Language.—Origin of Language: Remarks on the Idiom and Genius of Language.

Origin of Language. In the proper sense of the term, signifies the expression of our ideas and their various relations by certain articulate sounds, which are used as the signs of those ideas and relations. By articulate sounds are meant those modulations of the voice, or of sound emitted from the thorax, which are formed by means of the mouth and its several organs, the teeth, the tongue, the lips, and the palate. In a more general sense, language is sometimes used to denote all sounds by which animals of any kind express their particular feelings and impulses in a manner that is intelligible to their own species.

Nature has endowed every animal with powers sufficient to make known those sensations and desires with which it is necessary, for the preservation of the individual or the continuance of the kind, that others of the same species should be acquainted. For this purpose, the organs of all vocal animals are so formed as, upon any particular impulse, to utter sounds, of which those of the same species instinctively know the meaning. The summons of the hen is instantly obeyed by the whole brood of chickens; and in many others of the irrational tribes a similar mode of communication may be observed between the parents and the offspring, and also between one animal and another. But it is not amongst animals of the same species only that these instinctive sounds are mutually understood. It is as necessary for animals to know the voices of their enemies as those of their friends; and the roaring of the lion is a sound of which, previously to all experience, every beast of the forest is naturally afraid. Between these animal voices and the language of men, however, there is very little analogy. Human language is capable of expressing ideas and notions, which there is every reason to believe that the mind of the brutes cannot conceive. "Speech," says Aristotle, "is made to indicate what is excellent and what is inexpedient, and, in consequence of this, what is just and unjust. It is therefore given to men, because it is peculiar to them, that of good and evil, just and unjust, they only, with respect to other animals, possess a sense or feeling." The voices of brutes seem intended by nature to express, not distinct ideas or moral modes, but only such feelings as it is for the good of the species that they should have the power of making known; and in this, as in all other respects, these voices are analogous, not to speaking, but to weeping, laughing, singing, groaning, screaming, and other natural and audible expressions of passion or appetite. Another difference between the language of men and the voices of brute animals consists in articulation, by which the former may be resolved into distinct elementary sounds or syllables; whereas the latter, being for the most part inarticulate, are not capable of such a resolution. Hence Homer and Hesiod characterize man by the epithet μεταφορικός, or voice-dividing, as denoting a power peculiar to the human species; for, though there are a few birds which utter sounds that may be divided into syllables, yet each of these birds utters but one such sound, which seems to be employed rather as a note of natural music than for the purpose of giving information to others; and hence, when the bird is agitated, it utters cries which are very different, and have no articulation.

A third difference between the language of men and the significant cries of brute animals is, that the former is the product of art, the latter derived from nature. Every human language is learned by imitation, and is intelligible only to those who either inhabit the country where it is vernacular, or have been taught it by a master or by books. But the voices in question are not learned by imitation; and being wholly instinctive, they are intelligible to all the animals of that species by which they are uttered, though brought together from the most distant countries on earth. That a dog, which had never heard another bark, would notwithstanding bark himself, and that the barking or yelps of a Lapland dog would be instinctively understood by the dogs of Spain, Calabria, or any other country, are facts which do not admit of doubt. But there is no reason to imagine that a man, who had never heard any language spoken, would himself speak; and it is well known that language, the language spoken in one country is unintelligible to the natives of another country, where a different language is spoken. Indeed, it seems obvious, that were there any instinctive language, the first words uttered by all children would be the same, and that every child, whether born in the desert or in society, would understand the language of every other child, however educated or however neglected. Nay more, we may venture to affirm that such a language, though its general use might, in society, be superseded by the prevailing dialect of art, could never be wholly lost; and that even of one country would find it difficult, far less impossible, to communicate the knowledge of his natural and most pressing wants to the men of any other country, whether barbarous or civilized. The exercise of cultivated reason, and the arts of civil life, have indeed eradicated many of our original instincts, but they have not eradicated all. (This is doubtful. It is generally believed that the dog, in its natural state, howls; it being only where contact with dogs, the companions of man, has (either directly or indirectly) taken place, that it barks. Again, children do form their earliest words alike, and that independent of imitation.—a. o. l.)

There are external indications of the internal feelings and desires, which appear in the most polished society, and which are confessedly instinctive. The passions, emotions, sensations, and appetites, are naturally expressed in the countenance by characters which the savage and the courtier can read with equal readiness. The serene look, the smoothed brow, the dimpled smile, and the glistening eye, denote equanimity and good will, in terms which no man can mistake. The contracted brow, the glaring eye, the sullen gloom, and the threatening air, denote rage, indignation, and defiance, as plainly and forcibly as revilings or imprecations. To teach men to disguise these instinctive indications of their temper, and

