimes the excretory duct of the small gland situated at its base; and in other instances, as in the chick-pea, the glandular or secreting organ is seated at the extremity of the hair. The lymphatic class embraces a much greater variety of hairs, which differ widely from each other in consistence, direction, and form. They spring only from the parts of plants which are exposed to the air, and are not found, therefore, on the parts concealed below the earth or in water: they are also rare on plants that grow in the shade, are wanting altogether in etiolated plants, and are most abundant, in general, on plants that grow in warm places, and are well exposed to solar light. With respect to their seat or origin, as compared with that of the pores, although, says M. de Candolle, these two kinds of organs appear sometimes mingled, yet each has a determined place; for the pores are seated on the cellular or parenchymatous parts, while the lymphatic hairs constantly spring from the vascular or ligneous parts. Yet the hairs, he adds, are al-

most always placed on the same surface as the pores; and hence the superior surface of the leaf, which is commonly destitute of pores, has in general but few or no hairs. The pores, as before stated, seem to be the true organs of transpiration; and the various kinds of hairs serve to protect the plant against the excess of solar light, against variations of temperature, against humidity, or sometimes against insects. (Organog. Vég. tome i. chap. 10.) Hairs, also, of various kinds are met with on the different parts of the flower. These often have the colour of the petals or other parts on which they grow, and are distinguishable from the true lymphatic hairs, though sometimes associated with them on the same organ.

With regard to the structure of these minute bodies, structure, little that is satisfactory can be said. They seem to originate either directly from the cuticle or from the cortical texture beneath it; but not often from the ligneous texture, except in those instances where they are very long and rigid, as in the awns of wheat. Du Hamel observes that almost all of them are implanted on small bodies, similar to the bulbs which give origin to the hairs of animals. (Phys. des Arbres, tome i. p. 183.) They commonly resemble simple filaments, but often appear like elongated cells threaded on one another, and, instead of terminating in a sharp point, end in a small papilla or utricle, which yields in many instances a viscid or oily matter, or sometimes a coloured liquor, which has led many to regard them as exercising a glandular function.

One species of these supposed glandular organs has been more particularly examined, and their fluid analyzed, by M. Deyeux, who gives the following account of it. Soon after the seeds of the chick-pea (cicer arietinum) Hairs of are sown, its first leaves are seen to be covered with hairs, chick-pea. at the extremity of each of which is a transparent globule, about the size of a small pin-head, consisting of a fluid matter. It abounds most in mid-day, when the air is warm and dry, and is scarcely perceptible at night, or when the air is cold and moist; after rain, indeed, it does not again appear for two or three days. When these fluid globules were removed in a dry day by blotting-paper, they soon reappeared; they were acid to the taste, reddened litmus paper, and caused an effervescence in carbonate of potash when brought in contact with it. He regarded them as composed of oxalic acid, the properties of which they precisely resembled. (Mém. de l'Inst. Nat. tome i. p. 157.)

It is not easy to discriminate between some of the harder Definition; species of hairs, described in the former article, and those to which the appellation of prickles (aculei) has been assigned. They are defined by Du Hamel to be excrescences, often hard, and always terminated by a sharp point, which are developed with the other productions of plants, but are not inclosed in particular buds; so that they may for the most part be regarded as hard and solid hairs. They spring equally from the stem, the branch-position; es, the petioles of the leaves, and also from the leaves themselves in various plants; and in the chesnut and some others they are seen to cover the fruit. They are frequently straight, but in the rose and many others are curved at the point, as in fig. 18, Plate XLI.; and, according to Malpighi, possess sometimes in this plant a little head, which yields a viscid fluid.

Regarding their structure, Grew remarked that they structure. were connected only with the skin or the bark, and he therefore named them cortical, to distinguish them from thorns properly so called, such as those of the hawthorn,

Common Textures. which spring from the wood, and which he denominates ligoneous. These latter, he adds, always ascend, while the cortical thorns commonly point downwards. (Anat. of Plants, p. 33.) In proof of their origin from the bark, Du Hamel remarks, that if, after maceration in boiling water, the bark of such plants be stripped off, all the prickles come away with it, and leave not the smallest impression on the wood, nor even on the more interior layers of the bark itself. When a section also is made of the branch and prickle, as in fig. 18, Plate XLI., the wood y and the pith z are both seen to have no connection with the prickle, but the inner layer x of the bark is interposed between the base of the prickle and the wood. The prickle does not, however, spring from the skin, for it is formed of many layers like the bark. As the parts become more solid, it is less freely supplied with juice, and therefore hardens and turns brown. (Phys. des Arbres, tome i. p. 188.)

Prickle of the nettle. In the nettle (urtica dioica), Malpighi states, that beside the common prickles on the leaves, there are among them others of a different description. They possess more of a ligneous character, are hollow internally, and contain a juice which, when it gains admission beneath the skin, excites itching and tumour. (Anat. Plantar. p. 137.) Dr Hooke had previously given a much more minute account of the sting of this plant. Almost every part of it, says he, is covered with prickles like sharp needles. Each prickle consists of two parts, very different in shape and quality from one another; one is shaped much like a round bodkin, is very hard and stiff, exceedingly transparent and clear, and hollow from top to bottom. When this bodkin is thrust into the skin, it does not at all bend; but a certain liquor is then seen to move up and down in it, rising towards the top, when the point is pressed down on the base. This base is formed by a little bag, is more pliable than the bodkin part, and within it is a cellular structure, which contains a thin transparent liquor (see fig. 19, Plate XLI.): it is this liquor that rises in the tube, and, being deposited beneath the skin after it is punctured, excites the irritation that succeeds. (Micrographia, p. 142.)

The true thorn (spina) derives its origin, as Grew remarked, from the ligneous part of the plant, and in some plants is produced from the degeneration of some other organ, as of the leaf, or even the branch itself: we shall therefore defer the further consideration of its structure, until we come to treat of that of the branch.

Ambiguity of the term gland. Perhaps in the whole science of anatomy there is no word that has been employed with such latitude of signification, and is therefore exposed to so much ambiguity, as the term gland. In animal anatomy it was doubtless used at first to denominate certain organs, from the external resemblance which they bore to certain fruits or seeds; and in that sense it is still employed on several occasions. Afterwards it was understood to signify not so much the external form as the internal organization, and was considered to express a certain structure, by which alone the function of secretion could be exercised; but it is well observed by Dr Thomson, in his valuable work on Inflammation, that "the definition of a secreting glandular part must be taken from its function, and not from its structure; for nothing can be more various than the internal structure of those organs that are denominated glandular secreting organs: they consist sometimes of convoluted vessels, sometimes of follicles or small hollow bags, and sometimes of transparent membranes, in which neither convoluted vessels nor mucous follicles can be perceived."

(Lectures on Inflammation, p. 318.) Besides these more simple structures, it is well known that most of the internal viscera are likewise denominated glands, though differing in all their characters from those just mentioned.

The ambiguity which thus prevails in animal anatomy, in relation to the use of the term gland, has been increased tenfold in the applications that have been made of it to the organs of vegetables. It is justly observed by M. de Candolle, in reference to this subject, that the numerous approximations in structure between vegetables and animals have often promoted our researches into the former, but have sometimes led physiologists astray, and introduced into the language of botany many inexact expressions. In animal anatomy the term gland is understood to express some organ that exercises a secretory function; but in vegetable anatomy this term has often been applied to bodies that are not known to be real secretory organs. Thus the cells of the cellular tissue, which frequently contain resinous or oily matter, have been sometimes named cellular glands; the little globules or utricles at the extremities of the hairs on the edges of leaves, utricular glands; the small organs formed by the pores on the leaf, cortical or miliaery glands; certain fleshy tubercles on the leaves, urceolar glands; and the little scales that cover the fructification in ferns, scaliform glands. The nectarium of the flower commonly contains a sweet juice, and is therefore deemed a gland; but Linnaeus, with his usual disregard both of the structure and function of organs, considers as a nectary, not only the body which may secrete, but any other that may serve as a receptacle of the secretion; and indeed is said to comprehend under this term all those bodies which have no resemblance to the other parts of the flower, in whatever variety of form they may appear, or whatever purpose they may serve. (Wendow's Principles of Botany, p. 87.) In some other instances the term gland has been used, not to express the secreting organ itself, nor even the receptacle of the secretion, but the solid excreted matter on the surface of certain leaves; and others consider hairs, and every other protuberance that projects from the surface, and contains a fluid different from the common sap, as entitled to the distinctive appellation of gland.

Amid such diversity of opinion concerning the structure, position, and function of these minute organs, and in such vagueness in the methods employed to characterize them, it is extremely difficult to define their true nature, or declare the principle on which this definition should proceed. The mere existence of a fluid, distinct from the common sap in any organ, cannot be considered as bestowing on it the title of gland, otherwise the greater portion of some plants would come to be regarded as glandular: those varieties of structure which exercise no secretory function may also be excluded from the list of glands; and so likewise the hairs of plants, though containing peculiar fluids, may be excluded, since these peculiarities appear to arise frequently from circumstances foreign to the action of the organ itself; and even if they do not, some specific variation of the general name they bear is preferable to the employment of so ambiguous a word as gland. But where any organ is distinct from the common textures of the vegetable, and by the peculiarity of its structure is fitted to produce those changes on the vegetable fluids which we name secretion, it may be deemed a secreting organ. This secretory function, however, may sometimes be exercised, as in animal bodies, by membranous surfaces, and sometimes by small isolated bodies, to which, perhaps, may properly belong the denomination of glands.

But even though this method of defining glands were

lopted, it still is a matter of no small difficulty to distinguish their species by appropriate appellations. In animal anatomy no settled rule obtains; but the name of the and is assigned from some accidental circumstance of situation, figure, use, &c. In vegetable anatomy the botanist, regarding glands only as aiding the discrimination of species, refers commonly to their situation, and speaks of foliaceous, stipular, or petiolar glands, according as they happen to be seated on the leaves, the stipules, or the stoles. The anatomist imposes names according to their forms, as they chance most to resemble a globule, an acicle, or some other figure; and the physiologist is chiefly directed by ideas which indicate their functions, distinguishing them into mucous, oily, resinous, or nectariferous glands, according to the nature of the fluid they furnish. Of these different modes, that which proceeds from the apparent form, where it can be discovered, seems the most precise; but as this cannot always be accomplished, the situation of the organ, or the nature of the secreted fluid, must occasionally be had recourse to.

Of these bodies it is to be remarked that they differ in the respect from most of the corresponding organs in animals, almost all of them being seated on the external parts of the plant, like several of the more simple glandular bodies in animals. This arises from the greater simplicity of vegetable organization, particularly as it regards the sorbent system, the mode of growth, and the permanence of the organs produced; whence it happens that the living parts of aged perennial plants are situated only or near the surface; and it is only in such parts that true secretory functions can have place. Hence it is on ceculent stems, on leaves, flowers, and fruits, during the active state of vegetation, that the glandular functions of vegetables must be exercised; and which parts, therefore, are the appropriate seat of glands.

In some leaves a secretory function extends over a great part of the surface, as in some species of Cistus, of gar-maple, of larch, and others enumerated by Du Hamel, on which various collections of saccharine, gummy,

and resinous matter are found. (Phys. des Arbres, tome i. p. 183.) On the leaves of sage, Hooke mentions the occurrence of an infinite number of round balls resembling pearls, and which, says he, are nothing but a gummy exudation. (Micrographia, p. 142.) M. Guettard has described not fewer than seven species of glandular bodies on the leaves of different plants, to which he assigned names, chiefly from the appearance of their form; these are the miliary, the vesicular, the squamous, the globular, the lenticular, the utricular, and the urceolar glands. Of these reputed species, those called miliary are no longer held to be glands, but cuticular pores; and the squamous species is found to be identical with the thin scale that covers the fructification of ferns. Others add to this list the organ called nectary; but the very vague notions entertained of its nature and use altogether preclude the possibility of assigning to it any precise anatomical character.

Other writers have proposed to reduce all the bodies called glands to two classes, the cellular and the vascular, according as they conceive them to be formed of cellular tissue simply, or of this tissue and vessels combined. But such an arrangement would lead us, in some instances, to confound the mere receptacles of secreted fluids with the organs that secrete them,—would bestow a secretory function on organs considered by these writers to be non-vascular,—and convert the entire cellular tissue of the plant into a simple glandular body. The glands of plants may indeed possess the form and size of cells; but they are not, like cells, close cavities; they resemble more the follicles and mucous glands of the animal system, and, from the nature of their function, must always be regarded as vascular.

Of organs so minute, and so very imperfectly known and characterized, nothing can be attempted in the way of anatomical demonstration. This, however, is the less to be regretted, as the glandular system in plants appears in general to be of much less consequence in the vegetable than it is in the animal economy.

PART II.

THE ANATOMY OF THE INDIVIDUAL MEMBERS AND ORGANS OF VEGETABLES.

In the preceding Part we have described, in general, the nature of the elementary organs and common textures that compose the entire plant; and we come now to the second division of our subject, viz. the description of individual members and organs. Following the method of review, we shall first exhibit the anatomy of the seed, and the changes of form and structure displayed in its evolution. We shall next treat of the mature plant, and exhibit a few of the more remarkable varieties of structure observed in its several members, as the trunk, the branch, and the root. The organs that originate from these members, as seeds, leaves, flowers, and fruits, will then be duly noticed; and we shall terminate our descriptions by a brief exhibition of the formation and structure of the vegetable ovum in its progress towards the state of the perfect seed.

The limits within which, in a work like the present, we are necessarily circumscribed, will render our view of these individual structures, when compared with the immense variety that obtains in nature, exceedingly imperfect; but our purpose will be accomplished if we succeed in exhibiting a correct and tolerably comprehensive outline of the great features of vegetable organization, as they are displayed in the individual parts and organs of the more perfect plants.

CHAP. I.

THE ANATOMY OF SEEDS.

SECTION I.

The seed or egg of vegetables (semen vel ovum) is formed at the base of the pistil of the flower, in an organ called the ovary (ovarium), hereafter to be described. This organ takes the name of pericarp when the seed is mature, and is then said to be formed of three parts, enveloping each other. The outer membrane has been called epicarp, and the inner one endocarp, between which the thicker portion, called sarcocarp, is placed. In fleshy fruits, as the peach and apple, the sarcocarp is very thick, and contains the matter which nourishes the seed. Interiorly, the pericarp is either simple or separated by one or more partitions into cells, which contain the seeds. The seeds accordingly communicate with or are attached to the pericarp, at the point of the seed named hilum, from which point the ves-

Of the Seed. sels pass through the inner membrane of the pericarp, and are continued to its fleshy portion, which supplies the materials for its growth. These vessels form the umbilical cord (funiculus umbilicalis); but sometimes, it is said, this cord, from its extreme tenuity, or implication with other organs, cannot be discovered. When the seed has attained to maturity, the umbilical cord dries up and breaks, and the pericarp opens in various ways in different plants, to permit the escape of the seed.

Numbers of seeds: The seeds of different plants exhibit the greatest diversity in number, size, and figure. Sometimes they are few, in other instances very numerous. In one plant of white poppy Grew reckoned 32,000 seeds, and on the spike of a species of typha he numbered 40,176 seeds; so that, upon the three spikes which one stalk of this plant bears, there are every year produced more than 120,000 seeds. (Anat. of Plants, p. 198.) In figure the varieties in seeds are so numerous as to baffle description; and, with respect to size, many are so minute as not to be visible to the naked eye, and others so large as to reach several pounds in weight.

their size and figure. Umbilicus of the seed. The part at which the seed has separated from the ovary is indicated by a small mark or scar, called by Malpighi fenestra, by Linnæus hilum, and by Gærtner umbilicus. In some seeds this scar is of considerable extent, and is the only mark that is visible; in other instances there seems to be present a foramen in addition to the scar. All seeds, says Grew, have their outer coats open, either by a particular aperture, or by the breaking off of the cord, or by the entrance of the cord into the substance of the seed, as in those which have a shelly or stony covering. In the bean this aperture is placed on the side, in the chesnut on the top, in the gourd at the bottom; and in each case the point of the radicle is opposed to the aperture, and first pushes forth through it. (Anat. of Plants, book i.) In many seeds, however, the radicle is not thus opposed to the umbilicus, but, according to Gærtner, is variously placed with regard to it; and in the seeds of the apple and pear, Malpighi considers no proper umbilical aperture to exist, but only an hiatus to be formed by the relaxation of the tunics, through which the radicle makes its way. (Anat. Plantar. p. 9.)

Varieties in its form. There can be no doubt that, in every case, a connection subsisted between the seed and the ovary during the formation of the seed; but this, in different examples, may have much varied. In all cases an umbilical cord must have existed, and appears, in many examples, from inspection of mature seeds, to have been the only visible medium of connection. In the bean, however, besides the umbilical cord there are marks of a connection also between the coats of the seed and those of the ovary that contained it, so that the scar or cicatrix, in that and similar cases, is distinct from the umbilical aperture. The scar is the mark left by the separation of the tunics continued from the ovary; the foramen is the aperture produced by the separation of the umbilical cord. It will be afterwards stated that the outer coat of the seed appears always to have originated from the inner coat of the ovary, so that it forms a sheath about the umbilical cord. If, therefore, as in the bean, these coats are thick, and the umbilical cord short, then traces of the separation, both of the coats and of the cord, will remain on the seed; if, on the other hand, the coats are thin, and the cord more elongated, this latter will be closely invested by the former, and the umbilical scar will present the appearance only of a simple aperture, as is common to many seeds.

Opinion of M. Turpin. A late writer, M. Turpin, regards the umbilical scar as consisting of three parts, viz. the proper cicatrix itself, in

the centre of which he describes an aperture through which the nutrient vessels passed to the embryo; and near to it another smaller aperture, through which he believes the spermatid vessels to have passed. The former he calls the omphalode, the latter the micropyle, and declares that he observed it in more than 1200 seeds. (An. du Mus. d'Hist. Nat. tome vii. p. 199.)

From the situation of the umbilicus, the several parts or regions of the seed have been defined. They are six in number. The part where the umbilicus itself is placed is termed the basis, and the point at the opposite extremity the vertex of the seed; the upper or back part is named the dorsum or back, and the part opposite to it the venter or belly; while the two lateral portions are called the sides (latera). The point where the umbilical cord is inserted into the inner coat has been named the inter-umbilical umbilicus. This point usually coincides with that of the external, but, from a change in the relative position of the parts during their formation, this coincidence is not always to be observed. (Gærtner de Fructibus et Seminibus Plantarum, vol. i.)

When examined in its mature state, the seed is found to be composed of certain coats or tunics, which inclose a kernel or nucleus, that also consists of several distinct parts. At an early period of growth, while the parts are still green and succulent, two coats are easily distinguished, as in the transverse section of the bean (fig. 30, Plate XXXVIII.), in which the inner coat a appears much thicker than the outer, and the radicle b is seen rising through it.

When these coats are stripped off, the parts which form the nucleus are brought into view. They consist, in the bean and most other seeds, of two distinct parts, the lobes or cotyledons, as they have been called, and the radicle and plume. These several parts can be seen only by separating the two lobes from each other, as is done in fig. 31, where the letters c c denote the cotyledons, d the radicle, and e the plume. Such seeds as have thus two cotyledons are named dicotyledonous.

In many seeds, however, the part called cotyledon is single, and often bears but a small proportion to the entire bulk of the seed. Seeds which have thus but one cotyledon are named monocotyledonous; and to this division the seeds of wheat, of barley, and of all the grasses belong.

Many of the lower tribes of plants are entirely destitute of a cotyledon, and are called ACOTYLEDONS. Some writers assert that some seeds have more than two cotyledons, and such seeds they have denominated POLY-COTYLEDONS; but others, again, choose to consider these appearances not as distinct cotyledons, but only as deep fissures, or divisions into two primary lobes; and hence conclude, that all seeds having a cotyledon may be classed under the two divisions of mono and di-cotyledons.

ART. II.—Description and Structure of the Coats of Seeds.

Having given this general view of the several parts that compose the seed, we proceed now to a more particular description of their structure; and, as the tunics come first into view, we shall begin with them. These tunics, in some seeds, are two; in others, three in number. By Grew they were named coats or covers; by Malpighi, secundina; and by Gærtner, testa and membrana interna. We shall speak of them in the familiar terms of the outer, the inner, and the middle coats or tunics.

The outer coat or testa of Gærtner is described as a constant and essential part of the seed. It existed before the period of fecundation, and is sometimes the only apparent

covering possessed by the mature seed. Some seeds have indeed been considered to possess no tunic whatever, and have therefore, says Gærtner, been named acoea; but in such seeds there existed a coat before they arrived at maturity, and its apparent absence has been inferred from its extreme thinness, or its condensation with the sides of the surrounding ovary. (De Fructib. et Seminib. Plantar. vol. i. p. 132.) A distinguished botanist, however, Mr Brown, is said to have discovered two examples of seeds absolutely destitute of a covering, from their first appearance to their state of maturity. (Thomson's Annals of Philosophy, vol. i. p. 310.)

In different seeds this tunic possesses a very different structure, being in some thin and membranous; in others of a spongy or fleshy nature; and in others, again, it approaches to the consistence of leather or bone. But how various soever in this respect, it is always an entire tunic, and has no aperture but that of the umbilical foramen. Its colour is usually deeper than that of the other parts of the seed; and in this particular it presents every possible variety. It has rarely any connection with the nucleus, except in some monocotyledons. (Gærtner, vol. i. cap. 9.) With regard to its origin, Malpighi describes it, in its earliest state in the almond, as derived from the ovary itself, being composed of reticulated vessels, which spring from the surrounding organ. In other instances it is thicker, and is distinctly seen to be cellular as well as vascular; and in the bean and pea, little tubes are said by Malpighi to originate from the cells, and terminate by open mouths on the surface. (Anat. Plantar. p. 9.)

Both Malpighi and Grew discovered, in some instances, a very thin membrane to cover this outer coat, which, according to Gærtner, may be found in most seeds, if the parts be scrupulously examined. Its structure is sometimes membranous, often downy, and sometimes mucilaginous; it possesses occasionally considerable thickness, and at other times is a mere pellicle, and thence has been named pellicula. It covers the whole seed, and does not ever separate spontaneously from it. (Gærtner de Fructib. &c. vol. i. cap. 9.)

Besides this pellicle, another fine tunic named arillus is sometimes observed on the surface of the seed, as in the seed of euphorbia. It originates from the umbilical cord at the base, and extends more or less completely over the body of the seed. Its structure is very various, being sometimes soft and pulpy, at others thin and membranous, and in others forming a husky covering. It forms, in some instances, only a loose and partial covering; and in others it invests the seed so closely and completely, that it can scarcely be distinguished from the outer coat itself. Both in figure and colour it often varies greatly, but, like the pellicle before described, it is regarded rather as an accessory than a necessary integument.

To the exterior coat of the seed various appendages are sometimes attached, as down, wings, spines, hooks, all designed either as a defence to seeds or to facilitate their dispersion. They are distinguished and described by the botanist, but are in general of too fine a texture to be made the subject of anatomical demonstration.