To carry smiles and sunshine in their face, When discontent sits heavy at their heart,

constitutes a great part of modern manners. Yet, in spite of every effort of the utmost skill, and of every motive resulting from interest, the most consummate hypocrite, or the most knavish politician, is not always able to prevent his real disposition from becoming apparent in his countenance. He may, indeed, by long practice, acquire a great command over his temper, and the instinctive signs of it; but at times nature will predominate over art, and a sudden and violent passion will flash in his face, so as to be visible to the eye of every beholder. If these observations be true, and we flatter ourselves that no man will call them in question, it seems to follow, that if mankind were prompted by instinct to use articulate sounds as indications of their passions, affections, sensations, and ideas, the language of nature could never be wholly forgotten, and that it would sometimes predominate over the language of art. Groans, sighs, and some inarticulate lively sounds, are naturally expressive of pain and pleasure, and equally intelligible to all mankind. The occasional use of these no art can wholly banish; and if there were articulate sounds, naturally expressive of the same feelings, it is not conceivable that art or education could banish the use of them; merely because, by the organs of the mouth, they are broken into parts and resolvable into syllables.

It being thus evident that there is no instinctive articulated language, it has become an inquiry of some importance, how mankind were first induced to fabricate articulate sounds, and to employ them for the purpose of communicating their thoughts. Children learn to speak by insensible imitation; and when advanced some years in life they study foreign languages under proper instructors. But the first men had no speakers to imitate, and no foreign language to study. By what means, then, did they learn to speak? On this question only two opinions can possibly be formed. Either language must have been originally revealed from heaven, or it must be the fruit of human invention. The latter opinion is strongly supported by Lord Monboddo in his very learned and able work on the Origin and Progress of Language. But he candidly acknowledges, that if language was invented, it was of very difficult invention, and far beyond the reach of the greatest savages. Accordingly he holds, that though men were originally solitary animals, and had no natural propensity to the social life; yet, before language could be invented, they must have been associated for ages, and have carried on in concert some common work. Nay, he is decidedly of opinion, that before the invention of an art so difficult as language, men must not only have herded together, but also formed some kind of civil polity, existed in that political state a very long time, and acquired such powers of abstraction as to be able to form general ideas. But it is obvious, that men could not have instituted civil polity or carried on in concert any common work, without communicating their designs to each other; and there are four ways by which the author thinks that this might have been done before the invention of speech, viz.,—1st, Inarticulate cries, expressive of sentiments and passions; 2d, Gestures, and the expression of countenances; 3d, Imitative sounds, expressive of audible things; and, 4th, Painting, by which visible objects may be represented.

Of these four ways of communication, it is plain that only two have any connection with language, inarticulate cries and imitative sounds; and of these the author abandons the latter as having contributed nothing to the invention of articulation, though he thinks it may have helped to advance its progress. "I am disposed," says he, "to believe, that the framing of words with an analogy to the sound of the things expressed by them, belongs rather to languages of art, than to the first languages spoken by rude and barbarous nations." It is, therefore, inarticulate cries only that must have given rise to the formation of language. Such cries are used by all animals who have any use of voice, to express their wants; and the fact is, that all barbarous nations have cries expressing different things, such as joy, grief, terror, surprise, and the like. These, together with gestures and expressions of the countenance, were undoubtedly the methods of communication first used by men. We have but to suppose, says our author, a great number of our species carrying on some common business, and conversing together by signs and cries; and we have men just in a state proper for the invention of language. For, if we suppose their numbers to have increased, their wants would also increase; and then these two methods of communication would become too confined for that larger sphere of life which their wants would make necessary. The only thing, then, that remained to be done, was to give a greater variety to the instinctive cries; and as the natural process is from what is easy to what is more difficult, the first variation would be merely by tones from low to high, and from grave to acute. But this variety could not answer all the purposes of speech in society; and being advanced so far, it was natural that an animal so sagacious as man would go on farther, and come at last to the only other variation remaining, namely, articulation. The first articulation would be very simple, the voice being broken, and distinguished only by a few vowels and consonants. And as all natural cries are from the throat and larynx, with little or no operation of the organs of the mouth, it is natural to suppose, that the first languages were for the greater part spoken from the throat; that what consonants were used to make the cries were mostly guttural; and that the organs of the mouth would at first be very little employed.

From this account of the origin of language it appears, that the first sounds articulated were the natural cries by which men signified their wants and desires to one another, such as calling one another for certain purposes, and other such things as were most necessary for carrying on any joint work; then in process of time other cries would be articulated, to signify that such and such actions had been performed or were performing, or that such and such events had happened relative to the common business. The names of such objects as they were conversant with would be invented; but as we cannot suppose savages to be deep in abstraction, or skilful in the art of arranging things according to their genera and species, all things, however similar, except, perhaps, the individuals of the lowest species, would be expressed by different words not related to each other either by derivation or composition. Thus would language grow by degrees; and as it grew, it would be more and more broken and articulated by consonants; but still the words would retain a great deal of their original nature of animal cries. And thus things would go on, words unrelated still multiplying, till at last the language would become too cumbersome for use, and then art would be obliged to interpose, and form a language upon a few radical words, according to the rules and method of etymology.