The inner tunic (membrana interna of Gærtner) is a common but not constant part of the mature seed. It appears sometimes to be wanting, when in reality it is present. During the formation of the seed it is frequently so extenuated, or coalesces so completely with the outer coat, that it cannot be properly distinguished. In its earlier state it is represented by Grew as a very spongy and succulent body, and as thick and bulky as one of the lobes itself; but it dries and shrinks up as the seed approaches maturity, so that it is sometimes scarcely dis-

cernible. (Anat. of Plants, p. 47.) In the seeds of most plants it closely invests the nucleus, but is easily separable from the outer tunic. In those of the Gramineæ, where the bulk of the seed consists almost entirely of unorganized matter, no separation of this tunic from the contained parts occurs; but its inner surface is formed into a cellular tissue, in the cells of which the nutrient matter is lodged. In other instances the inner surface is prolonged into processes, which penetrate into the nucleus, and intersect it in various directions.

This inner tunic does not, like the former, exist before fecundation, but is formed subsequently to it. It is composed of vessels and cellular tissue. The cells are commonly larger than those of the outer coat. It has no aperture, says Gærtner, not even an umbilical one; but resembles a shut sac, over whose external surface the umbilical vessels creep, and open, in an insensible manner, within its cavity. (De Fructib. &c. cap. 9.) The distribution of the vessels throughout the whole of this coat Grew compares to that in the leaf.

Beneath this inner tunic Grew describes another fine membrane, which immediately invests the lobes or cotyledons of the seed. In the bean it is exquisitely thin, and so firmly continuous with the lobes, that some dexterity is required to accomplish its separation. It is spread not only over the convex surface of the lobes, but also over the inner or flat surfaces, where they are contiguous, extending likewise over the radicle and plume, and so over the whole nucleus of the seed. It does not, like the other tunics, cease to grow in germination, but is augmented and grows with the organic parts. (Anat. of Plants, b. i. ch. 1.) This tunic may be regarded either as a covering to the nucleus, or as an actual portion of it, as Gærtner, who speaks only of two coats, seems to have considered it. It appears, however, more proper to regard it as the inmost coat, and thus to conclude, with Grew, that the covers in most seeds are three. (Ibid. b. iv. ch. 3.) The coat last described must in that case be considered as the middle tunic, being situated between the outer one and that which is continuous with the nucleus. In many seeds, however, only two coats are distinctly visible.

The foregoing tunics not only contain the nutrient matter, but afford a mechanical protection to the organic parts, but seem fitted also, by their chemical constitution, to resist the operation of agents that might otherwise effect their decomposition, and that of the nucleus they inclose. From the experiments of Fourcroy and Vauquelin, on the tunics of certain seeds that grow in marshy situations, it appears that, besides the usual ingredients of vegetable substances, there exists in them a compound, formed by a combination of tannin with a peculiar matter of an animal nature, in union with a vegetable acid. This combination of tannin with the matter just mentioned renders these tunics insoluble in water, and enables them to resist putrefaction, although buried for long periods in the moist earth. The tunics of those seeds which do not possess this chemical constitution may, it is added, by their ligneous or horny texture, or by the oily matter with which they are penetrated, present similar obstacles to the action of decomposing agents. (An. du Mus. d'Hist. Nat. tome xv. p. 77.)

ART. III.—Description and Structure of the Nucleus of the Seed.

We proceed next to describe the parts contained within the above-mentioned coats or tunics, and which constitute the nucleus of the seed. These parts, as already observed, consist of the radicle, the plume, and cotyledons, together, in most instances, with the nutrient mat-

Of the Seed. ter destined to support their future growth. The three former bodies are completely organized, but the nutrient matter is wholly inorganic, varies greatly in quantity and proportion in different seeds, and is very variously situated with respect to the organized parts.

These parts, which in the progress of their evolution give birth to the new vegetable, derive their visible origin from a medullary point, that succeeds to the act of fecundation. By some these organized parts have been called the corculum, by others embryo, by others factus, and by others plantula seminalis. The primary point or particle from which they originate may with propriety, says Gærtner, be termed corculum, since it is the source and seat of vegetable life, and from it the whole vascular system of the embryo proceeds. In some instances this corculum increases so little as to be scarcely visible even in the mature seed, or exhibits only a palish spot, which has been termed cicatricula; in others it forms a roundish radicle, whose apex is free and rises above the nucleus, but whose base is firmly connected with it; in others, again, the corculum is still more disengaged, enlarging at each extremity, and producing at one end the radicle, and separating at the other into the two lobes called cotyledons, between which the first bud or plume of the future plant is situated. From this varying growth of the corculum, an embryo, more or less perfect, is produced. When the embryo presents only a mere germinating point, it is styled imperfect; when it exhibits a simple radicle, it is deemed incomplete; when it possesses both radicle and cotyledon, it is considered perfect; and when it consists of radicle, cotyledon, and plume, it is pronounced complete. (Gærtner de Fructib. &c. vol. i. cap. 13.)

Embryo. In the mature seeds of the less perfect plants the embryo is altogether invisible until after germination, and even in many other instances its characters cannot be accurately traced. Its general figure is determined by that of the radicle and cotyledons, and is exceedingly various in different seeds. In size it ranges from a minute point to that of a body of considerable magnitude. In consistence it is almost always soft and herbaceous, but its radicle possesses sometimes a ligneous hardness. No seed contains more than one embryo, except in cases of superfétation, of which Gærtner saw one instance in Pinus Cembra, a seed of which contained two embryos within one and the same cavity. (De Fructib. &c. vol. i. p. 168.) Malpighi also records a similar occurrence in a seed of Prunus Armeniaca; and the seeds of some Gramineæ, as will afterwards be shown, are capable of evolving an indefinite number of embryos. Every complete embryo is said to consist of three distinct parts beside cotyledons. These are the radicle, the stem, and the plume.

Its radicle. The radicle (radicula), called rostellum by Linnæus, is the most constant part of the embryo, being found in some seeds in which no other trace of that body can be discovered. In the seeds, however, of the less perfect plants, and even in those of some monocotyledons, no radicle is visible antecedent to germination. In some rare instances among dicotyledons, as in Nelumbo nucifera, no radicle exists; but in germination the stem first rises upward, and afterwards emits rootlets from its sides. (An. du Mus. d'Hist. Nat. tome xiii.)

The size of the radicle is very various, and so also is its figure, being either conical, cylindrical, filiform, or tubercular, &c. It always, says Gærtner, occurs solitary, except in secale, triticum, and hordeum, to which alone, of all known seeds, three, four, or six radicles, properly formed, and distinct from each other, are furnished to each embryo. (De Fructib. &c. p. 169.) This plurality

of radicles in the cerealia had before been remarked by Olfæ Malpighi. M. du Hamel describes the seed of misletoe (viscum album) as emitting numerous radicles like those of wheat. (Mém. de l'Acad. des Sciences, 1740.)

The stem (scapus) of the embryo is a continuation of its radicle, and connects it with the plume. It is frequently wanting altogether, nor, when it is present, can we fix precisely on the point where the radicle ends and the stem begins. What is called stem descends frequently into the earth, and becomes a true root; so that every part of the embryo situated beneath the cotyledons might without impropriety be denominated radicle. The place of junction between the radicle and stem was called by Grew the coarcture, from its presenting often an evident degree of contraction; but M. Bonnet and others have given it the more appropriate name of the neck (collum) of the seed.

The plume (plumula) is the first bud of the new plant. In seeds that possess but one cotyledon it is very generally wanting; and even in those which have two cotyledons it is not unfrequently absent, or is at least concealed within the stem. In most of the latter sort of seeds, however, the plume is met with. It is placed on the top of the stem or radicle, and lies between the cotyledons, by which it is variously compressed and folded on itself. In the greater number of seeds it is not entire, but at its free end is divided into several pieces, all closely couché together, like feathers in a bunch, and thence called the plume by Grew. In different seeds its several little leaves vary much in figure, size, and number. The structure both of the radicle and plume will be most advantageously displayed in connection with that of the cotyledons.

Of the organized parts of the seed, the organs called by The Grew lobes or dissimilar leaves, by Malpighi seminal leaves or cotyledons, remain to be described. The cotyledons derive their origin from the embryo itself, of which they constitute a part. The seeds, however, of some tribes of vegetables, as before remarked, are held not to possess these organs; and in many others the mass of nutrient matter has been confounded with them. When present they are either simple or divided. The simple cotyledon is formed by the mere extension of the corculum, and is in truth scarcely distinguishable from the stem itself. The double or conjugate cotyledons are produced by fissures, which usually divide that part of the embryo that is opposed to the radicle into two or more equal portions or lobes. These lobes have at first the appearance of mere tubercles, and in many seeds they retain this form unchanged; but in others they gradually expand into lamiæ or plates, which augment in size, and finally exhibit the proper form of cotyledons. This form is very various, as likewise is the size of these organs. Sometimes they are so small as to be scarcely visible, and sometimes so large as to form the chief portion of the seed. Their substance is either thin, or thick, or turgid. Their colour is commonly white, but sometimes yellowish, purple, or green, the colour into which they all pass during germination.

Concerning the structure of the cotyledons, it may be said that, in the more perfectly developed seeds, they are formed of cellular tissue, through which vessels are everywhere distributed; and, as we have already remarked, they are everywhere covered by a fine pellicle or coat, which prevents alike their adherence to the plume and to each other. This cellular structure of the cotyledon is well displayed by Grew (fig. 32, Plate XXXVIII.) in a slice of the cotyledon of the recent bean; and it is easily seen in a thin slice of almost any mature seed, if it be held

against the light after it has been soaked in water. This cellular structure extends into the radicle and plume, but in much smaller proportion, constituting, according to Grew, about three fifths of the plume, five sevenths of the radicle, and nine tenths of the cotyledon.

Through all the organs that compose the nucleus vessels are distributed, by the medium of which a general communication is established among them. This vascular system is likewise exhibited by Grew in the dissection of a bean (fig. 34, Plate XXXVIII.), in which the vessels are seen to branch off on each side from the radicle, and spread themselves by innumerable ramifications through the cotyledons. From the radicle vessels also pass upwards to the plume. These vessels of the radicle are visible when a transverse section is made through it, as in fig. 33, a, Plate XXXVIII., in which they are seen to occupy the middle of that body. When the section is made higher up at the neck of the embryo, as in the same figure, b, then the central trunk, surrounded by several smaller fasciculi of vessels, passing to the different parts of the plume, is still more clearly exposed. In many seeds, however, the organized parts are so small that their general structure cannot be traced, except during the progress of their germination. We shall therefore defer the description of them till we come to treat of their evolution, and shall then also go more fully into the structure of the parts just mentioned.

Within the cells of the cotyledon, in many dicotyledonous seeds, the nutrient matter destined to support the future growth of the embryo is entirely contained. In other instances this matter is only in part received into those organs; and in the Gramineæ, and other monocotyledons, it is often placed almost entirely exterior to the cotyledon. This matter is produced from a clear liquor that is secreted in the tunic during the formation of the seed. To this liquor Grew gave the name of albumen, from its likeness, not only in appearance, but, as he conceived, in use also, to the white of egg in animals. By Malpighi this matter, considered in connection with the tissue that contains it, is often called the flesh (caro) of the seed; and from its being sometimes situated around the embryo, it has been denominated perisperm by M. Jussieu. We follow Grew and Gærtner in the use of the term albumen, meaning to express thereby, not the primary animal compound to which chemists have of late assigned that term, and which is found but in few vegetables; but that compound substance which, whatever be its situation, quantity, or colour, constitutes the nutrient matter of the seed.

This albumen is a very constant part of the mature seed, but its proportion in some seeds is so extremely small, that the seeds in which it occurs have been termed exalbuminous; and in a few instances it seems to be entirely wanting. Its quantity, situation, and figure, in different seeds, are subject to very great variation. In the seeds of the Gramineæ, where the embryo acquires only a very small size, the albumen constitutes almost the entire bulk of the seed, and is placed wholly exterior to the embryo. In the Leguminosæ, on the other hand, the embryo is more completely developed, and the whole of the albuminous matter is contained within the cotyledons. In beet (beta), and many others, the albumen is partly received into the cotyledons, and lies in part exterior to them; and where this occurs the embryo sometimes encircles the albumen, and is sometimes encircled by it. In Rheum the embryo is placed in the centre of the albumen, in Rumex and some others it is applied on the side of it, in Atriplex the long cylindrical embryo surrounds the albumen, in Böerhaavia the embryo and its

cotyledons cover entirely the granulated substance of the albumen. (M. Jussieu, An. du Mus. d'Hist. Nat. tome v. p. 224.) In the onion (allium cepa) the embryo makes several curves within the substance of the albumen, and in dodder (cuscuta) it is twisted around it in a spiral form; so that the relative positions of the embryo and albumen, as well as their quantity, proportion, and figure, are subject to endless variation.

But however much in these respects the albumen may vary, it is always contained within an organized structure, of wheat. Sometimes this structure is that of the cotyledon, as already exhibited in the bean (fig. 32, Plate XXXVIII.), the cells of which contain this albuminous matter. Where the albumen is placed exterior to the embryo, as in the seeds of wheat, it is nevertheless contained in a cellular tissue. This is exhibited in fig. 27, Plate XXXVIII., copied from Leewenhoek, in which cells of an hexagonal form are seen to be filled with the albuminous particles that constitute the white matter or flour of that seed. This mealy part of wheat he describes as consisting of minute globules, inclosed in a kind of membrane so exquisitely thin as scarcely to be observed, within which the globules are contained, as it were, in cells. The globules appeared to be of different sizes, not perfect spheres, but having an indentation on one part, which led him to suppose that they were not formed by simple accretion, but by some mode of growth, and that "the membranes which inclose them in cells must be provided with so many veins or vessels, that every particle of meal may have its separate vessel." He even conceived the globules themselves to be inclosed individually in a thin skin or shell; but this opinion he never brought to ocular demonstration. (Select Works by Hoole, vol. i. p. 169.) Similar observations on the albuminous part of wheat have since been given by Mirbel; and Kieser and others have delineated the globular particles contained in the cotyledonous cells of the bean and other seeds; so that whether the albumen be situated in the cotyledons, or be placed exterior to them, it is in every case contained in a similar and distinctly organized structure.

In consistence the albumen is said to be either fari- Varieties naceous, fleshy, or cartilaginous; and it may exist in of albumen. various intermediate states. The farinaceous kind is readily reduced to powder, and is dissolved by water into a viscous mass. The embryo is generally placed exterior to this species of albumen, as in the Gramineæ. The fleshy albumen is more frequent. It is softer than the former, and dissolves by water into a gelatinous mass. It is often entirely contained within the embryo and its cotyledons, and yields the thick oil that is expressed from many seeds. Lastly, the cartilaginous species has a horny consistence, is difficultly soluble in water, and not easily reduced to powder. The embryo is never placed exterior to it, and when it contains oil, this is usually very thin. (Gærtner de Fructib. &c. vol. i. cap. 10.)

In many seeds the albumen serves as a support and defence to the embryo, as well as for nutriment. If it be removed previous to germination, as was done by Mirbel (An. du Mus. d'Hist. Nat. tome xiii. p. 157) in the seed of the onion, and by Dr Yule (Wern. Trans. vol. i. p. 591) in different species of Gramineæ, the embryo, though planted in a rich soil, and carefully tended, grows but feebly, and for the most part dies.

Besides the albumen above described, Gærtner has received the use of the term vitellus, but employed it to designate a very different part from that to which it was originally applied by Grew. The latter made use of this term to designate the inorganic matter of the mature seed, which in the early stage of its production he called

Of the Seed. albumen (Anat. of Plants, book iv. chap. 3); but it is employed by Gærtner to indicate a small membranous body, which in many seeds is placed between the embryo and albumen, and is closely connected with the former, but separates easily from the latter. The figure of this small body is described as being very various in different seeds. It is said not to rise out of the earth during germination; but, like the albumen, seems destined to afford nutriment to the embryo. In the Gramineæ it represents a thin scale interposed between the albumen and embryo, to which, from its shield-like form, he gives the name of scutellum. (De Fructib. Plantar. vol. i. cap. 11.) There can be no doubt that this scutellum of Gærtner is the little "conglobate leaf" first observed in wheat by Malpighi, and which later writers have denominated the cotyledon of that seed. In the next section its form and situation will be clearly displayed.

Classification of seeds: All seeds have by some botanists been distinguished into such as possessed one or more cotyledons, and such as were entirely destitute of them. In treating of seeds under the two divisions of mono and di-cotyledons, we would not be understood to deny the existence of seeds that possess more than two.

their evolution observed. Some seeds are so extremely minute, that, until lately, their existence was not clearly ascertained; and it is only during their germination that their general form and character can be detected. In many others, the organized parts are so small as to be scarcely capable of demonstration, except by following the progressive changes of form they exhibit in their evolution. We propose, therefore, to select, from each of the two divisions of mono and di-cotyledonous seeds, an example or two of the successive appearances displayed in their evolution, which will, besides, form the best introduction to a knowledge of the structure of the mature plant.

Evolution of wheat. In most of the monocotyledonous seeds the cotyledon does not appear above the soil during germination, but is retained within the coats of the seed, and consequently undergoes but little alteration in size. As an example of the evolution of a monocotyledonous seed, we shall select that of wheat (triticum hibernum), because its development has been studied with great care, and, in common with some others of the same natural family, it exhibits some striking peculiarities, which add greatly to its productive powers. The successive appearances exhibited in its evolution have been given with great accuracy by Malpighi, who has carried its anatomy farther, in some points, than most of his successors. In the earlier stages of growth, some very accurate representations of it have also been given by M. Poiteau; and Dr Yule has likewise obliged us with some valuable observations. From these different authorities, confirmed generally by our own observations, we shall endeavour to present a concise view of the structure and evolution of this very important seed.

Description of wheat. If we take a grain of wheat, and examine its convex side, we observe, at its base, a small oblong body (fig. 35, Plate XXXVIII.) lying in a semicircular depression, which is well defined through the tunics that cover it. These tunics are two in number; an outer one, to which the chaffy filaments at the vertex of the seed are attached, and which readily separates when moistened; and an inner one, which everywhere adheres closely to the cellular tissue that contains the albumen. If these two

tunics be raised and thrown back, as is done in fig. 36, the little oblong body k, and its semilunar appendage i, placed behind it, are brought into view; and, together, they constitute the embryo.

Let next a vertical section of another seed be made in the direction of the furrow that runs along its flatter side, and let this section pass through the embryo, as is represented in fig. 37: we then observe the seed to be composed almost entirely of albumen k, with which the embryo l, consisting of minute convoluted leaves, is in close contact. The part of the embryo that is applied against the albumen is the cotyledon, which on that surface is convex, and on the opposite one concave.

In fig. 38, the entire embryo has been removed from its connection with the albumen, and a front view of it, considerably magnified, is there given, in which the letter m denotes the cotyledon, in the concavity of which the plume n is lodged, and o indicates the protuberances from which the radicles afterwards spring. If now this same embryo be reversed, as is done in fig. 39, then the convex back of the cotyledon only is seen, with the extremity of the principal radicle at the base. It is this side of the cotyledon that was applied against the albumen; and its polished surface, says M. Poiteau, proves that it nowhere adhered by any organic structure. Gærtner also remarks that the connection between these parts is not organic, but merely superficial,—an observation that is true as far as relates to the embryo itself and the albuminous matter, but not as applied to the tunics which envelope them; for, at the base of the seed, the inner membrane, which contains the albumen, appears to be continuous, as Leewenhoeck remarked, with that which covers the cotyledon, being reflected from the albumen over the cotyledon, much in the same way as the pleura and peritoneum, that line the sides of the great cavities in animal bodies, are reflected over the viscera they contain. Such are the appearances presented by this seed antecedent to germination: let us next follow it through the several stages of that process.

After a seed of this species has been in contact for 24 or 30 hours with the humidity necessary to its germination, its embryo becomes swollen, and, when removed from the other parts, and moderately magnified, presents the appearance exhibited in fig. 40. In this figure the radicle is rendered more protuberant, and the fine tunic that invests it has undergone an alteration, being changed from a smooth, opaque, and solid texture, to one that is villous, transparent, and cellular. A vertical section of the same embryo, as exhibited in the next figure (41), shows the elongation of the principal radicle p, which caused the protuberance below, and the sprouting of the two lateral radicles p p, which push forth more slowly on the sides. These three radicles soon force their way through the sac that envelopes them, which then forms sheaths around their origins. In the same figure, the letter q denotes the plume, consisting of several convoluted leaves, and resting on the cotyledon. In fig. 42 the appearance of the seed, in a stage a little more advanced, is exhibited. The plume r is now seen to have risen above the cotyledon s, and the three radicles, surrounded at their origins by their proper sheaths, have greatly increased in length, and innumerable capillary rootlets are emitted from their sides. (An. du Mus. d'Hist. Nat. tome xiii. p. 383.)

The daily appearances exhibited in the evolution of this seed, as previously given by Malpighi (Anat. Plantar. p. 103), accord well with the above representations of M. Poiteau; and he has noticed some additional particulars of considerable importance. On the first day of germination, he represents the plume of the embryo as beginning to

the seed, open, and the protuberances, which indicate the eruption of the three radicles, as beginning to form. The radicles at this period are completely enveloped in a membranous sac or involucre; and the body of the embryo is closely connected with a "conglobate farinaceous leaf, by which nutriment is administered." This conglobate leaf is the cotyledon before mentioned, and its connection with the radicle and plume is well shown by Malpighi. In fig. 43, t, he exhibits a front view of the radicle and plume, as they appear when separated from the cotyledon; and at the letter v of the same figure a back view of the same body is displayed, in which the letter x points to the mark or scar that denotes the place of separation. Malpighi believed these parts to be united with each other by a little node, hereafter to be described; but it is by the medium of vessels that this connection between the cotyledon and the other parts of the embryo is maintained; and by this route alone can the nutrient matter or albumen be conveyed through the cotyledon to the radicle and plume. To these vessels M. Bonnet gave the distinctive appellation of mammary: the union they form between the different parts of the embryo is so close, that at this part, says Gærtner, the cotyledon, and radicle, and plume, form one undivided body. (Gærtner de Fructib. Plantar. vol. i. p. 149.)

2 day. On the second day of germination the exterior tunic of the seed, according to Malpighi, gives way; the plume rises upward; the radicles do not as yet pierce their investing sac, but this sac is turgid with juice, and is covered exteriorly by a fine white down: the cotyledon also, at this period, is rendered moist.

3 day. During the third day the cotyledon is quite turgid with juice; the plume is much enlarged, and begins to look green; the three radicles have pierced the enveloping sac, and are everywhere thickly covered with hairs; and above the first two lateral radicles, two small protuberances, y, z, fig. 44, the origins of two more radicles, are now seen to emerge, while the sac that envelopes them is observed sensibly to waste.

4 day. When the third day has elapsed; the plume, inclosed in a fine transparent membrane, is still more elevated, and acquires a greenish colour: the protuberances of the two new radicles are more prominent, and the three former radicles have greatly augmented; the cotyledon is much softer, and, as if milky, yielding, when compressed, a white and sweetish liquor.

5 day. After the completion of the fourth day, the plume, continuing to ascend, pierces the membranous covering a (fig. 45), and pushes into day a permanent leaf, green and convoluted, around which the membrane forms a sheath. Inferiorly, the first three radicles have greatly extended, and the two others, b, b, are much increased: the outer coat of the seed now begins to lessen, but still contains a sweetish liquor. M. Poiteau gives a section of the entire plantule about this period of its growth, which agrees very exactly with the figure of Malpighi. In this section (fig. 46) the plume c is seen to have pierced the membrane d that formerly inclosed it; the albumen e is diminished; the cotyledon f retains its situation and form; and the five radicles g, g are nearly of a length, and covered with hairs.

6 day. About the sixth day the plantule, still invested by its sheath, begins to open and expand; the seminal tunics shrink, and the surface of the outer coat is corrugated. If these tunics are cut open, the cotyledon within is observed in some parts to be firmer than before, and has the appearance of a concave leaf; but in other parts it is more vascular and filled with juice, especially in that part near to the mammary vessels.

After the eleventh day these tunics still adhere to the Of the Seed. plantule, but appear much wasted, and the juice they contain is mixed with bubbles of air; while the stem forms, 11th day, ing many knots, and the radicles emitting innumerable rootlets, continually augment in size. Where the vegetation has been very active, the whole original contents of the seminal tunics are by this time exhausted, and, when compressed, they yield only a watery fluid.

After the lapse of a month, when the parts already developed are still farther advanced, new buds break out from the primary seat of growth and rise upward, and new radicles push forth and descend. So readily are these radicles produced, that sometimes, if the primary ones be removed, others in crowds spring forth; at the same time new buds or shoots, protected in their proper sheaths, arise from the same part, and surrounding the primary plantule, are borne upward with it. Of these appearances accurate delineations are given, and they may be observed in every field of growing wheat.

The foregoing descriptions of Malpighi are in general Error of very correct, and his figures, though somewhat rude, ex-Malpighi, hibit faithful delineations of the objects they are destined to represent. In one or two points, however, he has fallen into error, which, in the above statement of his opinions, to avoid confusion, we corrected as we went along. Thus, though he distinctly points out the "conglobate farinaceous leaf" as the organ by which nutriment is administered to the radicle and plume, he assigns to the sac that, in an early state, envelopes the radicles, the function of placental, and even gives to the exterior tunic of the seed the title of seminal leaf. The true cotyledon, however, which never in this seed is produced into a seminal leaf, is the little conglobate body above mentioned; and the common tunics of the seed have no title to the appellation of seminal leaves. To this cotyledon Gærtner, from its shield-like form, gave, as before observed, the name of scutellum. He held it to be characteristic of the Gramineæ, and analogous to the organ to which, in some other seeds, he gave the name of vitellus. (De Fructib. Plantar. vol. i. p. 139.) But later writers, as Jussieu, Smith, Brown, and Poiteau, have all restored to it its proper office of cotyledon.

M. Poiteau has gone even farther, and asserted the existence of a second cotyledon in this seed, and in the oat, which he describes as situated directly opposite to the former. (An. du Mus. d'Hist. Nat. tome xiii. p. 388.) In this instance, however, he has mistaken the rudiment of the second bud for a second cotyledon, as Dr Yule ascertained by "tracing the growth of this supposed cotyledon from its first becoming visible to its final development as a plant." (Werner. Transac. vol. i. p. 594.) M. Mirbel of Mirbel. considers the sac that invests the plantule to be the cotyledon of this seed, and this cotyledon to form the first ensheathing leaf. (An. du Mus. d'Hist. Nat. tome xiii. p. 148.) But, as already remarked, the cotyledon never in this seed rises out of the tunics; and, as Dr Yule observes, differs totally in situation, structure, and consistency, from the ensheathing leaf of the plantule.

A very remarkable peculiarity in these plants is their great reproductive power, as displayed in the indefinite number of new plants which we have seen to be evolved from one primary seed. Malpighi not only observed this peculiarity, but has described the structure from which it originates. He considered the radicle and plume of the embryo to be connected with the cotyledon, not by the mammary vessels, as we have stated, but by a little body which he called the umbilical node. In a section of the Peculiarity lower part of the stem of the plantule, made after the of struc- third day of germination, he delineates this node as situ- ture in wheat.

Of the Seed. ated at the junction of the radicle and plume, as represented by the letter k (fig. 48, Plate XXXVIII.); and describes it as solid exteriorly, and softer and more membranous within. If a section of the same part be made on Malpighi, the fourth day, as in fig. 47, the stem i of the plantule will be seen, says he, to spring from this node, from which also the radicles equally take their origin.

of Leeuwenhoek. This peculiar property was also observed by Leeuwenhoek, though he seems not clearly to have apprehended the nature of the organs from which it proceeded. In the embryo of wheat he describes three points, from which not only three distinct radicles spring, but they are also, he adds, "the beginnings of three several spires or stalks of wheat; so that from every grain of wheat (which is well worthy of observation) there will arise not merely a single stalk, but three distinct ones, which are formed in the seed itself." Select Works by Hoole, vol. i. p. 169; and in vol. ii. p. 289, are to be found similar observations on the seeds of oats, barley, and rye.

of Mirbel. By M. Mirbel, the umbilical node of Malpighi is considered as a fleshy knot (un nœud charnu), by the medium of which the plume and radicle are united. The lateral radicles which issue from it he regards as distinct in their nature from the primary one, and as resembling those which spring from knots in the stem; he therefore names them articular roots, les racines articulaires. (An. du Mus. d'Hist. Nat. tome xiii. p. 149.) According to Dr Yule, however, this fleshy knot is to be considered as a tuber, analogous to the tuberous substance interposed between the bulbs and roots of the Liliaceæ and other monocotyledonous tribes; and which is destined to produce an indefinite number of young plants, a greater or less number of which are subsequently evolved by the joint agency of the roots and leaves. The "articular roots" of M. Mirbel he regards as in reality young plants, the roots of the Gramineæ being invariably fibrous. It is by means of these lateral shoots and their tubers that bushes, consisting of from sixty to several hundred stems, are sometimes seen to originate from one seed.

The above important peculiarities in the germination of the seeds of the Gramineæ are very perspicuously displayed by Dr Yule in the three figures which we have copied from his Memoir. In fig. 49, Plate XXXVIII., Dr Yule represents the embryo of wheat as it appears when detached from the albumen, a short time after germination has commenced; the ascent of the plume covered with its membrane, and descent of the three primary radicles, which have pierced their containing sac, are clearly exhibited; and the letter k points to the little cotyledon placed at the junction of the two parts just mentioned. In fig. 50 the germination of the same seed is shown in a more advanced stage; the plume l has now risen to a considerable height, and pierced the investing membrane; and at m a second bud or plume (which M. Poiteau mistook for a second cotyledon) is seen to shoot from the tuber like the first. The letter n denotes the seminal tunics. At a still more advanced period, four young plants, o o o o, fig. 51, of the second month, with their sheaths in part withered, are seen to have sprung from the same part; but the two seminal tunics of the seed, exhausted of their contents, still remain attached, as indicated by the letter p. (Wernerian Trans. vol. i. p. 589.)

The description given above of the evolution of wheat is applicable, with little variation, to the seeds of all the cerealia. The seed of the oat emits from four to six radicles, all of which break through their enveloping sac at the same place, and thus appear to be contained in one sheath. Such too is the case with barley, the plume of

which extends beneath the seminal tunics, and pushes out of the vertex of the seed.

This peculiar constitution of the seeds above mentioned is attended with important advantages in their culture, and explains the source of their great productive power. A single grain of barley was observed by Du Hamel to have produced 200 ears, each of which yielded 24 grains; so that one single seed planted in a good soil has produced 4800 grains. Curtis and others, by transplantation of the several plantules of wheat, obtained still higher returns from single seeds. For the same reason, these plants are better enabled than others to resist the injurious effects of accident or disease. If a seed, says Dr Yule, be buried under a stone or lump of indurated clay, the seminal plantules cannot shoot upward; but stems are then sent off in a horizontal direction, until they can effect their escape upward. Sometimes it happens that a small insect (Musca pumilionis) deposits its egg in wheat, and the grub is lodged in the very centre of the stem, just above the root, by which the stem is invariably destroyed, and the root so materially injured as to prevent its throwing out fresh shoots on each side, or stocking itself, as the farmers term it. Nevertheless, the plants thus attacked are not permanently injured; for, in the instance where these depredations occurred, the crop of wheat was good, and the ears large and fine through the whole field; so that these injured plants, by the production of lateral shoots, yielded an abundant crop. (Lin. Trans. vol. ii. p. 76.)

In the germination of other monocotyledonous seeds a similar succession of phenomena present themselves, with the exception of those which relate to the multiplication of so many individuals from a single seed. In all, the radicle first shoots forth, and the plume soon follows; the cotyledon is commonly of small size, and is retained within the tunics. As the embryo grows, the albumen is taken up and conveyed through the cotyledon to the young plantule; and before the albumen is exhausted the embryo is enabled to draw its nutriment from the soil in which it grows.

In many instances it appears that the primary radicle of seeds, after a short time, becomes dry and falls off, and is replaced by a great number of secondary rootlets. M. Poiteau regards this last circumstance as common and peculiar to monocotyledonous seeds. He has remarked it in many hundred palms, not one of which had a descending or tap-root. No plant in the numerous family of the Liliaceæ is said to possess such a root. The radicle of the Cyperaceæ does not, perhaps, perish so soon; but it does not continue long. This premature and constant destruction of the radicle he considers as the cause of the bulbs and truncations which occur, particularly in the Liliaceæ; for the lateral roots not being capable of receiving all the descending sap, it collects at the lower part of the stem, and occasions these different enlargements. (An. du Mus. d'Hist. Nat. tome xiii. p. 392.)

The effects which thus succeed to the spontaneous destruction of the radicle occur partly in other plants, in which the first radicle is naturally permanent, if it be artificially removed. Du Hamel found that if the extremity of permanent radicles were cut off, lateral rootlets were produced; that even mechanical obstruction to the descent of the radicle frequently gave rise to divisions in it, and the production of these lateral rootlets. He ascertained also, by experiment, that roots extend invariably, not by an elongation of parts already formed, but by new matter added to their extremities; and hence it is that roots, whether ligneous or herbaceous, do not elongate if even the smallest portion of their extremity be cut off. (Phys. des Arbres, tome i. p. 83.) The results of observa-

lections on the growth of carrots in different soils, by Mr Knight, correspond with those of M. du Hamel.

Of the Structure of Dicotyledonous Seeds, as displayed in their Evolution.

From the greater number of seeds which have two cotyledons, the phenomena of their evolution may be expected to exhibit at least as great variety as those of the division last described. In different species they differ in this respect as much from one another as they do from monocotyledonous seeds. Some seeds of this class raise their cotyledons above ground during germination; in others these organs are wholly retained within the tunics. Of each of these modes of evolution we propose to give an example, selecting, as before, those seeds which have been most accurately observed; or which, by the forms they exhibit, seem best calculated to illustrate the general laws by which their evolution is accomplished.

In the seeds of this division the embryo is commonly much more completely developed than in those of the class last described, so that the several organs of the plantule become distinctly visible. The radicle and plume are readily distinguished, and the cotyledons are frequently so large as to form nearly the entire mass of the seed. Within the cotyledons the albuminous matter provided for the nutrition of the embryo during its evolution is often entirely contained; and these organs, as before remarked, rise sometimes out of the earth, increase greatly in size, and after a certain period decay. In other instances no increase in size nor elevation above the surface occurs, but, like the greater number of monocotyledons, they remain beneath the soil, and yield gradually their nutrient matter for the support of the embryo. Even in plants of the same natural order, the Papilionaceæ for example, some, as lupinus, says Dr Smith, raise their cotyledons into the air and light; while others, as lathyrus, retain them under ground, concealed within the tunics of the seed. As an example of the latter, we shall give from Malpighi an abridged account of the successive appearances exhibited by the common pea (pisum), which, in its evolution, approaches in some respects to that of the seeds last described.

The figure and size of this seed are familiar to every one. After being placed for a day in circumstances favourable to its germination, it is much increased in size; its outer coat is rendered softer, and becomes more white and thin; the umbilical aperture continues closed, but near to it an irregular opening or laceration is visible. If the outer coat be now stripped off, the nucleus comes into view. It is seen to consist of two distinct parts or lobes, which are the proper cotyledons of the seed. These cotyledons are closely invested by the inner tunic; externally they have a convex surface, but internally, where they are in contact, their surfaces are nearly plain. Between them, in a small depression formed in their substance, lies the plume: it is composed of a number of yellowish leaves folded on each other, and bent inward, and is united by a little curved stem to a small white conical body, the radicle. These appearances are exhibited in fig. 1, Plate XXXIX., in which one of the cotyledons has been removed, so that the inner surface of the other, together with the plume and radicle, is fully brought into view: the letter a denotes the cotyledon, b the plume, and c the radicle. This radicle at its neck, or point of junction with the stem, sends off on each side a little stalk or pedicle to each cotyledon. In the above figure one of these pedicles has been cut through, and

the other, that remains attached to the cotyledon, is concealed behind the plume. It is by these pedicles alone that the two cotyledons are connected with each other.

When the second day of germination is completed, the 2d day. cotyledons are rendered more tumid, the tunics give way, and the radicle begins to protrude. Soon after, the cotyledons separate a little, and become somewhat concave internally. After the third day the radicle has pushed out through the tunics: it is white, except at its point, which is more deeply coloured, and it emits on all sides fine capillary rootlets; the cotyledons are now farther separated, and by degrees the stem of the embryo, with its curved plume, is disclosed.

About the fifth day the stem d, fig. 2, mounts up-5th day. wards: it is white, and bears on its summit the plume e, still curved, and now becoming green: the stem now also begins to exhibit the marks of knots at particular parts: the radicle f is farther advanced, and small protuberances, the origins of future rootlets, appear on it: the cotyledons g retain their place, are turgid and solid, and still surrounded by the lacerated tunics.

At the close of the seventh day the plantule is much 7th day. more advanced: the knots on the stem h h, fig. 3, are quite distinct, and its apex is furnished with broad green leaves, but which are not as yet unfolded. The substance of the cotyledons is still solid, and when compressed yields a bitterish juice: the radicle is much elongated, and has emitted numerous rootlets.

After the ninth day the plantule is completely formed: 9th day. its stem i, fig. 4, is now erect, and the leaves of the plume k are expanded: the cotyledons l are reduced in size: and the radicle m, or root as it may now be termed, with its numerous rootlets, is greatly augmented. Every part of the plantule except the cotyledons continues daily to increase: at the end of a month these organs are still found to adhere; but are wrinkled, thin, and exhausted of their nutrient matter, with which the stem and other organs are filled.

The progress of evolution in those seeds that raise Evolution their cotyledons above the earth is exhibited by Malpighi of the in that of the gourd (Cucurbita). This seed is of an ob-gourd. long figure, and has a flattened form. It possesses three distinct coats or tunics: the outer one is thick, tough, and coriaceous; the middle one thin, membranous, and of a greenish colour; and the inmost is that transparent colourless pellicle that closely invests, and is inseparably connected with, the cotyledons of the seed.

After this seed has been made to imbibe moisture, the outer and middle tunics readily separate, and expose the nucleus, which is seen to consist of two leaf-like cotyledons, which have no connection with each other, except by the medium of the little conical body, or radicle, at their base. The size and figure of these cotyledons, and the situation of the radicle that connects them, are represented in fig. 5, Plate XXXIX. Their external surface exhibits to the naked eye prominent lines, which indicate a vascular structure, the vessels of which proceed from the radicle at their base. They are commonly five in number, and from their main fascicular trunks ramifications are produced, which, in their distribution, form a finely reticulated appearance over the whole organ. On their inner side the cotyledons are quite plain, and closely applied against each other, but, as already remarked, are nowhere connected, except at the base. This surface is displayed in fig. 6, in which the great vascularity of the organ is rendered more apparent. It is between the two cotyledons that the plume, consisting of minute convoluted leaves, is lodged and cherished. In fig. 7 a part of the nucleus of this seed is represented a little enlarged, and

Of the Seed. the two cotyledons have been removed at different places by transverse sections, to show more clearly the situation of the radicle and plume. The letter n denotes the place from which one of the cotyledons has been removed, so as to bring the plume o into view; and p points to the conical radicle below.

1st day. When this seed has been twenty-four hours in circumstances favourable to its germination, it is rendered tumid, and the umbilical aperture at its base is enlarged by the swelling of the parts within: the cotyledons become turgid, and the plume is augmented in size. After the second day the outer coat is much softened, the middle one appears as if torn and decaying, and all the parts within still further augmented in size. During the third day the colour of the outer tunic becomes darker, the cotyledons are more swollen, their vessels more conspicuous, and the radicle pushes out through the umbilical aperture.

4th day. When the fourth day has elapsed, the plantule is still retained within the tunics; and if these be now removed and examined, the middle one is found to be dry and thin: the cotyledons (g, fig. 8) are whitish, soft, and flexible, but the vessels on their surface are much more distinct: the radicle r is elongated and covered with down, as likewise is the stem s. At the top of the radicle a protuberance is seen, which is white and soft; and farther down appear several smaller tumours, indicating the places of rootlets about to break out.

6th day. About the sixth day the cotyledons t, fig. 9, emerge from beneath the tunics, representing the "dissimilar leaves" of Grew, and the "seminal leaves" of Malpighi, but which we shall in future denominate cotyledonous leaves. They are thick, soft, and a little separated from each other; their position is pendent; the colour yellowish; they are very vascular, and between them the plume is still concealed. The radicle u, at this period, is much elongated, and rootlets everywhere spring from its sides; the stem v is also lengthened and curved, and, in common with the radicle, is everywhere covered with a white curling down.

9th day. Towards the ninth day the cotyledonous leaves (x, fig. 10) assume an erect position. At their points they are still yellowish, but elsewhere green, and their cellular tissue is filled with a greenish-yellow juice; they begin to separate a little, but still entirely conceal the plume. The stem y, at this period, is greatly elongated, and its lower extremity has become green; the protuberance that existed at this part is greatly lessened; and below it the radicle z is continued, from which numerous rootlets, covered with capillary productions, break out.

21st day. During the following days the cotyledonous leaves continue to enlarge, the stem to elongate, and the plume to augment in size; but it is not yet unfolded. About the twenty-first day the development of the plantule appears to be completed. A representation of its foliage at this period is given in fig. 13; the cotyledonous leaves, a', have now reached a great size, are of a deep green colour, and very vascular: they rise by a short pedicle from the summit of the stem. On each leaf seven fasciculi of vessels are visible, which, beyond the middle, terminate in a net-work, from which is produced the cellular structure that contributes to form the breadth of the leaf. In the axil, formed by the cotyledonous leaves, the plume b lay concealed; it is now disclosed by the removal of one of those leaves. At first the leaves of the plume are curled and convoluted, but afterwards they expand, and their figure is then seen to differ entirely from that of the cotyledonous leaves; they have notched margins, and their surface is covered with down. In this manner the

young plant continues to increase, and acquires at length its full magnitude, in the progress towards which the cotyledonous leaves waste gradually, and finally fall.

The seeds of the radish (Raphanus), of the lettuce (Lactuca), and of the kidney-bean (Phaseolus), are represented by Malpighi as exhibiting, in their evolution, a similar succession of appearances. In all these seeds, and in many others of this division, the radicle first pierces the seminal tunics; next the cotyledons come into view, and assume generally the form of leaves, between which the tender plume is for a time concealed, and at a later period is disclosed. In different seeds, however, the forms of these organs, and the periods of their successive evolution, are subject to the greatest variation, not only as relates to the species of seed, but to the soil, the climate, and season in which it is destined to grow.

Having thus surveyed the changes in external form, which the germinating seed exhibits, we shall conclude our description by a display of the peculiarities of its internal structure. This we shall find to consist entirely of vessels and cellular tissue, variously proportioned and combined. In a longitudinal section of the radicle of the pea in an early stage, Malpighi represents it, as in fig. 11, to be composed entirely of cellular tissue exteriorly, in the centre of which the vascular system, separating at the top into three divisions to supply the plume and cotyledons, is placed. A similar section of the radicle (fig. 14) on the seventh day exhibits corresponding sections of the rootlets it puts forth; which are seen also to consist of cellular tissue, and of vessels that come off from the central fasciculus of the radicle.

The stem, when about a month old, is composed of a thick bark formed of cellular tissue, within which are several fasciculi of vessels. In some very thin slices of the stem of the pea, viewed through a microscope of considerable power, the arrangement and distribution of the vessels and cellular tissue appeared to us as represented in fig. 15. The centre of the stem was occupied by cellular tissue, round which was a zone of vessels, c', forming four principal fasciculi. Exterior to this zone was a small ring of thickened cellular tissue, and beyond this the proper cellular substance of the bark. Near the circumference of the stem were four larger fasciculi, which may probably be considered as the "proper vessels," while those near the centre may be deemed the sap-vessels.

In the plantule of the gourd a similar structure is observable. In fig. 19 is represented a longitudinal section of its stem and radicle on the ninth day, as given by Malpighi. In this stem the vessels are disposed in a circle that surrounds the pith; but as they descend towards the root, they approach each other, and give off ramifications to form the rootlets. In a transverse section of the stem, on the 21st day, Malpighi describes it as hollow in the centre, around which six fasciculi of vessels are disposed, and the intermediate portion is occupied by cellular tissue. Hedwig and Kieser, however, enumerate not fewer than ten fasciculi of vessels in the stem of this plant, some of which are placed next the pith, and others near to the bark, as in fig. 28.

To these representations of the structure of the pea and gourd by Malpighi, we shall add that of the bean in its early stages, as delineated by Grew, who has traced the distribution of the vessels in the radicle and cotyledons with great minuteness. In fig. 12, Plate XXXIX., is exhibited a vertical section of a young bean, which is made to pass through the cotyledons, the plume, and radicle. From the extremity of the radicle the vessels ascend in fasciculi to the neck of the plantule, where they are seen to diverge towards each cotyledon, and ramify through it, while a

Other central vascular portion is continued to the plume. In fig. 20 the vascular and cellular structures of the germinating bean are shown in conjunction, in a highly magnified representation, presented here in a reduced size from Grew. In this figure a denotes the cotyledon, e the enveloping tunics, f the cellular tissue, g the vascular system continued from the fasciculus k in the radicle, and ramified through the substance of the cotyledon; the letter x points to the plume, which also receives vessels from those of the radicle; and k indicates a depression in the cotyledon, in which, antecedent to germination, the plume was partly lodged; a similar depression existed in the other corresponding cotyledon.

In the foregoing chapter we have traced the successive changes of form which the seed exhibits in its progress to constitute the perfect plant; we have next to display the structure of the plant itself in its more remarkable varieties and forms. The leading features of this structure have already been laid before the reader when discoursing on the common textures of plants: it remains now to exhibit individual examples of it, as they occur in the several members of the trunk, the branch, and the root.

Botanists employ different terms to distinguish the different kinds of stems or stalks that support the leaves and the organs of fructification. These are the stem (caulis), which is considered peculiar to herbaceous plants; the trunk (truncus), which is proper to herbs and trees; the straw (culmus), which is the appropriate stem of the grasses; and the stalk (scapus), which differs from the other varieties in bearing flowers only, and not leaves. For the peculiarities in external form and character which distinguish these several kinds of stems, as they occur in different species of plants, we must refer to the writers on botany; and shall proceed to exhibit a general outline of their internal structure.

Perhaps there is no plant in which the simplicity of vegetable organization is more clearly displayed than in the sugar-cane, which belongs to the family of Gramineæ. In its stem or culm the cells and vessels are comparatively large, retain much of their more perfect forms, and are quite distinct from each other. When treating of the cellular tissue we referred to the cells of this plant exhibited in the very thin transverse slice, fig. 16, Plate XXXIX., as illustrative of their hexagonal figure, and of their being bounded on every side apparently by a single membrane. In some parts, when the observer is viewing these cells through the microscope, some of them appear quite transparent, from the upper and lower bounding membranes being entirely removed, and the light, in consequence, being freely transmitted; but in others one or both of these membranes remain, and though they are exquisitely thin, yet, when viewed by a strongly reflected light, a degree of refraction seems to be produced, which communicates to the surface of the membrane an irregular appearance, such as it has been attempted to express in the darker cells of the same figure. The deception that arises from viewing two layers of these cells in conjunction, which imparts the appearance of double sides,

as in fig. 21, was before noticed; and the longitudinal appearance of the same organs, as seen both in a single and double series of columns (fig. 22 and 23) was at the same time described.

It is through this cellular structure that the vessels which constitute the other component part of the culm of this plant are distributed. They occur in fasciculi, which, towards the centre, are placed at considerable distances from each other, and preserve a symmetrical arrangement; but nearer to the circumference they are more numerous, and their distribution is much less regular. In fig. 25, Plate XXXIX., a very thin transverse slice of this plant is exhibited, in which this regular disposition of the vessels at and near the centre x x, and their crowded state near the circumference t t, are well shown; the cells, too, at the centre are larger, and have a more perfect form than those near the circumference of the figure. In fig. 27 a very thin longitudinal slice of this same plant is delineated; it is considerably magnified; the letter u denotes the cellular tissue, and v v two fasciculi of vessels which ascend through it.

In the palm, which, though belonging to the division Stem of trees, we shall notice in this place, a similar disposition the palm; of the elementary organs is observed; but both the vessels and cells are smaller than in the sugar-cane, and the fasciculi of vessels are also much more numerous. This structure is exhibited in the transverse section of the trunk of the palm, fig. 26, in which the dark spots indicate the vascular fasciculi, and the whiter portion the cellular tissue. As before observed in the sugar-cane, the vessels are seen to be less numerous at the centre than at the circumference, where they are much crowded together, and very irregularly distributed, in consequence of the peculiar mode in which the growth of these trees is accomplished.

In fig. 24 of the same plate we have also copied from M. Desfontaines a portion of the longitudinal section of the trunk of another species of palm (dracana draco), which displays more clearly the irregular direction which the vessels take in the lax cellular tissue through which they are distributed. It was before remarked that these plants do not naturally produce branches, but that their vascular system is expended wholly in the production of leaves at their summit, and their trunk is perfectly cylindrical. If, however, the top of the plant has been cut off, or broken by accident, a division into branches is said to take place. (Mém. de l'Institut. Nat. tome i. p. 486.)

As in these plants the ligneous and cortical textures are uniformly blended together through the entire stem, it must be presumed that the sap-vessels and proper vessels are everywhere associated. This fact is accordingly pointed out by Malpighi as occurring in several species of the Gramineæ, who, as we before remarked, delineates a proper vessel as existing in each fasciculus of sap-vessels.

The next variety of structure we shall notice is that of the certain herbaceous plants, in which the proportion of cellular tissue in the stem much exceeds that of vessels; and, though belonging to the division of dicotyledons, they in several points resemble in structure those just described. In a section of the pea already given (fig. 15, Plate XXXIX.) the greater portion of the stem was seen to be made up of cellular tissue, through which several fasciculi of vessels were dispersed. The stem of the gourd has been examined with great minuteness by M. Kieser, who has represented in several sections, both transverse and longitudinal, its appearance in different parts at different periods of its growth. The total number of fasciculi of vessels that exist in the mature stem of this plant he

makes, with Hedwig, to be ten; but the number of vessels in each fasciculus he shows to vary at different periods, and in different parts of the same plant, according to their relative age. In fig. 28, Plate XXXIX., is a piece of the stem of the gourd, cut horizontally near the root, of its natural size and appearance, as represented by Kießer. The portion $\backslash ell$ of this figure is represented in fig. 29, magnified 130 times: it comprehends only one of the 10 fasciculi of vessels that exist in the stem. In this magnified view the letters $m\text{'}m\text{'}$ indicate the elongated cells of the bark; $n\text{'}n\text{'}$ the larger cells, which form the parenchyma of the stem; $d\text{'}d\text{'}$ the cells which surround the fasciculus of spiral vessels; $p\text{'}p\text{'}$ the fasciculus of spirals, containing 29 vessels, in part simple and in part punctuated spirals; $r\text{'}r\text{'}$ the great cells situated towards the centre of the stem: at $q\text{'}$ are four spiral vessels of different sizes, filled more or less with small punctuated vesicles, which spring from the walls of the vessel, and in two instances nearly close up their cavities.

The horizontal sections of the stem of the gourd exhibited above are taken from the same plant as the vertical sections of the vessels in figures 13, 14, 15, &c., Plate XXXVIII. The number of fasciculi that exist in the stem of this plant is 10; but the number of vessels in each fasciculus continually increases as the plant approaches maturity, and varies even in different parts with the different stages of its growth. Thus, of six sections made of the stem, that taken near the summit, and therefore of latest growth, exhibits only six or seven vessels in each of the 10 fasciculi, corresponding in this respect with the number found near the bottom in the earliest stage of growth. In the next section, made through the internodal space below, the vessels in each fasciculus have increased both in number and size; in the third section, the vessels have still further increased in size, and in number amount to 19 in each bundle; still lower down, in the fourth internode of the same stem, both the number and size of the vessels go on augmenting. In the fifth section, taken at the largest part of the stem, a still further increase in size is found, and the number in each fasciculus amounts to 23 vessels. Again, the section made in the stem of the same plant very near the root, and taken at a more advanced age in autumn, exhibits vessels approaching in character to those of the ligneous part of trees, and now amounting to 29 in each fasciculus. It is this last section which we have copied above in fig. 28 and 29, where the vessels may be seen smallest next the pith, and gradually increasing in size as they recede from the centre towards the circumference. Lastly, in a section of the branch of the root of the same plant, in place of 10 fasciculi of vessels met with in the stem and in the principal root, there were only four, but each fasciculus contained 37 vessels: the cells of this part were smaller than in the stem, and the whole construction was comparatively more vascular and compact.

The general characters of these vessels, from the earliest period of their development to the mature state of the plant, and the successive changes in size, in form, and appearance which they undergo, have been already exhibited in the series of figures 13, 14, 15, 16, 18, 19, of Plate XXXVIII., representing vertical sections, corresponding to the horizontal ones just described. As the plant increases in age the vessels augment in number, and in character they approach to those of arborescent plants: their sides are thickened, their transparency is almost lost, and the cavities of some are filled by vesicles. The balance, also, in the different orders of vessels at mature age is nearly the reverse of that which obtains in youth. In the young stem, says Kießer, the simple spirals are to the punctuated variety as 17 to 6, while in the adult

plant there are only 6 simple spirals to 23 of the punctuated variety in the stem; and in the root the simple spirals, which at first were equally visible, have altogether disappeared. A metamorphosis of the simple into the more complex varieties of vessel must, therefore, during the progress of growth, be deemed to take place.

In the stem of the gourd, as thus described, we observe an arrangement of parts in many respects agreeing with that of the sugar-cane. In both, the cells in certain parts are large, and their figure is well preserved; and the vessels are distributed in distinct fasciculi through the cellular tissue. In the gourd, however, the fasciculi are fewer in number, and do not suffer that displacement during the progress of vegetation which those of the cane experience. In neither plant are the vessels so numerous, or so situate, as to compress the cellular tissue into transverse partitions; nor is there in the transverse section of either any appearance of concentric layers. In neither stem is there any proper pith, the central part of the sugar-cane being occupied by cells and vessels, and that of the gourd being alike destitute of both.

In the sections of various herbaceous stems represented by Grew and Malpighi, the number of vessels and their disposition exhibit the greatest variety, and approach more or less to the arrangement of parts that is found in shrubs and trees. Thus, in holly-hock (Alcea) Grew describes the vessels of the bark as yielding a thin mucilage, and within them are the sap-vessels, postured in short rays, which comprise twelve or sixteen vessels. In Scorzonera, the "proper vessels" in the bark yield a milky fluid, and are postured in a radiated manner with the sap and spiral vessels that extend to the pith, all of which form parts of the same radiated lines. (Anat. of Plants, p. 103.) In endive, Malpighi describes the structure of the stem as approaching in character to that of trees. Near the circumference the vessels, says he, are disposed in lines directed towards the centre, and are separated by small ranges of cellular tissue, which is compressed into a solid and dense form; and more interiorly the vessels are larger, and extend a considerable way in rays through the stem. (Anat. Plantar. p. 25.) Where the vessels have this radiated position, the cellular tissue between them assumes the form of membranous partitions; but the greater portion of the stem of this plant is occupied by cellular tissue, not thus compressed into a membranous form. We thus clearly observe that, even in herbaceous stems, membranous partitions, extending more or less completely through the stem, are formed, whenever the vessels are disposed in radii, and are sufficiently numerous to compress, on either side, the cellular tissue that envelopes them.

ART. II.—Of the Trunk in Shrubs and Trees.

In shrubs and trees the several textures that compose the trunk are commonly distinct from each other; and of some parts that are but imperfectly distinguished in herbs become in them well defined. In the trunks of these plants, however, though a general uniformity of structure is observed, yet particular modifications of it exist in every species, and these are still further varied by the peculiar nature of the individual plant, and the mode and circumstances of its vegetation. Before descending to particular examples, it may be useful to exhibit a concise view of the parts that compose the trunk, and the terms employed to denote them.

Externally in every tree we have the skin or cuticle, the structure of which has been already described in a former section. In very young plants, and in young

branches, it is succulent, and its surface is entire; but in older trunks and branches it is frequently dried and broken, and in some instances is thrown off; so that the exterior covering of the plant appears to be formed by the cellular tissue of the cortical texture.

Beneath the skin is the bark, constructed, as we have seen, of cellular tissue, and of vessels collected into fasciculi, which at first are straight, and run parallel to each other; but, by the subsequent augmentation of the parts within, are separated at certain places, and touch only at a few points, so as to form a reticulated appearance. In the annual shoot only a single ring of vessels is observed, and these, with the tissue in which they are placed, form the cortical layer, as it has been called, of that period. Within this layer a new production of vessels and of cellular tissue takes place; and this being annually repeated, constitutes the series of layers of which the bark is ultimately composed. The new layer that is thus annually formed, and which appears to exercise an active vegetative function, was more particularly distinguished by the appellation of liber by the ancients, from its being the substance on which, before the invention of paper, they were accustomed to write. Very frequently, instead of a ring, the vessels of the bark are collected into distinct parcels or clusters, which, in the progress of vegetation, assume very different shapes.

Next to the bark is placed the wood, constructed, like the bark, of vessels and cellular tissue. Like it, too, it consists, in the young plant and in the annual shoot of the older one, of a single ring of vessels, which immediately surrounds the pith. In the following year a new ring of vessels is formed around the first, and in every succeeding year this process is repeated; so that the wood consists at length of a series of rings inclosing each other, the number of which denotes the age of the tree. The outer ring of newly formed vessels is more succulent than those of older growth, and is generally of a whiter colour; whence it has received the names of sap-wood or alburnum. By Du Hamel and most French writers it was named aubier, and by Hill it was called the blea. In every annual shoot the newly formed liber and alburnum are in contact, but in every succeeding year they are separated more and more from each other by the interposition of new matter between them; so that at length the first layer of bark and the first ring of wood occupy respectively positions the most distant; those, namely, of the centre and circumference of the tree.

The vessels that are thus annually formed, and constitute the alburnum, are disposed in radii which extend more or less completely from the circumference to the centre. In some trees the vessels are much more numerous than in others; and, in the progress of vegetation, they frequently suffer great alterations in size and external figure, as was before observed when treating of the vascular system. In the alburnum the size of the vessels is very uniform, and so it continues ever after in some trees; but in others it is very various, some being enlarged more than others by a greater influx of sap. With this increase of size the vessels seem to acquire the spiral character. They are also frequently closed up in the more aged parts of trees, by the production of vesicles within their cavities; and, in common with the smaller vessels and the cells, they are in some trees ultimately filled with gummy or resinous matter. In other instances, where the "proper juices" of the plant differ but little in quality from the common sap, the vessels, when no longer employed in active vegetation, become dry, and appear in some plants like empty capillary tubes; and in others, from the minuteness of their size, no aperture is visible. In this

state they have been regarded as solid filaments, and denominated the lignous fibres of the plant.

From the disposition of the vessels in radii more or less regular, the cellular tissue by which they are surrounded will be more or less compressed in the same direction, and form those transverse partitions between the several rays of vessels which we have already so frequently noticed. In addition to these fine partitions, described both by Grew and Malpighi, which are thus interposed between every ray or line of vessels, larger portions of cellular tissue are observed at certain distances, which extend in the same direction through the wood, and in some instances are continued through the bark also. By the authors just mentioned these partitions were described as stretching from the bark towards the pith, and were named insertments by Grew, and transverse utricular ranges by Malpighi. Others have held them to proceed rather from the pith to the bark, and have called them medullary rays. We shall venture to name them in future transverse septa, which appellation denotes simply their direction, and the fact of their forming partitions between the vessels, without having any reference to their origin. For similar reasons, the finer partitions of cellular tissue, previously described, may be denominated the lesser transverse septa.

Beside these transverse septa, there are sometimes portions of cellular tissue intercepting the vessels in a longitudinal direction. This was remarked both by Grew and Malpighi in the oak and several other trees: they appear sometimes to be very short, and sometimes to form nearly a continuous ring around the trunk. It is probable they are formed simply by the pressure of the vessels acting in a direction opposite to that by which the transverse septa are produced: they may, for distinction's sake, be named the longitudinal septa. In many trees they are not to be observed, or, if they exist, are so thin as to form only a sort of fascia on the adjacent vessels. To the cellular tissue promiscuously intermingled with the vessels some writers give the name of parenchyma of the wood; others term the elongated cells of this part lignous fibres.

The only other part of the trunk that remains to be mentioned is the pith. It is situated at the centre, and is surrounded commonly by a ring of vessels, but sometimes in part by thickened cellular tissue. Its proportion in shrubs is usually much larger than in trees; and in the young shoots of trees it occupies more space than in the older branches. At first in the young plant it is succulent, but afterwards becomes dry; and in aged trees it frequently is entirely obliterated, or at least rendered solid.

In a description of the growth of the trunk in a young plant of chesnut, Malpighi has given clear views of the gradual development of these several parts, and exhibited the appearances successively displayed in the first years of growth. (Anat. Plantar. p. 35.) In the vine some of these parts are very distinct. Fig. 1, Plate XI., represents a transverse section of the annual shoot of this plant. The pith a, in the centre, is large, and composed of cells of different sizes. From the pith to the skin in the vine, and, according to Grew, in the elm and some other trees, extend the transverse septa b. In some places the cellular structure of these septa in the recent plant is still visible, as at c; but generally they have the appearance of thickened membranes. Between every two septa two or sometimes three lines or rays of sap-vessels are interposed, which from their size are readily visible, and their place and disposition are indicated by the dotted marks d: the circular ring that bounds these sap-

Of the Trunk. vessels seems to be formed of condensed cellular tissue, and exterior to it the "proper vessels" are placed in the cellular tissue of the bark.

Trunk of the apple. In the next figure (fig. 2) is exhibited the transverse section of an apple-branch of one year's growth, which, like the former, is considerably magnified. The angulated figure and relative size of the pith at this period are shown. From the pith to the bark proceed numerous very fine septa, possessing somewhat of a curved direction; and between them innumerable black points, indicating the sap-vessels, are seen. These vessels are disposed in radii, and are bounded exteriorly by the thickened tissue of the bark: the bark is proportionally large, and in the midst of its cellular tissue clusters of "proper vessels," e, are observed.

In a similar section of the apple-branch of two years' growth, fig. 3, more highly magnified, the pith f is seen to have assumed a rounder form: from the pith towards the bark the transverse septa extend, between which the vascular radii are situate. The letter g denotes the place of junction of the two years' growth; h the newly formed alburnous vessels, situate between the bark and the wood; and i two ranges of "proper vessels," placed in the cellular tissue of the bark. Grew gives a highly magnified view of a section of this same plant about the third year of its growth, in the 25th table of his work, the appearances of which correspond with those just related, except that the transverse septa, as in all his figures, are made to represent straight lines, which, in the earlier periods of vegetable growth, does not appear to be quite correct.

of the oak. Proceeding next to the plant of more mature age, we shall copy from Grew part of a section of the oak, which in fig. 4 is exhibited on a reduced scale. The letter k denotes the bark, l the alburnum, m the wood, and n the pith. In this section the larger transverse septa o, and the smaller ones p, are very distinct, between which the sap-vessels of different sizes, indicated by the dark points, are situate; the letter q points to the "proper vessels" of the bark, some of which are collected into round parcels, and others form a ring.

Beside the transverse septa that intersect the diameter of the tree, we observe, in this figure, other lines, r, placed at right angles to them, and which appear to be interposed between each annual ring of wood. Grew considered it probable that these lines were produced in the oak, and also in the fig-tree and walnut, by a peculiar sort of sap-vessels that existed once in the bark, as did the turpentine vessels distributed through the wood of the pine. (Anat. of Plants, p. 115.) Hill also calls them sap-vessels seated on the outer edge of each annual circle, which in the young plant contain sometimes a limpid liquor, and at other times appear empty. He notices also the existence of "proper vessels," containing a thick juice in the alburnum of this tree. (On the Construction of Timber, p. 9.)

of very old oak. An examination of this wood at an age still more advanced may contribute to throw some further light on its structure. In fig. 5 is given the representation of a transverse section of the central portion of a piece of the wood of very old oak, of its natural size and appearance. The pith s, at the centre, was completely obstructed, and its cellular character obliterated; the letters ttt denote the larger transverse septa, rendered completely solid everywhere, and in some places partially obliterated; and u the white lines that extend in the opposite direction, and mark the boundaries of each year's growth. In the substance of these concentric lines numerous traces of minute apertures were visible; and from one line to another, small irregularly curved lines, as represented in the

drawing, were everywhere observed to extend. No vacuity was anywhere perceived to exist, but the whole formed one compact and solid mass.

A small and very thin slice of this wood, comprising a part of one of the larger transverse septa, and small portions of two of the concentric lines above mentioned, was submitted to a pretty highly magnifying power; and a delineation of its appearance is given in fig. 6. In this figure the letters x x denote the two concentric lines; w one of the smaller irregular lines that frequently extend from x to x, but are often intercepted, and form irregular spots or patches, as traced in several parts of the drawing; x the smaller transverse septa, situated at unequal distances from each other; and y the larger septum, the cellular character of which is entirely obliterated. Several small irregular lines, running in the same direction with the larger concentric ones, are also observed: these, at their junctions with the lesser transverse septa, produce a number of little squares, the area of which are filled up with a dotted or punctuated membrane.

This microscopical representation seems to confirm the opinion of Grew, that the concentric lines x x are chiefly composed of vessels; for their apertures, now rendered visible, are of different sizes, and the canal of many of the larger ones is filled up with a vesicular substance, often observed in the vessels of aged plants. The substance that surrounds these vessels appears to be condensed cellular tissue, rendered perfectly solid by the resinous matter with which it is filled. The irregular thickened lines w that run transversely to the former are similarly composed of vessels and condensed cellular tissue; and the smaller patches, we have noticed, are made up of the same. Both the large and small transverse septa consist of condensed cellular tissue alone; and such, too, appears to be the matter of which the fine lines or longitudinal septa that intersect them are composed. Some observations of Du Hamel on the formation of the ligneous layers may aid our inquiries into the nature of these latter septa, and the minute structure of the wood.

In a transverse section of the trunk of the oak, the elm, or the fir, says this very intelligent writer, the ligneous layers are distinctly visible, and it is commonly believed that each of these layers is the product of one year's growth. If, however, we cut obliquely a piece of oak, each layer is then seen, with the aid of a common lens, to be composed of a number of thinner layers, which mutually cover each other. By macerating pieces of wood in water, he was able to separate the annual layers into a great number of leaflets, thinner than the finest paper; and these primary layers, as they may be called, he afterwards ascertained by experiment to be formed successively during the whole period of active vegetation; so that the layer which is the product of one year's growth is itself formed of a number of layers exceedingly thin. (Phys. des Arb. tome i. p. 31, and tome ii. p. 19.)

Now it is probable that between each of these primary layers cellular tissue is interposed; and this tissue, by its compression, gives origin to the minute lines which in this tree have been noticed. In several trees, however, no similar lines are visible, because in all probability they do not, like the oak, afford secretions by which the tissue is thickened and rendered apparent. Even in a very thin slice of oak, when viewed in the microscope by a strongly reflected light, these lines almost entirely vanish, while those of the two orders of proper transverse septa remain.

M. Kieser examined the same wood in the transverse slice of a tree 100 years of age: and his representation of its structure is copied in fig. 23, Plate XL. In this figure, in

which the appearances are magnified 130 times, $a\text{'}a\text{'}$ indicate the cellular tissue, or, as he calls it, the parenchyma of the wood. At the top and bottom, the cells are obstructed and appear like obscure points (as is the case also in fig. 6 of the same plate); but in the middle of the figure the cavities of the cells are said to be still visible: the letters $b\text{'}b\text{'}$ denote the apertures of enlarged spiral vessels, all of which were filled with vesicles, but a few only are so represented, as in $c\text{'}$; $d\text{'}$ indicates the smaller vessels dispersed through the cellular tissue; $e\text{'}$ a portion of one of the larger septa, and $f\text{'}$ the smaller ones, which are often displaced by the augmentation in the size of the vessels.

In this highly magnified representation by Kieser, there is no appearance of the concentric lines or septa which intersect the transverse ones, as noticed by Grew and Malpighi, and exhibited in fig. 5 and 6; but the two orders of transverse septa remain. We before remarked, that in a slice of wood so very thin as, when viewed by a strong light, to become translucent, these lines in a great measure vanished; and in the figure of M. Kieser this evanescence is made to extend even to the thickened portion that intercepts the annual layers. It is therefore probable that all these appearances of concentric or longitudinal septa are produced simply by the thickening and compression of the cellular tissue in the oak, and such other trees as have viscid juices, and whose vessels are subject to irregular enlargement; for in whatever position these thickenings occur, numerous vessels are always to be observed. Hence in very thin slices they entirely vanish, while the proper transverse septa, being formed by the condensation of continuous portions of membrane, remain visible. In these figures it may also be remarked, that the surface occupied by cellular tissue seems equal nearly to that occupied by vessels; but it is probable that the apertures of many minute vessels dispersed through the tissue have not been distinguished from the cells. In a highly magnified representation, however, of the wood of this tree by Grew (Anat. of Plants, tab. 3, fig. 7), the proportion of the vascular to what is deemed the cellular part is still less.

The vessels of the bark in the oak, as delineated by Grew, fig. 4, Plate XL., are represented as disposed partly in parcels and partly in a ring; and in all the representations he has given of different trees, the greatest diversity in regard to the position of these vessels is found to exist. From the difference in the qualities of their juices, in the texture of the parts, and in the modes and circumstances of their growth, this variation may be expected to occur. In the oak, the pine, and many others, it seems certain also that a part of the cortical texture, or at least vessels containing the "proper juices" of the plant, are constantly mixed with those of the ligneous texture, and with them contribute to form each annual layer of wood. For examples of the diversity of structure exhibited in the trunk of a great many different trees, we may refer to the plates of Grew.

SECTION II.

Of the Structure of the Branch and its Appendages.

ART. I.—Of the Branch.

From the trunk springs the branch, which in structure resembles very exactly that of the trunk itself; and hence the description of the one serves entirely for that of the other.

Branches, in common with leaves, originate from buds, of which we have subsequently to treat, introductory to a description of leaves. At present we shall describe

only the mode of connection between the trunk and the branch. Of the Branch.

The branch, says Grew, springs not from the surface, but so deep as to take with it not only the bark and the wood, but the pith also, making its way at the parts where the vessels are separated by the transverse septa, and carrying with it the skin, which is extended with it. (Anat. of Plants, p. 28.) The separation of the vessels of the trunk to form the branch is not like the separation of a few filaments from a skein of thread; but they spring, says Du Hamel, from a centre, and bear with them a part of each portion of the tree. Hence, if a tree that is divided into two branches be cut a foot above the bifurcation, the surface of the sections resembles that of two trunks cut horizontally: if the section be made lower down, at the place where the branches spring, the same appearances of two series of concentric circles are seen at the axis of the trunk, but they are surrounded by other layers common to those which belong to each of the branches; and still lower on the trunk the concentric circles belonging to each branch diminish, and are finally lost in those which form the trunk from which they sprang. (Phys. des Arbres, tome i. p. 93.) These primary branches divide into others, and these again into still smaller branches, which form angles with each other, more or less great, according to the species of tree, and other causes.

The position of the branches on different trees is very various, but in the same species it is generally uniform. From the measurements of Du Hamel, it appears that in many trees the solid matter of the branches that go out from the trunk exceeds that of the trunk itself in the proportion nearly of five to four. (Phys. des Arbres, tome i. p. 96.)

Some have supposed that the branches proceeded always from, and were nourished by, the pith; but this opinion has been combatted by Du Hamel and Hill. The latter gives a longitudinal section of the trunk of a species of American dog-wood (copied in fig. 7, Plate XL.), representing the origin of two buds $g\text{'}g\text{'}$, which are seen in the state of pushing out on either side through the vessels of the wood. In this trunk the pith is of a brown colour, the inmost ring of vessels green, and the outer ones white. Through the white rings the inner green vessels are seen to shoot, and the brown pith is left entire behind; though the new branch at length obtains a pith for itself, which, however, has no connection with that of the trunk. In a similar section of an older trunk (fig. 8), the branches are seen to spring also from the inmost circle of wood; the pith $h$ is not at all disturbed, but each branch is furnished with its own pith $i\text{'}i\text{'}$. In a section of the vine (fig. 9), a similar origin of the branches, from the inmost circle of wood, is observed; and not only is the pith $h$ of the branch distinct from that of the trunk, but the pith of the trunk itself is intercepted by the shooting of a branch $l$ across it. (On the Construction of Timber, p. 35.) This latter fact of the interception of the pith by the shooting of a branch across it was previously noticed by Grew in the vine and some other plants. (Anat. of Plants, tab. 19.)

When a bud thus protrudes to form a branch, the perpendicular vessels of the trunk are compelled to separate; and they afterwards meet above, and pursue their former direction. This is shown in fig. 11, Plate XL., copied from Du Hamel, in which the bud $m\text{'}$, in the act of sprouting, is seen to push to either side the vessels of the trunk, which again meet above it.

The vessels both of the wood and bark, according to the same author, have their direction determined chiefly

by the course of the sap. If the sap preserve a perpendicular direction, as in trees that have no branches, the vessels are perpendicular also; but if it move to one side, these vessels then change their direction. This is strikingly evinced in a tree that has been cut over above a branch; for then all the sap being obliged to pass to the young branch, the vessels all at once take the same direction; so that if a tree has been cut over in winter, and at the end of the succeeding spring its branch below be examined, the new vessels of the branch will be seen to cross those of the trunk, as exhibited in fig. 12, n, Plate XL. (Phys. des Arbres, tome ii. p. 53.)

In figures 7, 8, and 9, copied from Hill, the young branch, in every instance, is seen to originate from the inmost ring of vessels that surrounds the pith; and he was of opinion that all buds and branches sprang from this part alone. To this vascular circle he gave the name of corona, and held it to be the most important part of the vegetable body,—that it was like no other part of the plant, but contained within itself the essence of them all. (On the Construction of Timber, p. 21.)

It is however notorious, that trees in which not only this first circle, but almost every other circle of vessels, has perished, produce leaves and shoots from the trunk where the bark is entire, as this author himself admits, p. 44 of his work. There does not appear, either in the anatomical character, or in the functions of this circle, any thing that distinguishes it from the others, except priority of formation, and its being in consequence the seat from which the first buds and branches spring.

A much more correct idea of the origin of buds and branches was entertained by Du Hamel, who illustrates it by the diagram fig. 10, Plate XL. Let us suppose this figure to represent a tree of four years' growth, as indicated by the four ligneous cones which at its base envelope each other. In the first year a bud o springs from the inmost ring, which by the fourth year is seen to consist of four ligneous layers. In the second year a second bud p springs from the ring of that year, and consists only of three layers; the next year a bud q is developed on the first bud o, and possesses only two layers; and the bud r, developed the fourth year on the outmost ring of wood, is seen to consist only of one layer. (Phys. des Arbres, tome ii. p. 53.) In this way every ligneous layer may be considered equally capable of giving origin to buds and branches.

Connected both in origin and in structure with branches are those appendages, frequently observed upon them, called thorns. They are very conspicuous on the hawthorn, and are constituted, says Grew, of the same parts as the bud itself, and in a like proportion. They spring from the outer portion of the ligneous texture, and may be considered as abortive buds. (Anat. of Plants, p. 33.) Malpighi describes them as being frequently produced in the axils of leaves, and assuming for a time the form of a branch, but at length degenerating into a thorn. He considers them to arise from defective nutrition, and adds, that they disappear frequently under higher culture. (Anat. Plantar. p. 138.)

According to Willdenow, most species of our fruit-trees naturally possess thorns, which disappear entirely and become branches under higher culture. Even in the black-thorn the prickles diminish in number under improved culture, but do not entirely disappear.

Sometimes, however, instead of being produced from abortive buds, thorns owe their origin to the degeneration

of other organs; the petioles of some pinnate leaves which are persistent become thorns, and the same thing is observed of the peduncles of some flowers. Certain stipules also, as those of Mimosa, are said sometimes to change into thorns (Principles of Botany, p. 270); and in the date, according to M. de Candolle, a lobe of the leaf has been converted into a thorn. (Théorie Élément. de la Botanique, p. 344.) In other instances, as in the holly, the leaf produces thorns round its entire margin, which are formed by the vessels that bound it. (See fig. 11, Plate XLI.)

When the bark is removed from a branch that possesses thorns, the thorns still remain, by which they are distinguished from prickles that originate from the bark. In fig. 20, Plate XLI., is a representation of a branch deprived of its bark, in which thorns of different forms are seen to spring from the ligneous texture; and in fig. 21 of the same plate is a longitudinal section of another branch covered with its bark, in which the ligneous and cortical textures of the thorn are displayed. For the descriptions of the varieties of thorns, and their external appearance, we must refer to the writers on botany.

From the similarity in structure which these bodies exhibit to the branches from which they spring, we have placed them as appendages to those members. As is well known, they are met with only in certain plants, and their obvious use is to attach the different parts of the plant to one another, for mutual support, or to the objects in their neighbourhood. In a general manner they are denominated fulera: the most remarkable species are those which, like the claws of ivy, called claviculæ by Malpighi, are not rolled into a spiral form; and those of the vine, named tendrils (cirri), which possess a spiral conformation.

According to Malpighi, the claws or claspers of ivy possess a roundish form, and are covered with hairs, of which yield a viscous humour, by which they are agglutinated to stones or to walls and trees. Their forms are exhibited in fig. 22, e f, Plate XLI., copied from Malpighi; and in the lower part, g, of the same figure, their origin and structure are displayed. They are seen to spring from the ligneous texture of the stem, and to possess a similar structure.

The tendril is described by the same author as springing from knots between the origins of leaves. It possesses a round stem, which is sometimes covered with hairs; and this stem frequently divides into several branches. At first it is very tender, but gradually acquires solidity, and assumes the spiral character; its colour is green, and it is composed, like the trunk, of all common textures. (Anat. Plantar. p. 139.) By Willdenow it is regarded as an abortive leaf, being simply a prolongation of the petiole, without the leafy expansion. (Principles of Botany, p. 272.) Its form, as it occurs in the vine, is exhibited in fig. 23, Plate XLI.; but in this particular it is subject to the greatest variation.

The leaf itself, as well as the petiole, is said to be sometimes prolonged into a tortuous appendage resembling the tendril; and similar transformations are sometimes exhibited by the peduncle and petals of the flower. In one species of vine mentioned by Malpighi, the extremity of the tendril, which is at first pointed, becomes gradually incurvated and reflected, and at length is formed into a roundish body, which is furnished with small papillæ, that yield a viscous fluid, by means of which it attaches itself to walls or wood, so strongly as not to be easily separated.

In dodder (cuscuta), small tubercles are formed on the stem, which, according to M. de Candolle, are so organized as to fix themselves on any other plant, and derive nourishment from it. (Théorie Élément. de la Botanique, p. 344.) For descriptions of the several varieties of these organs enumerated by botanists, and for the terms employed to express them, we must refer to their works.

Considered in relation to their general form, roots, says M. de Candolle, exhibit two very distinct appearances. Some roots have a conical trunk, simple or ramified, but springing always from a single base, and their radicle only elongates or ramifies; others spring always in a bundle, more or less marked, from a common base, which is so confounded with the neck of the plant that it is sometimes taken for the base of the stem, and sometimes for the trunk of the root. The first form of root is met with in the greater number of exogenous plants; the second occurs chiefly in the endogenous, and also in those exogenous which have fascicular roots.

Of the conical roots, some ramify greatly, and have numerous fibrous or capillary rootlets; others produce few or no branches, and are furnished with few fibrous roots. In the former class, the principal differences arise from the degree of ramification; in the latter, two varieties of form are very distinct—that of the carrot, named fusiform, from its tapering to a point, like a spindle—and that named rapiform by De Candolle, from the swollen form it assumes at its origin, and sudden shrinking to a point below, as in the turnip.

The fascicular roots present also many varieties of form: 1st, Those of the Gramineæ, in which each rootlet, though distinct at its birth, ramifies so much that the divisions resemble the fibrous productions of the preceding class: 2dly, Those of the hyacinth, in which many cylindrical rootlets spring from a common base, and descend nearly parallel, or with a slight divergence: 3dly, Those of the orchis, in which, besides cylindrical rootlets, there are others swollen into tubercles: 4thly, Those of Ranunculus, where all the fibres that spring from the base are more or less swollen, so as to form a bundle of oblong tubercles. Between all these distinct forms there are several intermediate varieties. Under the name of tuberous roots are included all those which have swellings in any part of their length, and that, too, whether they sprang originally from a single or common base. It may, however, be observed, that many tubercles which spring apparently from roots, are in reality productions from inferior branches of the stem, concealed beneath the soil, as M. Dunal first proved to be the fact in the potato. (Organog. Végét. tome i. p. 252.)

The root is held to consist of two parts; the main body with its branches, and the fibrous productions called capillary rootlets. These parts are commonly hid beneath the soil, but in some plants they are attached to rocks or stones, in others they are fixed in other vegetables. Some plants, again, have no fixed attachment, but float loosely on water; and others are considered to possess no root at all.

With respect to structure, the body of the root of trees, says Malpighi (Anat. Plantar. p. 145), may be regarded as an elongation of the trunk beneath the soil, being constructed of the same textures, disposed nearly in the same manner. In various herbaceous plants the root has been examined with much minuteness by Grew. (Anat. of Plants, book ii.) The cuticle, says he, is of different

colours in different plants: sometimes its surface is smooth, at others rough; and its texture is either thin and transparent, or thick and opaque. At an early period of the the cuticle appears to be a continuation of that of the radicle, but afterwards seems to be derived from the bark. It is composed chiefly of cellular tissue, through which vessels are distributed in the direction of the length of the root. The existence of these vessels, though scarcely visible by the microscope, may be inferred from the fact that it is more difficult to tear the skin in a longitudinal than transverse direction. If, also, the skin be cut transversely, and then allowed to dry, all the parts where there are no vessels shrink below the surface of the cut end; but where there are vessels there is no shrinking. Or if, after making a transverse section of the root, the skin be pressed very gently with the edge of the nail, sap will sometimes arise from the skin, as from any other part where vessels are placed; which does not happen with fluids contained in the cellular tissue, unless the cells be broken. (Ibid. p. 62.)

Beneath the cuticle lies the bark, which is found to differ very much, both in colour and thickness, in different plants; scarcely in some instances exceeding the skin in thickness, whilst in others it is thicker than the woody part of the root. It is composed of vascular fasciculi and cellular tissue, the cells of which are smaller than those in the trunk: these cells are generally of a spherical form, or nearly so, often transparent, but not visibly pervious into one another. In some roots the cellular tissue is disposed in rays extending from the inner to the outward edge of the bark; in others it extends only in part through the bark. The cells themselves are sometimes uniform, and sometimes they differ both in size and figure. They are often filled with lymph, or other matter. (Ibid. p. 64.)

In the cellular tissue vessels running lengthwise through the bark like small threads are observed; and in a transverse section the sap is seen to exude from them. They often form a net-work, and seem as if inosculated; but are only contiguous, like the braces in a drum. They are never single, but form clusters of 20 or more together: their form is cylindrical, and they are of the same size through their whole course. Some vessels frequently separate from a cluster and are transferred from place to place; but they never ramify so as to be successively propagated one from another. Their position in the bark is very various, forming sometimes a ring at its inner edge, and dispersed sometimes in clusters through it, or disposed in the form of lines or rays. The fluid they contain varies in colour and consistence in different plants, being sometimes thin and clear, sometimes thick and mucilaginous, often milky or resinous. Sometimes the fluids of the bark vary in the same root, which, when cut transversely, will yield both a limpid and milky sap. (Ibid. p. 67.)

The woody part of the root, like that of the trunk, is of the made up of cellular tissue and vessels. This tissue varies greatly in proportion in different roots, but is less abundant in the root than in the stem. It is disposed sometimes in rings, but oftener in lines and rays, like the transverse septa in the stem. Through this tissue two kinds of vessels are distributed, one containing a lymph, and the other being empty; whence they have been denominated by Grew the succiferous and air-vessels. In some roots these vessels are disposed in clusters or fasciculi; in others they form rings; and in others they have a radiated form. Both kinds are much intermixed, and everywhere surrounded by cellular tissue. The number of empty vessels is sometimes small and sometimes great;

Of Roots. and the same may be said of their size, even in the same root. Malpighi, says Grew, has described them as being constructed of a thin transparent zone, of a silvery colour, having a spiral form; to which Grew adds, that the spiral zone is not a single piece of a flat form, but consists of two or more round and true fibres, not inosculated side to side, but knit together by other smaller fibres,—a construction similar to that which Kieser has since demonstrated. Their spiration, however, is better seen in the stem, and better in young parts than in old ones. Though these spiral vessels seemed empty, or containing only air, yet, says Grew, "whether these vessels may not, in some vegetables, and at some times, contain liquor, is doubtful." (Ibid. p. 74.)

Root of wormwood. In illustration of the structure above stated, Grew exhibits, in twelve successive tables of his work, the external forms and internal structure of various roots. As a specimen of this structure, we have copied his representations of the root of wormwood. In fig. 13, Plate XL., is a delineation of a transverse section of this root, of its natural size; and in fig. 14 is a magnified representation of a portion of the same section, here presented on a reduced scale, in which A denotes the cuticle, B the bark, and C the wood. The smaller letter a indicates the ligneous rays formed by the vessels, and b the cellular tissue or transverse septa continuous with that of the bark. The dark spots in the bark are described by Grew as balsam vessels, intermingled with lympheducts; and those in wood as the spiral or empty vessels. It will be seen that the vessels of the woody part are smallest at the centre, and gradually augment in size as they approach the circumference, as Kieser has since observed; and that, with the exception of a few vessels said to convey lymph, all those of the wood are considered as air-vessels, and therefore as possessing a spiral conformation. In his chapter on roots, Malpighi also frequently describes tracheæ as occurring in the dissections made by him of the roots of various plants.

Root of the gourd. In fig. 21, Plate XL., is exhibited a horizontal section of the root of the gourd, of the natural size, as given by Kieser, representing four fasciculi of vessels, intercepted by cellular tissue. One of these fasciculi, consisting of 37 vessels, is represented in fig. 22, magnified 130 times. Of these vessels, the largest, as usual, are situated exteriorly, and they pretty regularly diminish in size as they approach the centre. Two of these vessels are nearly filled with vesicles, and the whole of them are generally of smaller size than those in the stem of the same plant. A vertical section of one of the largest vessels near the circumference, and of two of the smallest at the centre of this root, have been already exhibited in figure 19, Plate XXXVIII. They all belonged to the variety of punctuated spirals; so that all the vessels in the root of this plant, as well as in the stem, may be considered as possessing the spiral conformation.

Spiral vessels in roots. To the testimony of Grew, Malpighi, and Kieser, as to the existence of spiral vessels in roots, we may add that of Du Hamel, who says that if a young root, as that of the horse-chestnut, be duly macerated in boiling water, and its bark afterwards removed, the ligneous portion will be found so tender as to be divisible into fine filaments; which filaments, when viewed with a microscope, will be found to be spiral vessels or tracheæ, which, he adds, might lead one to think that all the ligneous fibres are really spiral fibres, while they are yet tender. (Phys. des Arb. tome ii. p. 13, 14.) Mirbel and some other writers, however, deny altogether the existence of spiral vessels in the root; and M. de Candolle declares that one of the chief differences in structure between the stem

and the root consists in the total absence of tracheæ in the latter. (Organog. Végét. tome i. p. 242.) Yielding credence, however, to the positive testimony of Grew, Malpighi, Du Hamel, and Kieser, rather than to the negative assertions of their opponents, we are disposed to conclude, that all the vessels which convey the ascending sap in the root possess, like those in the stem, a spiral structure.

It cannot fail to be observed also, in the sections of roots just exhibited, that the proportion of vessels to cells is much larger in the root than in the stem of herbs, and that consequently the cells are smaller and more compressed. The vessels occupy also the very centre of the root, to the exclusion of all cellular tissue, in the form of pith, in exogenous plants; but in the endogenous, the internal structure of the roots, as compared with that of the stem, is said to present no sensible difference. In exogenous plants, also, the diametral insertions, or septa, and the appearance of the bark, are as distinct in the root as in the stem; while in the endogenous the vessels and cells are intermingled through the entire root, as well as through the stem. In trees, the structure of the root is more distinct than in herbs, and still more nearly resembles that of the trunk.

In treating of the evolution of seeds, we exhibited several examples of the manner in which the fibrous or capillary rootlets spring from the trunk, or larger branches of the root. These rootlets are more simple in structure than buds, being made up chiefly of the ligneous portion of the plant. They are the organs which absorb nutriment from the earth; and hence, as Du Hamel observed, the earth is exhausted chiefly where these capillary rootlets are distributed, and not in the neighbourhood of the larger roots. On examining the roots of trees after severe winters, he often found the capillary rootlets dead; which led him to suppose that trees lose in the earth their capillary rootlets nearly as in the air they lose their leaves. He found, that even after slight frosts many of these rootlets were dead; and that, when the temperature became milder, new rootlets were developed, which abundantly replaced those destroyed. (Phys. des Arb. tome i. p. 89.)

All roots raised from seeds have a tap-root, that grows in length in proportion as the earth is easily penetrable. The seeds of the oak, sown in a deep and rich sand, have produced plants, says Du Hamel, the stems of which were only six inches in height, while the root extended nearly to four feet. If, however, the root meet below the surface with any obstacle that opposes its elongation, it then continues short, and subdivides into many lateral branches. This result may be also obtained by simply cutting the extremity of the root; for roots, as Du Hamel first ascertained by experiment, do not elongate but by their extremity. (Phys. des Arb. tome i. p. 83.) Hence, when injured by insects, or in any other way, so as to check its elongation, the root, by throwing out lateral branches, will often derive more nutriment by these new productions than it would have obtained by a single trunk.

Roots are provided through their whole extent with a quantity of germs, or a capacity of producing them, when placed in favourable circumstances; and this capacity is possessed also by the branch. If the branch of many trees be cut off and planted in the earth, it emits rootlets from its sides beneath the soil. If even the bark of a branch be removed so as to intercept the descending sap, and the part be then surrounded with moist earth, the upper portion of the divided bark will emit rootlets into the earth. The simple immersion of cuttings of the willow in water is sufficient to cause the production of rootlets,

which spring from the cutting, says Malpighi, just at the surface of the water, not below it. They appear at first like small tumours, and are composed of ligneous fibres and tracheæ, which gradually force their way through the bark. (Anat. Plantar. p. 146.)

In like manner, as a branch may thus be made to produce rootlets, so may a root be made to produce branches. If a root be cut off, or an incision be made in its bark, so as to intercept the descending sap, a swelling, or bourrelet, will form around the upper extremity, from which new rootlets will spring; or if this cut end be left exposed to the atmosphere, then, says Du Hamel (Phys. des Arb. tome i. p. 102), in place of rootlets, new shoots or branchlets will be developed. It is thus that many plants, as mint, which spread their roots horizontally, produce new stems from them when they happen to come sufficiently near to the surface. From the facts above stated, Du Hamel maintains that the tumour or enlargement (bourrelet) formed either on the cut stem or the root of a plant is essentially the same; that in both it contains germs proper for producing either branches or roots; and that the one or the other is developed, according as the bourrelet is exposed to the air or is placed in the earth. (Phys. des Arbres, tome i. p. 102.)

Others have maintained not only a perfect identity of structure between the root and branch, but also between the bud that gives origin to the branch, and the rootlet that springs from the root; and further, that they are convertible into each other. In proof of this it is alleged, that if a willow plant be bent down so that its head shall be fixed in the earth, it will soon push forth rootlets; and that if, after these are developed, the trunk with its former roots be raised so as to be exposed to the air, there will shortly be formed a new head, furnished with leaves and branches. These facts, however, prove only a substitution, and not a conversion, of these organs into one another; for the young buds of the head die when placed under the soil, and the new roots spring from points not previously occupied by buds; while, in like manner, all the young rootlets developed beneath the soil perish in the air, and new buds are produced from the older trunks of the root. In some plants the stem seems to perish, and the root in the following year produces a new stem; and from this it has been inferred that the root alone preserves the vitality of the individual: but the stem in these plants does not, says M. de Candolle, entirely perish; and in the phenomena exhibited by cuttings the contrary takes place, for the stem develops new roots. We ought therefore, says he, to regard the stem and the root as of equal importance. Their union constitutes the body of the plant. (Organog. Végét. tome i. p. 249.)

The knots or articulations in the stems of plants, by retarding the descent of the juices, favour the production of new parts; and according as the stem may be exposed to dryness or humidity, either branches or roots may be formed. Hence stems that have many knots most easily propagate by slips and layers; and fleshy plants, as the Cactus, which possess a thick and moist cellular tissue, have a disposition to propagate roots even in the open air. It is the same with roots possessing tubercles, which are a sort of magazines of nutrient matter, and possess, therefore, a singular aptitude for emitting new productions. (Organog. Végét. tome i. p. 248.)

We have already stated that roots grow in length, not by an extension of parts, but only by the addition of new matter to their extremity. This power of growth Du Hamel found to reside within two or three lines of the extremity, both in ligneous and herbaceous roots. (Phys.

des Arb. tome i. p. 84.) It is chiefly by their extremities, also, that they absorb fluids; for Senebier found, that if the extremity of a root, as that of a carrot or radish, was immersed in water, the water was absorbed; whilst if the whole root, with the exception of the extremity, was immersed, little absorption took place:—results which have been confirmed by the experiments of Carradori. These facts have led M. de Candolle to consider the extremities of roots, and of all fibrous rootlets, as furnished with a peculiar hygroscopic power. This power he conceives to depend, as we before remarked, on the existence of a minute cellular structure at this part, to which structure he has given the name of spongioles (spongiolæ). These spongioles, according to him, are formed of a very compact cellular tissue, composed of roundish cells. He distinguishes several varieties of these organs, as the radical, the pistillary, and the seminal, according as they are seated at the extremities of the radical fibres, of the pistil, or in the coats of seeds. So great, he adds, is their absorbing power, that they absorb the colouring molecules of liquids, though these molecules will not enter the ordinary pores, which are infinitely greater in size. (Organog. Végét. tome i. chap. 7, p. 89.) In his 10th plate he exhibits the appearance of these spongioles in the radicles of pan-danus odoratissimus, where they are better seen than in any other plant examined by him.

The nature and qualities of the soil exert a great influence on the form and size of roots. If the soil be free soil on and pervious, the root often descends to a great depth in the form of a tap-root; but if it encounter obstacles which oppose its descent, it then continues short and divides into lateral branches. Roots also extend into the portions of soil that are richest, while the barren parts are nearly destitute of roots. (Phys. des Arbres, tome i. p. 82.)

CHAP. III.

THE ANATOMY OF THE ORGANS OF VEGETABLES.

SECTION I.

Of the Structure of Buds.

In the preceding chapter we have described the larger buds and more permanent members of the plant. We have now to delineate the form and structure of organs, whose existence is of shorter duration, and which present a more diversified character. They may all be said either directly or indirectly to originate from buds, with which therefore we shall commence our descriptions.

A bud, according to Gaertner, is an organic body, sprouting from the surface of a plant. In the beginning it is distinct from the proper and permanent members of the plant, but after some time becomes a part of it; or, if separated, is capable, by the increase of its own proper substance, of growing into a new plant, perfectly similar to its parent. (De Fructib. Plantar. p. 3.)

Buds may in general be easily distinguished from seeds; but in some of the lower tribes of vegetables the conformity in external appearance and internal structure is said to be so great as to render this a matter of much difficulty.

Gaertner enumerates four species of buds, two of which are leafless, and two are provided with leaves or scales. The first, propago, is the most simple species of bud; it is leafless, and possesses different forms: sometimes it is entirely naked, and sometimes is inclosed in a bark-like case: it at length separates spontaneously from its parent, and is dispersed like a seed.

Of Buds. The second species is termed gongylus: it is described as a perfectly simple, leafless, somewhat globular and solid bud, concealed within the bark of the parent plant, and never separating spontaneously till the bark decays by age. It has a great affinity to the tuber of roots, but differs from it in not being a proper member of the plant, and in not possessing the principle of multiplication which resides in the tuber.

Bulbus. The third species of bud is a compound germin, and named bulbus. Its figure is somewhat globular: it is nearly leafless, and formed of a very short keel (carina) and thick succulent scales, and at length separates spontaneously from its parent. Of this species there are two varieties, the one solid, constructed of a solid fleshy body, and having generally the rudiments of the new plant fixed outwardly upon it; the other coated, being formed of several concentric scales, in the centre of which the young plant is cherished.

Gemma. The last species of bud, and that which is strictly so named, Gærtner denominates gemma. This is a bud composed of a subulate keel, and distinct herbaceous leaflets: it resembles a branch in miniature, and never separates spontaneously from its parent. It is named an eye (oculus) when it puts forth flowers alone, or flowers and leaves together; and simply a gem (gemma) when it is unfolded into leaves alone.

Of these several species, the last, gemma, is generally considered as alone entitled to the appellation of a true bud by those who consider scales and leaves as essential to its constitution; but this definition, says Gærtner, would exclude even bulbs, and it is better to derive our idea of a bud from a general agreement in properties, and particularly from similarity of origin in formation and evolution, than from the ever-varying condition of external form. The two former species of buds occur in the lower tribes of vegetables; the two latter are observed on the stems and branches, and roots of various plants. On the present occasion we shall notice the latter only.

Structure of bulbs. That variety of bud which is called a bulb, and which springs from, or is variously attached to, the roots of many plants, has already been stated to exhibit considerable difference in structure. It is sometimes constructed of several thick scales or leaves enveloping each other, and is sometimes formed of a more continuous and solid substance. To the first description belongs the bulbous substance of the lily and tulip. Grew has given a section of the latter, made in the month of September. It displays the tunicated structure of the bulb, at the base and in the centre of which the young flower, destined to appear in the following spring, is observed. (See his Anatomy of Plants, table 63.) Of the tunicated variety, the common onion affords also a good example. The coats that compose its bulb are to be regarded as fleshy leaves, and the true root, according to Du Hamel, is the fleshy plate that supports the bulb, from which the rootlets spring. The common potato affords an example of the solid bulb, on whose surface numerous buds, all capable of producing entire plants, are seen. Botanists enumerate a great many other varieties of bulbs, for which we must refer to their writings; and to the works of Malpighi and Du Hamel for the anatomy of many of them.

Names of branch-buds. The true bud of the stem and branch was distinguished by the ancients into two kinds, according as it produced either a leaf or a flower, and to it they assigned different names. "Germen autem est id quod ex ipsis arborum surculis primo vere exit, ex quo deinde folium producitur," says Pliny; and this he distinguishes from the flower-bud, "nam gemma proprie floris est." By Grew the term germen, and by Malpighi the word gemma, is employed to

denote each variety of bud. By Linnæus the term germen is used to denote, not the bud of a branch, but the rudiment of a seed; and not the rudiment of the seed only, but the organ also that contains it. The word gemma he uses to denote a bud. Gærtner adopts the word germen in a generic sense, to express every species of bud; and, with Malpighi and Linnæus, employs the term gemma as indicative of that species now under consideration. This, therefore, as the most generally received appellation, we shall in future employ.

During summer, says Du Hamel, buds are gradually formed in the axils of the leaves, viz. in the angle which the petiole forms with the branch. They are at first exceedingly minute; are seen in winter chiefly on the young branches, sometimes on the larger ones, and more rarely on the trunk. They exhibit different forms, according to the kind of tree that bears them, and are attached to it by a very short pedicle. Their position on the branch was considered by M. Bonnet to be reducible to five classes; sometimes they are situated on opposite sides of the branch, but placed alternately, and sometimes they are placed exactly opposite to each other. In other instances they form a kind of ring round the branch. Sometimes they have a spiral disposition, and at others constitute a sort of double spiral around the branch. In those cases where the buds stand opposite on the branch, the extremity of the branch is frequently terminated by three buds; but where the buds are only alternate, the young branch is commonly terminated by a single bud. In the pine the true buds are placed, not in the axils of the leaves, but at the extremity of the branch alone. Some buds stand out a considerable distance from the branch; others are placed in close contact with it; and these varieties occur sometimes on the same branch in regard to the buds that issue from its sides and extremity. The shape of buds is also very various—some being long and pointed, others short and round; some again are hairy, others smooth; some very small, and others large. (Phys. des Arbres, tome i. p. 99.)

Beside these varieties in position and form, which serve to distinguish the buds of different genera and species, there are also many sorts of buds to be observed on the same tree, whose characters are discoverable by their form. Those which are pointed usually produce leaves and branches, and from those which are large and round commonly proceed flowers. The former are named leaf or wood buds (gemmæ foliiferae), the latter flower or fruit buds (gemmæ floriferae): others which possess both leaves and flowers have been called mixed buds (gemmæ mixtae). In some trees, as those of the apple and pear, two varieties of wood-buds occur, one of which is small, produces only a small bunch of leaves, and in the end becomes a fruit-bud; the other is larger, and gives origin to a branch.

As the rudiments of the flower appear in the bulb of roots the season before they are destined to bloom, so those of the leaf and the flower are distinguishable at the same period in the bud of the branch. They are to be perceived, says Du Hamel, in autumn, and continue to grow even during winter, appearing to be clandestinely formed in that season, when the movements of the sap seem to be suspended; and are thus prepared to shoot forth on the return of spring. It is, however, only in perennial plants that these phenomena are observed. Annual plants do not produce buds, and even those whose roots survive the fall of the stem produce buds only on their roots. (Phys. des Arbres, tome i. p. 103.)

Climate appears to exert the greatest influence on the formation and evolution of buds. In cold regions, as we

have just observed, the bud is formed many months before it is destined to shoot into a leaf or branch; but in warmer regions scarcely any interval occurs between the periods of formation and evolution. The buds, in such climates, are said to unfold themselves immediately from the bark into branches, without having remained in the form of buds for any length of time. Sometimes, in the milder seasons of our own climate, the evolution of buds rapidly succeeds to their formation, and the vegetative process with us emulates the productive powers of more favoured climates. In some examples, however, the specific characters of particular plants overcome these general tendencies of climate; and thus hot climates are said to possess some bud-bearing plants, and in colder climates there are a few shrubs which are said never to bud. (Willdenow's Principles of Botany, p. 213.) It seems, however, more correct to consider all plants as bearing buds, from which the branches, the leaves, and the flowers, successively proceed; and to say that in warm climates in general no suspension of the vegetative process occurs, as in cold ones, and no marked interval is observed, therefore, between the formation and evolution of buds. The few exceptions that occur respectively in warm and cold climates must be considered in reference to the specific characters of the individual plants. The process by which buds are actually formed has been called Gemmation or Gemmification.

Having made these few general remarks on the nature and formation of buds, we have next to exhibit a few examples of their structure and evolution, confining ourselves at present to a description of those which produce either branches or leaves. The mature bud consists of two parts; one that forms the new branch or leaf, and in the language of botanists may be termed persistent; the other serving only a temporary purpose, and falling when that purpose is accomplished. To the former may properly be applied the term germ or gem, and the latter, from its office, may be called hybernaculum. In its leading characters the bud bears a near analogy to the more perfect seed, for the germ very exactly resembles the plumbe; and the hybernaculum, as we shall see, in structure, office, and duration, approaches near to certain cotyledons.

The leaves or scales which constitute the hybernaculum, and which in future we shall denominate the hybernacular leaves, vary much in number, size, and figure, in different buds. Even in the same bud the inner ones are thinner, and much more tender and succulent, than the outer, and are besmeared with a viscid humour which intimately unites them; while the outer ones are commonly hard, hairy, and of a scaly texture. Like the cotyledonous leaves of seeds, those of the hybernaculum sometimes grow for a certain time with the germ, and fall successively at periods more or less distant: they are also not less distinct in figure from those of the germ, than the leaves of the cotyledon are from those of the plumbe. This arises from the peculiarity in the distribution of their vessels, which do not spring from one common central trunk, as in ordinary leaves, but are derived from several distinct fasciculi at the base, like those of the cotyledonous leaf, as seen in that of the gourd, a, fig. 13, Plate XXXIX. According to the manner in which they are disposed or folded up in the bud, botanists have assigned them different names, for an account of which we must refer to their writings. In the opinion of Du Hamel, they all derive their origin from the inner layer of the bark, of which they seem to be only a prolongation. (Phys. des Arb. tome i. p. 103.)

The germ of the bud, which is contained within and

protected by these enveloping leaves, is composed of one or more leaves generally folded and curled, but in some instances open and expanded. At first they are very small, and their form is indistinct, so that the pedicle alone is distinctly visible; from which branch off the vessels that form the middle rib, and are afterwards expanded to construct the lobes of the leaf. From the figure which the germ possesses before its expansion being like that of a keel, its vascular portion at this period has been named carina, and its softer part medulla or pith.

With respect to the particular portion of the branch from which the germ internally derives its origin, opinions have much varied. Some have held that it proceeded from the pith alone; others from the first circle of vessels that immediately surrounds the pith; others from the tender wood alone; and others from the pith, the wood, and the bark conjointly. Grew held this latter opinion. (Anat. of Plants, p. 28.) Malpighi describes the germ as a tender, ligneous substance, formed of vessels and cellular tissue, and surrounded by its proper cortical texture. (Anat. Plantar. p. 45.) According to Du Hamel, it originates from the ligneous texture and the pith (Phys. des Arbres, tome i. p. 103); and Hill considered it to spring from the first circle of vessels alone, but not to carry with it any portion of the pith.

As every germ is composed of the cortical and ligneous textures, it may be said to originate in part from both, as all these writers seem to admit. Sometimes also the pith of the germ is continuous with that of the branch; but in other instances no such connection subsists, and there is nothing in the character of the pith of the trunk that renders it at all essential to the constitution of the germ. That in many instances, also, the germ springs from the first circle of vessels, is most certain; but it is not less certain that buds spring from trees long after this first circle of vessels has lost its vegetative power, or has been entirely destroyed.

In the oak Malpighi describes the entire bud as consisting of many scales enveloping each other, and forming an oval body. When evolution commences, these open and expand, and in part fall; but two generally remain and protect the springing germ for a long time. In fig. 15, Plate XI., is represented one of the hybernacular leaves d of the oak-bud, at the base of which the germ e is placed. The leaf d is described as possessing an oblong form, and is very evidently vascular; the germ is exceedingly small, and is represented as possessing at this period only one fasciculus of vessels. By degrees three fasciculi become apparent, which are continued through the germ, and form three pointed extremities, as in fig. 16. These parts augment, and the vascular fasciculi separate, so as to produce successively the appearances in fig. 17 and 18. At length the small curled leaf f, fig. 19, rising between the two hybernacular leaves g g, is seen to resemble the plumbe of the seed, and to possess a form altogether different from the enveloping leaves, which in appearance resemble more the cotyledonous leaves of certain seeds. In fig. 20 is given a vertical section of the germ, considerably magnified, in which h denotes the pith that occupies the centre: it is inclosed by vascular fasciculi, i, that send off through the bark ramifications to the several little processes that compose the serrated border of the leaf. (Anat. Plantar. tab. 10 and 14.)

But the method of nature in the evolution of buds, continues Malpighi, is not always the same; for the hybernacular leaves do not always waste and fall as those of the germ increase. On the contrary, in many trees, these leaves, especially those about the base of the bud, losing

Of Buds. their primary figure, assume new forms, and are converted into the permanent leaves with which the branch is adorned. Examples of this sort occur in laurel, in the apple-tree, the almond-tree, and many others. In other instances, as in the rose, the permanent leaves seem to be generated out of those of the hybernacle, from the apex of which they emerge; and gradually the latter is changed into a sort of petiole, to the sides of which two slender appendages, the remains of the former hybernacular leaf, adhere. Similar transformations are said to occur in many other plants.

Buds on the ends of branches. The buds, such as they have been described, pullulate variously from the sides of stems and branches, and always above the insertion of the fallen leaves; but they spring also from their extremities, and produce the annual elongation of the branch or stem. The manner in which this takes place, and the appearance which the parts exhibit, have been well illustrated by the observations and dissections of M. du Hamel. In the evolution of the seed, the plume, as we have seen, rises above the earth, and produces the stem, which puts forth leaves. When these leaves fall in autumn, the stem continues, and is terminated by one or more buds. The roots, as already remarked, do not increase in length but at their extremity, and therefore never elongate after the smallest portion of their extremity has been cut off. It is not the same with branches; for the newly formed part of the young shoot actually elongates, especially at its extremity, where it is most succulent and tender,—less in the parts lower down, where it is harder,—and in its more ligneous parts not at all; as Du Hamel, by very simple and decisive experiments, ascertained. (Phys. des Arbres, tome ii. p. 14.)

in horse-chesnut. The appearance, structure, and evolution of a bud, at the extremity of a branch of the horse-chesnut tree, are given by the same ingenious author, whose candour, ability, and success in the prosecution of these curious researches render him worthy to rank by the side of Malpighi and Grew. In fig. 1, Plate XLII., is represented part of the annual shoot of this tree, terminated by its appropriate bud, formed in autumn, and which is the commencement of the next year's shoot. In fig. 2 a vertical section of the same bud is exhibited, in which the number and disposition of the hybernacular leaves that envelope and protect the germ in the centre are displayed. In the stem of the shoot the letter a denotes the pith, which is surrounded by the wood b b, and this again is covered by the bark c c. In fig. 3 is represented a section of part of the bud detached from the woody part of the shoot, and a little magnified, to show that the leaves of the hybernaculum take their origin from the inner portion of the bark.

their structure. Proceeding next to the interior part of the bud, Du Hamel represents it as composed of numerous small leaves (fig. 4), which are more and more minute as we proceed inward, and are covered with fine hairs. In fig. 5 is a branch-bud of the peach, as seen in February, after all the enveloping scales have been removed. It is composed of greenish filaments, ranged nearly as they appear in the figure. When some of these filaments were detached and viewed with the microscope, they appeared toothed at the edges, as in fig. 6, and were covered with hairs. All these filaments were afterwards detached in order to disclose a small body lodged within them, and which appeared to consist of two small leaflets, folded and toothed at their edges, but not covered with hairs. It is represented in fig. 7. It occupied the centre of the shoot, and seemed to be connected with the pith.

In the next figure, 8, is represented the bud of the horse-chesnut in the state of evolution. The letters d d indi-

cate the scales of the hybernaculum thrust aside by the shooting of the germ e, accompanied by two permanent leaves f f. The letter g denotes the place of a second bud. A vertical section of the same bud is shown in fig. 9, from which all the scaly envelopes have been removed. It exhibits an entire shoot of one year's growth attached to part of one of two years. From h to i denotes the growth of two years, and from i to k that of one year, with the germ k in the centre, and the lateral leaves as in the preceding figure. The letter l marks the pith, m m the wood, and n n the bark of the young branch. From l to n the pith is white; from n to o greenish; and towards i it is of a brownish-red colour. From i to k, which marks the extent of the annual shoot, the pith is green and succulent, and at p p it is seen to be prolonged into the lateral branches. The wood of two years' growth, from m to i, is white, and forms a continuous tube round the pith, except where the branches go off. It is covered by another layer so thin as to be scarcely visible, but which will in the end become wood; and this layer is covered by the bark. The ligneous layer of the annual shoot appears to be a prolongation of the new layer of the older one, and, like it, possesses at first an herbaceous character. The cortical layer also seems to be a prolongation of that of the older shoot. As to the pith, though in both shoots it is continuous, it is to be observed that that of the older branch is white and dry, and that of the young shoot green and succulent. (Phys. des Arbres, tome i. p. 117.)

It is thus by the formation of a bud in autumn at the extremity of a branch, and the shooting and growth of that bud in the succeeding spring and summer, that the trunk of the tree and its branches are annually elongated. During the first season the shoot retains in great part its herbaceous characters, but in the second it becomes perfectly ligneous. In the axils of the permanent leaves of the young shoot, the rudiments of new buds become apparent, even in the first season.

It was before remarked that trees receive an additional circular layer every year, and that from these new layers buds successively spring, so that the earliest branches may contain as many ligneous layers as the trunk, and those of later formation a smaller number, according to the year in which they shot forth, and the circle of wood from which they sprang. Combining this growth in breadth with that in length, it will appear, says Du Hamel, that at the base and centre of a tree 100 years old there is wood of 100 years of age; whilst at the exterior part of the same tree, and at the extremities of its branches, there is wood of one year's age only. As the latitudinal growth was before illustrated by a diagram, a similar mode may be adopted to explain the longitudinal increase.

Let fig. 10, Plate XLII., represent in q r the ligneous portion of a tree that has proceeded from a seed in spring, and is observed in autumn. The following spring a second shoot proceeds from the bud r, which reaches as far as s; but at the same time there is a second ligneous layer formed on the first shoot q r, by which its thickness is proportionally augmented; and at the end of the second year the tree has the form and extent of the unshaded portion of the figure q s. The next spring the bud s opens and sends out another shoot to t; and ligneous layers are added as before to the two preceding shoots; and thus the tree is extended from q to t. The fourth year the same processes are repeated, and the tree extends from q to w; and each annual shoot, from the base to the summit of the figure, is seen to be composed successively of four, three, two, and lastly of one layer in thickness. This figure, therefore, illustrates the mode in which trees increase at the same time in height and breadth. The lig-

neous layers may be compared to a series of cones which envelope each other, and which annually augment the diameter of the tree by the two thicknesses of the layers. It shows also that trees grow much more in height than in breadth, and that this growth is effected by the successive formation and evolution of buds at the extremity of the stem, precisely as the first shoot issues from the seed. (Phys. des Arbres, tome ii. p. 50.)

It is well known that if a bud be taken from one plant and inserted beneath the bark of a kindred species, it will grow, and retain in its new situation its original qualities and habits; or some buds, if directly planted in the earth and protected from too rapid exhalation, will produce a tree similar to the parent. Hence some writers, as Dr Darwin, have considered each bud as an individual plant, and that a tree, properly speaking, is a family or swarm of buds, which annually produce as many new plants. The bulbs produced beneath the soil by the tulip and many other plants resemble, in this respect, the buds produced in the atmosphere. Vegetables, therefore, are propagated by two methods—the oviparous, as by seeds, and the viviparous, as by buds and bulbs; and the individual plants, whether springing from seeds or from buds and bulbs, are all annual productions. The reproduction of plants, therefore, appears to be of two kinds, solitary and sexual; the former being accomplished by buds and bulbs, the latter by seeds, or through the agency of sexual intercourse.

SECTION II.

Of the Structure of Leaves.

These are organs of great importance in the vegetable economy; they are not however universal; for the Cactus, some species of Solanum, and a few other plants, are considered to be destitute of leaves. Like other parts of the plant, they may be regarded with reference either to their external form or their internal structure. The former view belongs more especially to the botanist, who has very happily applied his descriptive language to portray the almost infinite diversity of figure, size, and character, which the leaves of different species of vegetables exhibit. On this branch of the subject we shall but lightly touch, recalling simply to notice the leading distinctions of the botanist, and such only of them as may appear to be more immediately connected with the structure of these organs.

The leaves (folia) are distinguished and denominated according as they are simple or compound. Simple leaves are such as have only a single leaf on the stalk or petiole that supports them, and where all the parts of the leaf are continuous with one another. Compound leaves are those which are made up of more than one piece, or where the leaf is formed of parts or leaflets articulated together. In regard to their place, situation, and insertion, leaves are said to be determinate. By the place of a leaf is meant the part of the plant to which it is attached. By situation is understood the disposition of the leaves on the stem or branch, which corresponds to that of the buds. By insertion is expressed the mode of connection between the leaf and the stem or branch; and the direction of leaves is considered to bear reference to the position in which they stand to the stem.

The foregoing observations apply to leaves considered in connection with other members: when we regard them singly, we remark several parts which are common to almost all leaves, and to which particular names have been assigned. The part at which the leaf springs from the branch or stem, whether directly or by means of a pe-

tiolate, is called the base, and the point opposed to this, the apex, of the leaf. The expanded portion is termed limbus, which has its two sides, faces, or surfaces, as they are indifferently called. The prominent lines that appear on these surfaces were named riblets (costule) by Malpighi, and have very improperly been called nerves by most writers; for the term nerve denotes an organ of a totally different nature, of which not even the existence has yet been demonstrated in any part of the vegetable system. To the line that circumscribes and forms the boundary of the leaf the term margin or border is commonly applied.

The figure of leaves exhibits the greatest diversity, to their characters which various terms are employed. Their margin is either entire or variously fissured, notched, or toothed. Their surfaces are naked and smooth, or clothed with hairs and studded with excrescences; sometimes they are plain and flat, at others furrowed or plaited; and in all these particulars the opposite surfaces present also the greatest diversity. In size, leaves exhibit the most remarkable differences; and, with respect to substance, some are exceedingly thin, and have a membranous texture, others very thick, and of a fleshy nature. As to colour, they exhibit every shade of green, from the most gay and lively to the deepest and most obscure. In some the tint of colour approaches to blue, in others to red; and previous to their fall they all undergo those changes of colour which produce the diversified beauties of an autumnal scene. The period, however, at which this occurs is very different. Some leaves fall early, before the summer has passed, and are termed caducous; others retain their place till autumn, and then fall, and are named deciduous; others continue beyond the summer, and are styled persistent; and others, which have a still longer duration, are denominated perennial.

In the preceding section on the structure of buds we exhibited the form and appearance of leaves in the earlier from buds, periods of their existence. They were shown to possess a very minute size, and to be curiously folded up and concealed within their various scales or coverings, which effectually protected them from the rigour of the winter season. The germ of the bud, from which the leaves originate, was stated to be composed of the same elementary parts as the stem and branch from which it sprang. From the lignous ring in the annual shoot it derived its vessels, which, after having traversed obliquely the cortical layers, were prolonged into the pedicle by which it remained attached to the branch. This pedicle was continued to form the keel of the germ in its folded state; and, in its more expanded forms, to constitute the vascular riblets of the leaf. By subsequent growth it gradually extends and elongates, becomes more lignous, and is formed at length into the slender stalk or petiole by which the leaf remains attached to the branch. (Phys. des Arbres, tome i. p. 123.)

The vessels which are given off from the branch to form the petiole are not collected into one bundle, but of the petiole constitute several fasciculi, which are disposed in different ways about the centre of the stalk. In some stalks there are three, in others five or six, and in others seven or more fasciculi, all placed, says Grew, either in an angular or circular posture, and at a greater or less distance from the centre. This centre is generally occupied by a pith, but sometimes it is hollow or tubular. In table 49 of the Anatomy of Plants, several transverse sections of the petioles of the leaves of different plants are given, exhibiting the different modes in which these fasciculi are dispersed through the cellular tissue that forms the greater portion of the stalk. In every instance the pith and bark

Of Leaves. of these petioles are represented as forming one continuous substance. In some plants the fasciculi of vessels that come off from the stem are not collected into a cylindrical figure, but, after being variously implicated with each other, expand at once into a leaf. Such leaves rise by a broad base, and, from being destitute of a petiole, are termed sessile, forming often a sort of sheath about the stem, of which several varieties are enumerated by botanists.

The petiole, as it springs from the branch, may be compared to a small stem, which at the basis of the leaf expands into numerous branches. Frequently it is continued through the centre of the leaf, forming its middle rib, and giving off in its course numerous branches. At other times, on reaching the base of the leaf, it separates at once into three or more equal portions, which form as many distinct leaves or parts of a leaf, supported by their respective petioles. The structure of the petiole is the same as that of the branch, being composed of the ligneous and cortical textures, in which sap and "proper vessels" are discoverable; and the whole is invested by the cuticle.

Where the petiole terminates, the proper or expanded portion of the leaf (limbus) commences, the figure of which is determined by the number and distribution of the vessels. These vessels divide and ramify in various modes, till at their termination they form a finely reticulated structure. In many recent leaves this structure is very visible; but when the softer parts are removed by spontaneous decomposition, the vascular system of the leaf remains nearly entire, and its extent and form are then more completely exposed. In fig. 11, Plate XLI., this distribution of the vessels is exhibited in a leaf of holly. From the central fasciculus branches are everywhere given off, which further subdivide, and form smaller ramifications, that terminate at length in a minutely reticulated structure. Around the margin of the leaf the vessels are continued, and at certain parts are prolonged into the thorns that bound the circumference of this leaf.

In this and similar instances the vessels appear to be ramified out of greater into less; but, as already observed, this does not appear to be really the fact, the vessels being all of the same size everywhere in the leaf, and all continued through it, like so many distinct tubes. This structure Grew represents in a highly magnified view of the leaf of borage, in table 50 of his work; part of which is represented, on a reduced scale, in fig. 13, Plate XLI., designed to show that the ramified vessels are all clusters of tubes of the same size, which, though they separate continually, and come into contact, are never produced or ramified one out of another. Neither do they ever inosculate or anastomose with each other, until, according to Grew, they come to their final distribution.

The vessels thus distributed through the leaf belong chiefly to the order of spiral vessels, as was shown by Grew in the leaf of the vine and many others; and, from the observations of Malpighi, Darwin, and Knight, already related, it appears that "proper vessels" are everywhere associated with them. It would seem, from the observations of Darwin, that these two orders of vessels communicate in the leaf by continuation of canal, as the arteries and veins do in animal bodies. In the stalk he considered them to be disposed in two concentric rings, the inner one of which carried out the sap to the leaves, and by the outer one it was returned to the bark. (Phytologia, p. 43 and 58.) In branches of the apple and horse-chestnut, Mr Knight also observed coloured fluids to rise through the vascular fasciculi in the petiole of the leaf. These fasciculi were surrounded by others free from colour,

which conveyed a different fluid, and, on being traced downwards the stalk, were found to enter the inner bark, and to have no communication with those of the wood. The returning vessels he describes as being situated parallel to and surrounding those which carry up the sap. (Phil. Trans. 1801, p. 336.)

Beside the vascular system which, by its ramifications, forms the skeleton of the leaf, there is another structure that requires to be noticed. We have seen that the vessels, in their ultimate distribution, form a net-work more or less fine and minute in different leaves; so that a great number of intervascular spaces are produced. These spaces or areas, says Malpighi, are occupied by cellular tissue, which springs from the vessels themselves, and seems to depend from them; and by its means the thickness of the leaf is formed. In the recent part of the borage leaf, represented in fig. 13, Plate XLI., this cellular structure, occupying the reticulated spaces formed by the vessels, is observed; but it is most clearly seen in thick leaves, when the cuticle has been removed, and the vessels in part dissected away. The cells or utricles which form this parenchyma, as it has been called, are of different figures, and are mutually contiguous; they are composed of a membrane formed into the shape of a little vesicle or bladder, from the middle of each of which a little vascular production issues; they are all connected with each other, and with the vascular system of the leaf. In fig. 12, Plate XLI., this cellular structure is represented in the leaf of Cactus by Malpighi, in which the oblong cells are described as proceeding from the central vessel, and mutually communicating with each other. (Anat. Plantar. p. 52.) A similar idea of the formation of the cells seems to have been entertained by De Saussure. In different leaves the cells possess very different sizes and forms. In the experiments of M. de la Baisse, they are said to have become tinged by coloured fluids conveyed from the vessels. Whether they communicate with each other is not known; but analogy, derived from other similar structures, would lead us to suppose that they communicate only with vessels, as seems to be the case in the corymbosous leaf of the seed.

M. de Candolle describes the leaf as consisting of a superior and inferior face, between which is an intermediate part, of very various thickness in different leaves, to which he gives the name of mesophyll, answering to the parenchyma of preceding writers. The leaf is formed, he says, by ramifications of the vessels, and by cellular tissue, which fills up the intervals of these ramifications. It probably comprehends two systems; one which, receiving the ascending sap, brings it into contact with the air, and permits the exhalation of its superfluous parts; the other a system which receives the elaborated sap, and reconducts it to the stem, where it serves for nutrition. Physiological phenomena, he adds, prove the existence of these two functions in leaves; but anatomy has not yet distinguished the organs by which they are accomplished. (Organog. Végét. tome i. p. 271, 273.) It has been before observed that M. de Candolle follows Kieser in regarding the alleged spaces between the cells, called intercellular canals, as the organs in which the vegetable fluids are conveyed; but the facts and arguments already stated establish, we think, with sufficient certainty, that, in the wood, the bark, and leaves, the sap is conveyed by appropriate vessels. The consistence of the leaf, continues M. de Candolle, depends much on the relative proportions of the cells and vessels that compose it. When the vessels are numerous, the cellular tissue is small in quantity, and the leaf is fibrous, firm, and thin; when, again, the vessels are few and distant, then

Of leaves, the cells are more abundant, and the leaf is softer and more fleshy.

The two surfaces of this organ often differ much in appearance and structure. On the upper surface the vessels are commonly but little prominent; it is therefore smoother, has but few hairs or pores, and is sometimes destitute of them. The lower surface, on the other hand, presents the vessels more prominent, and is furnished with hairs and pores much more abundantly. These differences are more particularly observed in the leaves of trees: in those of herbs the two surfaces more nearly resemble each other. In aquatic plants, though the upper surface is smooth and green, and the under one pale and rough, yet the former only, as being exposed to the atmosphere, is furnished with pores. It is to be remarked, however, that whether the two surfaces exhibit an aspect either very different or very similar, each seems destined to enjoy a special function. Hence, if the upper surface be turned downward, and the lower one upward, the leaf tends to resume its natural position; and if it be retained by force in a position not natural to it, it dies. (Organog. Végét. tome i. p. 274.)

In the cellular tissue of herbaceous stems, and of all leaves, are contained small globules, which by the action of light are rendered green, and bestow on the plant its verdure. In the elongated cells of Conferva reticulata they are represented by Kieser as small green globules. (See Plate II. of his Mémoire.) They are of a resinous nature, and have been named chlorophylle by chemists; but as a matter similar in composition is said to reside in the flower, M. de Candolle deems the foregoing name improper, and proposes to call this matter chromule. (Ibid. tome i. p. 19.)

In the substance of the leaf Malpighi describes small bags or follicles as existing between the cells and vessels, and which, in most leaves, yield a peculiar humour. They spring from the vessels, have a roundish or oblong form, and the lip that bounds their orifice is often furnished with hairs. This orifice points towards the upper surface of the leaf, or that which regards the heavens. In different leaves the follicles have different forms; and they afford a secretion which, adhering to the contiguous parts of the leaf, renders them shining. They are numerous in the leaf of the fig, lemon, orange, &c.; and are rendered more visible as the cellular tissue of the leaf wastes and dries. (Anat. Plantar. p. 52.)

There are other small leaves called stipule and bractea, which are attached to leaves and flowers, and are of some importance to the botanist. In the general features of their structure they resemble those already described.

The whole structure of vessels and cells that compose the leaf is covered by a cuticle variously furnished with hairs and pores, and other appendages already described.

SECTION III.

ART. I.—Of the Structure of Flowers.

The flower is that assemblage of organs which, by the mutual action they exert on each other, give origin to fruits and seeds. All the parts hitherto described contribute to the growth and perfection of the individual plant, and have been named the Nutritive or Conservative System; but the organs composing the flower are destined to continue the race, and from thence have been denominated the Reproductive Organs. In a limited sense the bud, produced by the conservative system, may be deemed a reproductive organ, since new individuals may successively be propagated from it; but after a certain period, plants thus produced seem to degenerate, and be-

come incapable of producing fruit. Reproduction with-out fecundation presents no organic apparatus which is proper to it; but sexual reproduction, or that accomplished by fecundation, exhibits numerous and varied organs, which merit all our attention, not only because they exercise the principal functions of vegetation, but because, from the constancy and symmetry of their forms in the same species, though varied infinitely in different species, they afford the base on which Classification rests.

In relation to the organs of reproduction, vegetables, as before remarked, have been divided into two great classes: organs in 1st, such as have flowers visible to the naked eye, and in 2dly, such as have flowers (if there be any) which can be rendered visible only by the microscope, and in which the sexual organs are not distinct. The former class, named phenogamous or phanerogamous, comprises all exogenous, and the greater part of endogenous plants; the latter, or cryptogamous division, comprehends some endogenous vascular plants, and all those which possess a simple cellular structure. At present we shall treat only of the first or flowering class.

The flower is essentially constituted by the presence of Varieties one of the two sexual organs, or of two re-united by a in sexual common support, with or without the exterior envelope organs. destined to protect them. In some instances the flower bears only a male or female organ, and is then named unisexual; but more commonly these organs are present in the same flower, and it is then called hermaphrodite.

This unisexual flower, again, is simply male or female, Descrip- according as it possesses either stamens only, or pistil only. tion of In some instances, though the same plant bears both male sexual and female organs, it is not hermaphrodite, since these organs occur in different flowers; in others, again, the male and female flowers exist only in different plants. Lastly, we sometimes find both male, female, and hermaphrodite flowers mingled together, either on the same or on different footstalks. The distinctions founded on this separation, re-union, and mixture of the sexual organs, have served as the basis of certain classes in the Linnæan system.

Sometimes the male or female organs alone, protected in a small scale, constitute the flower; but in general they are surrounded and protected by others, named the corolla and the calix. All these parts are commonly borne on a stalk called the peduncle, which, expanding at its extremity, forms the receptacle, or torus as it has been called, upon which all the parts above mentioned are supported. In fig. 1, Plate XLII., the sexual parts which constitute a perfect flower are exhibited. The letter a denotes the corolla, formed, in this flower, of six leaves called petals; b the stamens or male organs, which surround the pistil or female organ c; and d indicates the peduncle of the flower. In the next figure (fig. 2) the corolla has been removed, and the parts within are then more clearly seen. They now consist only of the sexual organs supported on the peduncle, and at their base are embraced by the calix e. The pistil, in the centre, is composed of three parts, viz. the ovarium f, the style g, and the stigma h; while each stamen consists only of two parts, the filament i, and the anther k, which latter contains the prolific matter named pollen. Such an assemblage of organs constitutes a perfect or complete flower; but it often happens, that in flowers where the sexual organs are present, one or other of the accessory organs is wanting; thus, in the tulip only one of the protecting envelopes is present, and botanists differ as to the name it should bear. Linnæus considers it as a corolla, because it is coloured like ordinary petals; but M. Jussieu regards it as a calix, from its analogy in structure and use to that organ.

Of Flowers. The flower is sometimes attached immediately to the branch by its base, when it is termed sessile; but in general it is borne on a footstalk or peduncle. This peduncle of the flower, like the petiole of the leaf, is sometimes single and sometimes ramified: in the latter case its divisions take the name of pedicels.

Floral leaves or bractæ. Around the base of the flower we often meet with small leaves which, in size, colour, form, and consistence, differ from ordinary leaves. They are called bractæ, and from the axils they form with the stem or branch the flower-buds often spring. From thus being the birth-place of the flower, which draws from them a great part of the nutrient matter, the bractæ are smaller and more membranous than the ordinary leaves of the plant, and partake often of the colour of the flower. They must, however, be regarded as leaves, modified by the circumstances in which they grow, as is demonstrated by the fact, that examples frequently occur in which bractæ are changed into true leaves. (De Candolle, Organog. Végét. tome i. p. 438-9.) When these bractæ approximate so as to form a sort of envelope around the flower, they take the name of involucre,—when this envelope becomes a sheath, it is named spatha,—in the Gramineæ it is termed gluma, &c.; of all which forms there are numerous modifications, to which botanists attach different names. The bractæ approximate more or less in character to the calix, in proportion as they are coloured or exhibit the form of whorls; and the transition of the organs of vegetation into those of florescence is so gradual, says M. de Candolle, that the more we study it, the better do we comprehend that unity of composition which constitutes the base of philosophic organography. (Organog. Végét. tome i. p. 447.)

Flowers originate from buds. The flower, as well as the leaf, originates from a bud, and in many respects the buds of flowers resemble those of leaves. They are covered and protected in the same manner, but they are generally larger, and have a more rounded form. They consist of two parts, the gem or eye (oculus), as it is sometimes called, and the hybernaculum, or protecting envelope. In some trees they spring from the extremity of particular branches; in others from the branches in common with the leaf-buds; and in others from the axils of the leaves. In fig. 3, Plate XLII., is a portion of the branch of the peach-tree, bearing two flower-buds l l, between which the smaller and more pointed leaf-bud m is placed.

Flower-buds formed in autumn. Grew discovered and has exhibited many examples of the complete formation of the flower many months before it is destined to bloom. (Anat. of Plants, tables 63, 64.) Of this fact Du Hamel has also given several examples. In fig. 4 is a longitudinal section of the flower-bud of the peach-tree, made in the month of February, to show the complete formation of the stamens and pistils within the surrounding envelope of the bud. In fig. 5 is another representation of a similar bud, from which the enveloping scales have been removed: it displays the calix, which at this period completely envelopes the other parts of the flower. When the leaflets of the calix are separated, as is done in fig. 6, then the stamens and pistils are fully disclosed: the petals also of this flower were visible, but at this period were very small. Even the anthers were found to contain a fine dust, but no rudiment of the future embryo could be detected in the ovary. In the flower-bud of the pear, examined in February, the parts of fructification were also visible, but indistinct; a month later, all the parts were more advanced, and the stamens, petals, and pistils were distinct, and towards the end of March, even the rudiments of future seeds were visible in the base of the pistil. (Phys. des Arbres, tome i. p. 200.) Thus it appears that through the winter season the several parts

of the flower are clandestinely formed, though still extremely minute. As they enlarge, their position within the bud is very various in different species: they gradually expand and form the perfect flower, whose several parts we have next briefly to describe.

The peduncle of the flower, by which all the other parts are connected with the stem or branch, is frequently single, but often divides into several parts or pedicels, as they are named, each of which supports one or more flowers. The modes in which the flowers are disposed on the pedicels have received different names from botanists, and to the general circumstance they assign the term inflorescence. The peduncle in structure resembles the stalk of the leaf, being formed, like it, of the several common textures. In size, in length, and other external characters, it exhibits all the varieties already noticed in the petiole of the leaf.

At the extremity of the peduncle is placed the flower-cup or calix, which has very different forms, and has received in consequence different names. Even in its most common forms it exhibits great variety. It is sometimes composed of a single piece, and has a tubular form; in others, of many pieces called sepals (sepala), which vary in number, position, and size, and for an account of which we must refer to the writers on botany. In some instances this organ falls when the fruit has set; in others it continues till the fruit is mature; and in others the fruit is formed within it, to which it becomes a permanent covering, and exercises the office of pericarp. The colour of the calix is commonly green, but sometimes partly red, or white, or yellow. In some flowers it is entirely wanting. From the dissections of Grew and Malpighi it appears, like the leaves, to be constructed of a cuticle, a pretty thick cellular tissue, and of vessels, all of which exist in the peduncle, and are derived from the common textures of the plant.

Above the calix, the corolla, the chief ornament of flowers, is borne. It is inclosed by the calix, but surrounds the interior parts of the flower, and is of every colour but green. It consists of one piece, or of several: these pieces are called petals, and according to the number of these pieces the corolla is variously denominated. The monopetalous corolla, or that composed of one piece, is also distinguished into several parts, expressive of form, position, or quality; and in the polypetalous variety, which consists of several pieces, each petal has its claw (unguis) situated at the base, and by which it is attached to the calix or receptacle, and its expanded part (lamina), which is of very various figure, size, and colour. At the base of the petal a tuft of hairs is often observed, and its surface is frequently covered with a fine down, or sometimes with jointed hairs bearing globulets, as occurs on leaves. With regard to structure, Malpighi describes it as composed of the same textures as the common leaf, as is distinctly visible in the thicker varieties of it. Grew showed it to possess spiral vessels,—a proof, as Du Hamel observes, that it is partly derived from the ligneous texture, since such vessels are not found in the bark. The odour these organs possess leads to the belief that they contain also a peculiar juice. (Phys. des Arbres, tome i. p. 215.)

The organ next to be noticed is the nectary (nectarium). In a strict sense this term designates any excretory gland which is situated on any one of the floral organs: and the juice it yields bears the name of nectar. The excretory glands observed on flowers deserve a common designation, because, whatever be their position on any part of the flower, whatever the nature of the juices of the plant, or whatever the size, form, or consistence of the gland itself, the secretion it yields is always saccha-

flowers, and very similar in all known plants. In regular flowers the nectaries may occur on all the organs, symmetrically disposed. Their most common seat is the receptacle or torus. In some flowers, however, they are seated on the ovary, in others on the corolla, calix, stamens, &c. They have the form sometimes of distinct tubercles; but in different plants they exhibit various other forms. In irregular flowers the nectaries are not thus symmetrically placed, but are found on one part and not on the corresponding part of the same flower. Sometimes they supply the place of abortive stamens or pistils. The juice they secrete is greedily sought by insects. (Organog. Végét. tome i. p. 535.)

ART. II.—Of the Sexual Organs of Flowers.

Having exhibited this general view of the flower, we must now describe somewhat more particularly the forms, characters, and structure of its most important parts—the sexual organs.

The male organ or stamen consists, as before remarked, of two parts,—the anther, with the pollen it contains, and the filament on which the anther is borne. The filament, though a common, is not an essential part, and is sometimes wanting, when the stamen is said to be sessile: the perfection of the stamen resides, therefore, in the anther, and especially in the pollen it contains.

In regard to number, length, position, direction, &c. the stamens exhibit great diversity in different plants. The number of stamens in each flower varies from one to twenty or more; and on this number, progressively increasing in flowers truly hermaphrodite, the first ten classes in the Linnæan system are founded. In length, the stamens are equal or unequal; and this disproportion is sometimes symmetrical and sometimes not,—peculiarities which give rise to the formation of other classes in the Linnæan arrangement. In position, the stamens are sometimes opposed to the divisions of the petals, and sometimes they alternate with them. Sometimes, from being shorter, the stamens are wholly included within the corolla: in other instances they protrude beyond it. Their direction is erect, pendent, or horizontal; and their summit is variously inclined to or reflected from the centre of the flower. In some flowers the stamens are united by their filaments; in others by their anthers; and in others they are joined to and almost confounded with the pistil, giving origin, in their various modes and degrees of union, to the formation of other classes in the Linnæan system.

The filament which bears the anther is most commonly straight and filiform: sometimes, however, it is flattened or wedge-shaped, or awl-shaped. It is sometimes as small as a hair, sometimes large and flat like a petal, and its summit is either pointed or obtuse. To the summit the anther is commonly attached; but in some flowers this part is prolonged beyond the attachment of the anther. In structure, the filament resembles the corolla; and it often happens that these organs are transformed into one another. Thus the rose, in its wild state, has only five petals and many stamens; but by cultivation the stamens gradually degenerate into petals, and the flower becomes sterile. Malpighi describes the filaments as originating from the ligneous texture, being formed of spiral vessels and elongated cells; and as they are sometimes produced from elongated floral leaves (petals), they must necessarily, he adds, be formed of the same parts. (Anat. Plantar. p. 64.)

The anther is that essential part of the stamen that contains the fecundating matter. It is borne on the summit of the filament, and is generally formed of two small

membranous sacs, attached immediately to each other, or united by an intermediate connecting body. Anthers, therefore, are generally bilocular; but in some plants there is only one sac or loculum, and in others as many as four. The sacs are either round, oval, or elongated: each sac commonly exhibits on one of its sides a longitudinal furrow: this side is properly called the face, and the part opposed to it the back, of the anther. The anther is attached to the summit of the filament, sometimes by its face, sometimes by its back, &c. In form, the anther is subject to great variety; and the two sacs that compose a bilocular anther are joined together in very different modes. In discharging the pollen the sacs open in different ways in different genera of plants, most commonly by the longitudinal furrow that runs along the face of the anther; but in some plants the opening, or dehiscence as it is called, is through two small holes or pores situated at the summit of the two sacs; and sometimes it is by a sort of little valves in some one or more parts of the anther. As the filaments of the stamens sometimes cohere, so likewise do the anthers, in a manner to form a sort of tube. In other plants the stamens, instead of being free or simply united together, coalesce so as to make one body with the pistil. The colour of the anther is often yellow, orange, violet, white, &c., but never green or truly blue.

The pollen, contained in the anthers, consists of numerous regularly figured small particles, which possess a very different figure, size, and colour in different plants. Many of their forms and sizes are delineated by Grew in table 58 of his Anatomy of Plants. In fig. 9, A, Plate XLII, copied from Grew, one of the filaments, bearing its anther, as it appears when detached from the pistil of the mallow, fig. 8, is exhibited. The particles of pollen are considerably magnified, and still more highly in fig. 9, B. In fig. 10 of the same plate, several representations of the forms of the pollen, as given by Du Hamel, are also exhibited. The number of these particles in each anther extends, it is said, from a few hundreds to many thousands. In the cell of the anther they are said to float in a viscous liquid, and to be nowhere attached to its sides, though it is probable they once were; and M. Turpin has even designated a salient point in each cell as the part which produced the pollen. (De Candolle's Organog. Végét. tome i. p. 465.)

In some flowers the pollen consists of transparent grains; in others they are of a white, purple, blue or brown, and more frequently of a yellow colour. Their surface is either smooth or rough, or covered sometimes with a viscous matter. When examined under the microscope at the period of maturity, they may be seen to burst, and yield a fluid, in which, according to Du Hamel, small granules are seen to float. This fluid is described as being thin, or viscid, or oily. If a grain of pollen be thrown on the surface of water, it may be seen to swell insensibly, and finally burst. At this moment a minute quantity of fluid matter escapes, which spreads on the surface of the water, and forms a sort of slight cloudiness. It is to this liquid matter that the fecundating property of the pollen has been attributed. (Éléments de Botanique, par Achille Richard, p. 199.) This fluid is said by Hedwig to be discharged at once on the bursting of its containing capsule; but Koelreuter considers it to be slowly transmitted through pores in the side, or hairs on the surface of the capsule.

Instead of being formed of distinct granules, the pollen is sometimes met with in a solid mass. In many genera of the families Apocynæa and Orchidæa, all the pollen contained in one sac is united into one body, having the form of the containing sac; in other instances it forms smaller

Of Flowers. masses, and these are united by a sort of elastic net-work; while in others the masses are granular. The pollen is very inflammable, and in many plants its odour is said to bear a striking resemblance to the corresponding secretion in animals. (Éléments de Botanique, par Achille Richard, p. 200.)

Sponta- It has lately been said by some, that the particles con-
neous mo-tained in the grains of pollen exhibit spontaneous motions,
tion in the like the spermatic animalcules of animals; others ascribe
pollen. these motions to causes distinct from life; and others,
though they allow similar motions to be exhibited by the
molecules of unorganized matter, maintain, nevertheless,
that in the granules of pollen they are independent of
physical causes, and resemble the less rapid motions of
some of the simplest animalcules of infusions.

The pistil. The parts of the flower hitherto described are con-
structed, says Malpighi, with reference to the female or-
gans, in which the seed, the last result sought by nature,
is curiously formed and matured. This organ, called the
pistil (pistillum), consists of three parts. These are re-
presented in the pistil of the almond, fig. 11, Plate XLII,
where the letter p indicates the stigma, q the style, and r
the ovary. To this latter organ Malpighi and Grew gave
the name of uterus, Linnæus that of germen, and Gærtner
that of ovarium. In different plants the pistil exhibits
great variety in form, size, number, and mode of attach-
ment; upon which many of the orders in the Linnæan
system are founded.

The ova- The ovarium occupies almost always the inferior part
rium. of the pistil. It is the organ in which the seed is pro-
duced. When cut open it exhibits one or more cavities or
cells, in which are contained the rudiments of the seeds
or ovula; and it is in it that the change of the ovula into
perfect seeds is accomplished. Its form is various, but
most commonly ovoidal. It is seated commonly on the
receptacle, together with the stamens; but frequently it
is placed below the flower. Its cavity consists sometimes
only of one cell or loculum, in which one or more
ovula are found. More frequently there are two or more
cells containing ovula: these are sometimes disposed in
regular series, sometimes they are scattered and without
order, and sometimes they are united to one another in a
globular form. When these ovula are fecundated they
become seeds; but in some plants a certain number are
constantly abortive.

The style. The style is a prolongation from the summit of the ova-
rium, and supports the stigma. It is commonly so situate
in the flower as to be surrounded by the stamens; but
sometimes it is entirely wanting, and the stigma is then
said to be sessile. The ovary in different plants may be
surmounted by one or more styles; while in other plants
there is but one style to many ovaries. Most frequently
the style occupies the summit of the ovary, but sometimes
it springs from its side, and very rarely from its base.
The forms of the style are numerous:—the most common
is the filiform; but frequently it is thick, angulated, or
club-shaped. It is commonly a hollow tube which com-
municates with the ovary: there is usually but one style
to one ovary, but sometimes more. In some instances the
pistils correspond in number with the loculments into
which the ovary is divided; in other instances every
seed that is formed has its distinct pistil, while in other
examples only one pistil is allotted to a great number of
seeds. In some flowers the style is so connected with the
ovary that it falls after fecundation; in others it con-
tinues after that event, and forms a part of the future
fruit.

The stig- The stigma, which forms the summit of the pistil, some-
ma. times terminates the style by an open mouth. Sometimes

it appears like a small bud; in other instances it is va-
riously divided or forked. Sometimes it is smooth; and
is sometimes covered with hairs. The number of stigma-
ta is determined by that of the styles, or divisions of the
styles, and therefore varies from one to six or more in dif-
ferent plants. Sometimes the stigma is attached to the
summit of the ovary, without the intervention of a style;
but most frequently it is seated on the summit of the style,
or occasionally attached to its side, or to that of the ovary.

In consistence, the stigma is thick and fleshy, or thin
and membranous, or formed of small glandular bodies.
In form, it varies exceedingly; and its surface is smooth,
pubescent, or plumous, &c. In structure, it is ordinarily
glandular; and at the period of fecundation its surface is
rendered moist by a peculiar viscous matter. When the
pollen falls on this matter, its grains burst, and the gra-
nules, called by some favilla, are supposed to be absorbed
by the spongioles seated in this part, and conveyed to the
ovula by the vessels which form the pistillary cord, and
thus accomplish their fecundation. M. Bulliard is said to
have traced coloured liquors, previously absorbed by the
stigma, along the vessels in the interior of the style, to
the ovula contained in the cells of the ovary. (De Can-
dolle, tome i. p. 480.) In many plants, however, either
from the structure of the pistil, the nature of the pollen,
or the circumstances in which fecundation is effected, the
pollen seems never to come into contact with the ovula.

SECTION IV.

Of the Structure of Fruits and Formation of Seeds.

In the preceding section we described the structure of
the flower antecedent to fructification: we have now to
exhibit as concisely as possible the changes of form it
undergoes after that event, and particularly as it regards
the production of the seed.

After fecundation has been effected, the calix, corolla,
stamens, and even style of the pistil, commonly fade and
fall; the ovary alone remains, and undergoes very differ-
ent changes of form in different plants. In the latter pe-
riods of its enlargement it is usually called pericarp (peri-
carpium), a term which is understood by botanists to ap-
ply also in certain cases to the calix, the corolla, or any
other apparatus of organs that serves as a support and
defence to the seed.

In its early state the ovary is described by Gærtner as
possessing at first a simple cellular structure, which at a
later period takes the form of distinct cavities or locu-
laments. Within these cavities minute globules are af-
terwards seen, which are the rudiments of future seeds. Ac-
cording to Mr R. Brown, the ovulum, in the unimpreg-
nated ovary, is attached only to a part of the mem-
brane that lines its cell or internal cavity; but soon after
fecundation in some cases, and still oftener during the
growth of the seed, this membrane coheres so closely with
the proper coat of the seed as to be no longer distinguish-
able or separable from it. In the ovary of plants of the
family Compositeæ he observed two slender filiform cords,
which, originating from the base of the ovulum or its
short footstalk, ran up, and were more or less connected
with the parietes of the ovary, until they united at the
top of its cavity immediately under the style, between
which and the ovulum a connection was then formed. In
some species of this family, as in tussilago odorata, these
cords were easily separable from the ovary, and could
be removed entire along with the ovulum. These cords
he regards rather as vessels conveying nutrient mat-
ter, than as organs by which impregnation was accom-
plished. A similar appearance of vascular cords con-

meeting the ovulum with the style was observed in the family Brunonia; and in certain liliaceous plants the bulb-like seeds, which separate from the plant before the embryo becomes visible, are supplied with distinct spiral vessels, which, entering at the umbilicus, ramify regularly through the fleshy mass, and appear to have a relation to the central cavity, where the embryo is afterwards formed. (Linnaean Trans. vol. xii.) Thus, then, the ovulum in the plant, like the egg in the animal, is, to a certain extent, produced without fecundation; but if that function be not performed, it soon degenerates and wastes. The progressive changes that occur in the ovary itself, and in its contained ovula, subsequent to impregnation, have been observed by Malpighi in several plants, whose observations, as they relate to the almond, we shall briefly detail.

In fig. 11, Plate XLII., is represented the pistil of the almond, as it appears soon after fecundation, in which the ovary r is seen to be somewhat enlarged, and has an oval form. In the next figure (fig. 12) a longitudinal section of the same pistil is exhibited, exposing the cavity of the ovary s, within which a small vesicle, t, is placed. As the ovary (fig. 13) enlarges, it becomes rounder, and its style is contorted and diminished in size. If in this stage it be laid open, as in fig. 14, its vessels, which in the peduncle w are disposed cylindrically, are observed to be dilated at the place of the calix x, and to give off branches to the ovary itself, and to the shell y that now begins to be formed within it. In the centre, the ovulum z is now seen to be much increased in size. All the parts continue to augment; the ovary (fig. 15) becomes rounder, and the appearance of the style is obliterated. On exposing its cavity in this stage, as in fig. 16, the ovulum a in the centre is observed to be much increased, and the outer layer b of the shelly covering that invests it now begins to harden. Such are the changes of form and structure exhibited by the ovary: we have next to trace more minutely those of the ovulum that is produced within it.

The earlier appearances of this body have been exhibited in fig. 12 and 14. When removed from its seat, a few days after fecundation has been accomplished, and viewed by a moderately magnifying power, it exhibits the form and appearance represented in fig. 17. Externally it is covered by a vascular tunic, c, derived from the inner coat of the ovarium, a part of which coat adheres to it at d. If in this stage the ovulum be laid open by a vertical section, as in fig. 18, it is seen to be composed of two tunics or sacs, one within the other; the inner one is filled with a cellular tissue, that contains a transparent juice. To this inner tunic the term chorion may be properly applied.

At a period a little later, when the ovulum is examined, a tubular body (e, fig. 19) is observed to extend through the chorion or tunic last mentioned. Shortly after, this tube expands at its apex, and is found to contain a small vesicle. In fig. 20, which represents a section of the entire ovulum, the outer tunic, the chorion, and the tube f expanded at its apex, are exhibited. To it the appellation of amnios may be given; for it is the organ in which the embryo, or corculum, as at this early period it has been called, is first seen to emerge. Through several successive days the expanded portion of this tube enlarges and forms a sort of sac, which is filled with cellular tissue, and the summit of which, says Malpighi, the embryo is seen to occupy. In fig. 21 this amnios is separated from the other tunics, and at its summit the embryo g is observed. If removed from its seat, the embryo presents the appearance k (fig. 22), and when expanded, as in fig. 22, i, is seen to consist of a body and two little wings.

Having thus viewed the several parts of which the ovulum is composed in their separate state, let us next observe them in connection, and trace the series of appearances they exhibit, and the effects they produce on each other. In figure 23 is given a vertical section of the entire ovulum in a more advanced state. The outer coat k still envelopes the others; the embryo l occupies the summit of the amnios m', whose lower part, still tubular, is continued through the chorion n'. In the next figure (fig. 24), from which the outer coat has been removed, the embryo o' and the amnios p' are represented as enlarged; but the chorion q' is partly exhausted of its juice, and has fallen down in a collapsed state. At this period the embryo, when separated from the amnios, has the form r', fig. 25, and in its expanded state is represented by the letter s' of the same figure. The bulk of the embryo t', fig. 26, continually augments, and encroaches on the capacity of the two tunics v', x', whose forms are constantly changing; and from being successively emptied of their juices, with which the embryo becomes filled, they are gradually pressed downward. At last the embryo y', fig. 27, is so much augmented as to fill the cavity of the outer tunic; and by this time the amnios and chorion, exhausted of their fluids, exhibit the shrunk and corrugated forms in which they appear at the bottom of the figure. According to this representation, the outer tunic k', fig. 23, derived from the ovary itself, and the fine membrane that immediately invests the embryo, form the only permanent coverings of the mature ovum or seed; for during the progress of formation the chorion and amnios (which are successively produced subsequent to fecundation) are again obliterated by the growth of the contained embryo. Malpighi describes the process of formation in many other seeds to be nearly similar; for his descriptions of which we must refer the reader to his work (Anat. Plantar. p. 71).

In the above descriptions of Malpighi, the several parts seem to be clearly exhibited, except in one important particular, namely, the situation and course of the umbilical cord. In almost every instance he designates the tube, which we have represented as the first form of the amnios, as the umbilical vessel (vasculum umbilicale), which the subsequent appearances it exhibits show to be erroneous. In the descriptions of Grew this deficiency in the representations of Malpighi is supplied. He has particularly observed the formation of the seed in the apricot, which in many respects resembles that of the almond; and we shall subjoin an abridged account of his observations.

In this fruit the pericarp that envelopes the seed is in the seen, in its mature state, to be composed of the pulpy part a, fig. 28, within which is the osseous envelope b, and at the centre the kernel or true seed c. At an early period both the pulp and stone are observed to consist of cellular tissue; and through the stone the vessels passing from the peduncle are continued. At the base of the figure the letter d denotes one fasciculus of vessels continued through the stone, and turning inward, where it reaches the apex of the seed. These vessels form the umbilical cord or seed-branch of Grew, while the fasciculus e that runs on the opposite side is continued to the flower. In fig. 29 a vertical section of the ovulum, as well as pericarp, is exhibited as it appears at a very early period; in which f denotes the pulp, g the stone through which the umbilical vessels pass and enter the outer tunic k of the ovulum, around which they make a ring. Within this tunic is another, i, filled with cellular tissue; and through its axis a small tube extends, at the apex of which the embryo k is first seen to emerge.

In fig. 30 these several parts of the ovulum are exhibited on a larger scale; the letter l denotes the outer tunic that immediately lines the stone; m the inner one, corresponding to the chorion of Malpighi; and n the tube answering to the amnios of the same author. Through the outer tunic Grew represents the umbilical vessels to pass and be continued to the middle tunic, the cavity of which is occupied by large cells that contain a pure lymph. At first this tunic is entire, but soon there appears in it the small duct n. This duct is not at first wider than a hair, and is dilated at each extremity into an oval cavity that contains a pure lymph. A few days after a soft node is seen to emerge in the upper cavity of this tube. This node (o, fig. 31) is described to possess a conical figure, and to be another tunic filled with very minute cells. It is at first entire, but when about the size of a carraway-seed it becomes a little hollowed near its apex, at which part the vessels enter and terminate in another very small node (fig. 32), which is the first appearance of the embryo of the seed. This embryo, when about one fifth part as big as a cheese-mite, begins to be distinguished by a little fissure, which marks the division of the lobes, as in fig. 33. When the lobes have increased, and are more fully formed, the node contracts at its base (fig. 34), indicating the place of the umbilical cord, which subsequently becomes the radicle of the seed. (Anat. of Plants, p. 209.)

This description corresponds nearly with that of Malpighi, as far as regards the situation and general form of the tunics, and the place in which the embryo is seen first to emerge. It also displays the course of the vessels to form the umbilical cord; but the growth of the embryo in this seed does not seem to produce the obliteration of some of the tunics in the manner delineated by Malpighi.

An example of a different kind is observed in the pear. Its structure has been described by Grew, and more minutely by Du Hamel. In its mature state it consists of a pulpy matter, in the centre of which are five loculaments that contain each two seeds. These appearances are exhibited in the transverse section, fig. 35; and in the longitudinal section, fig. 36, the seeds are further shown to be attached by a small umbilical cord. The pulp of the fruit is made up of a very fine cellular tissue, filled with the proper substance of the fruit, and is everywhere furnished with vessels. Through this pulpy matter a number of solid particles are met with, which are more particularly accumulated at the top and about the core. They are formed of an assemblage of small particles, of a stony consistency, with which a little knot of vessels (fig. 37) is everywhere connected. In fig. 38 is a thin transverse slice, showing the relative position of these stony particles, as indicated by the knots of vessels with which they are associated. The stony matter is not observed at an early period, but seems to be deposited from the juices in a more mature state.

By long maceration in water the pulpy matter is dissolved, and the vascular system is obtained separate. In the peduncle of the fruit fifteen principal fasciculi of vessels are contained. Ten of these are distributed to the seeds and flower, and the five others are dispersed through the pulp. This vascular structure is represented in fig. 39 after the removal of the pulpy part; the larger vessels embrace the core, and, after variously ramifying, terminate in the little vascular processes before described as connected with the stony matter of the pulp. In fig. 40 is represented one of the loculaments of the capsule, with the seed in it, receiving vessels from fasciculi continued from the peduncle; and in fig. 41 an entire seed is represented, and also a section of the same, in which the umbilical vessels that enter at the base are shown, as in

other instances, to be continued beneath the tunic to the apex of the seed. (Phys. des Arbres, tome i. p. 242.)

The last variety we shall notice in the formation of the seed is that of wheat (triticum), as given by Malpighi. In fig. 42 is represented the pistil of the flower of this plant, consisting of the ovarium g, the two styles r, and the feathered parts that form the stigmata. Previous to fecundation, the ovarium is found to contain a little vesicle, fig. 43, which is the rudiment of the future ovulum. After fecundation the styles soon fall, and the ovarium acquires a more pointed figure, as in fig. 44. If now it be opened, the little vesicle has changed its appearance, and contains within it another smaller vesicle, u, fig. 45. The ovarium continues to alter its shape, and assumes a more oblong form, fig. 46, and the appearance of styles is now quite obliterated. Gradually the little vesicle is formed into a small plantule, convex anteriorly, fig. 47, but more hollowed within, fig. 48, and which is situated at the base of the ovarium. The two portions thus described in figures 47 and 48 are the minute germ and cotyledon of this seed, which are represented in their appropriate place in fig. 49, in which the letter x denotes the germ, resting in the concavity of the cotyledon; y the albumen that forms the chief bulk of the seed; and z the ovarium, which in this seed continues permanent, and forms its outer tunic. (Anat. Plantar. p. 73.)

In his description of the progressive changes exhibited by the impregnated ovulum, Gaertner follows Malpighi and Grew, and refers to the figures given by those authors as illustrative of them. The liquor, however, in which the embryo is first seen, he confounds with the sac that contains it, calling it the amnios; but the term amnios designates the membrane or sac only, and not the fluid contained within it. Sometimes, he says, there is no proper sac for this fluid, but it is contained in the chorion, or in an appropriate cellular tissue. (De Fructibus, &c. tom. i. p. 60.) The changes in the ovary during the formation of the seed have likewise been observed by Mr Keith, whose account of them agrees with that of Malpighi and Grew. (System of Physiological Botany, vol. ii. chap. 8.) Very recently, also, the same subject has occupied the attention of several foreign writers, and particularly of M. Mirbel, whose observations we shall briefly detail.

According to M. Mirbel, little has been done to advance this part of physiology since the days of Grew and Malpighi, till the late observations of M. Schmitz and Mr R. Brown. The latter remarked that we ought not to judge of the structure of the ovulum from that of the developed seed. Acting on this idea, M. Mirbel applied himself to examine the ovulum, from the moment the ovulum begins to appear, and followed its progressive changes of development in plants of the same species. The ovulum at first he describes as a small pulpy excrescence of a conical shape, attached to the containing cell, and which does not appear to have any proper envelope or aperture of any kind. In plants whose ovaries produce many seeds, the ovula exhibit different degrees of development, according as they are nearer to or more distant from the vascular cords that supply nutriment. M. Schmitz, it is said, first discovered in the apex of each ovulum a small orifice or hole, which penetrates the two membranes or sacs, of which, at this early period, the ovulum is composed. To these sacs, which are placed one within the other, M. Mirbel has given the names of primine and secondine. The primine orifices at their apex, which at first form but one aperture externally, soon become more distinct from the enlargement of the parts within, and gradually a third membrane is seen to protrude through them. No sooner has it become visible in the form of a small pulp, than it begins,