Those who think that language was originally revealed from heaven, consider this account of its human invention as a series of mere suppositions hanging loosely together, and the whole suspended from no fixed principle. The opinions of Diodorus, Vitruvius, Horace, Lucretius, and Cicero, which are frequently quoted in its support, are in their estimation of no greater authority than the opinions of other men; for as language was formed and brought to a great degree of perfection long before the era of any historian with whom we are acquainted, the antiquity of the Greek and Roman writers, who are comparatively of yesterday, gives them no advantage in this inquiry over the philosophers of France and England. Aristotle has defined man to be a λόγος πράγματος, or imitative animal; and the definition is certainly so far just, that man is much more remarkable for imitation than invention; therefore, say the reasoners on this side of the question, had the human race been originally mutum et turpe pecus, they would have continued so to the end of time, unless they had been taught to speak by some superior intelligence. That the first men sprung from the earth like vegetables, no modern philosopher has ventured to assert; nor does there anywhere appear sufficient evidence that men were originally in the state of savages. The oldest book extant contains the only rational cosmogony known to the ancient nations; and that book represents the first human inhabitants of this earth, not only as reasoning and speaking animals, but also as in a state of high perfection and happiness, of which they were deprived for disobedience to their Creator. Moses, setting aside his claim to inspiration, deserves, from the consistency of his narrative, at least as much credit as Moschus, or Democritus, or Epicurus, derives from his prior antiquity. If antiquity could give this subject any weight, he would deserve more, as having lived nearer to the period of which they all write. But the question respecting the origin of language may be decided without resting on authority of any kind, merely by considering the nature of speech, and the mental and corporeal powers of man. Those who maintain it to be of human invention, suppose men at first to have been solitary animals, afterwards to have herded together, without government or subordination, then to have formed political societies, and by their own exertions to have advanced from the grossest ignorance to the refinements of science. But, say the reasoners, whose cause we are now pleading, this is a supposition contrary to all history and all experience. There is not upon record a single instance well authenticated of a people emerging by their own efforts from barbarism to civilization. There have indeed been many nations raised from the state of savages; but it is known that they were polished, not by their own repeated exertions, but by the influence of individuals or colonies from nations more enlightened than themselves. The original savages of Greece were tamed by the Pelasgi, a foreign tribe; and were afterwards further polished by Orpheus, Cercops, Cadmus, and others, who derived their knowledge from Egypt and the East. The ancient Romans, a ferocious and motley crew, received the blessings of law and religion from a succession of foreign kings; and the conquests of Rome, at a later period, contributed to civilize the rest of Europe. In America, the only two nations which, at the invasion of the Spaniards, could be said to have advanced a single step from barbarism, were indebted for their superiority over the other tribes, not to the gradual and unassisted progress of the human mind, but to the wise institutions of foreign legislators.

This is not the proper place for tracing the progress of man from the savage state to that of political society; but experience teaches us, that in every art it is much easier to improve than to invent? The human mind, when put into the proper track, is indeed capable of making great advances in arts and sciences; but, if any credit is due to the records of history, it has not, in a people sunk in ignorance and barbarity, sufficient vigour to discover that track, or to conceive a state different from the present. If the rudest inhabitants of America and Language, other countries have continued, as there is every reason to believe they have continued, for ages in the same unvaried state of barbarism; how is it imaginable that people so much ruder than the rest as to be ignorant of all language, should think of inventing an art so difficult as that of speech, or even to frame a conception of the thing. In building, fishing, hunting, navigating, and the like, they might imitate the instinctive arts of other animals, but there is no other animal that expresses its sensations and affections by arbitrary articulate sounds. It is said, that before language could be invented, mankind must have existed for ages in large political societies, and have carried on in concert some common work; and if inarticulate cries, and the natural visible signs of the passions and affections, were modes of communication sufficiently accurate to keep a large society together for ages, and to direct its members in the execution of some common work, what could be their inducement to the invention of an art so useful and difficult as that of language?

Let us however suppose, say the advocates for the cause which we are now supporting, that different nations of savages set about inventing an art of communicating their thoughts, which experience had taught them was not absolutely necessary; how came they all, without exception, to think of the one art of articulating the voice for this purpose? Inarticulate cries, out of which language is fabricated, have indeed an instinctive connection with our passions and affections; but there are gestures and expressions of countenance with which our passions and affections are in the same manner connected. If the natural cries of passion could be so modified and enlarged as to be capable of communicating to the hearer every idea in the mind of the speaker, it is certain that the natural gestures could be so modified as to answer the very same purpose; and it is strange that, among the several nations who invented languages, not one should have stumbled upon fabricating visible signs of their ideas, but that all should have agreed to denote them by articulated sounds.

Every nation whose language is narrow and rude, supplies its defects by a violent gesticulation; and therefore, as much less genius is exerted in the improvement of any art than was requisite for its first invention, it is natural to suppose that, had men been left to devise for themselves a method of communicating their thoughts, they would not have attempted any other than that by which they now improve the language transmitted by their fathers. It is vain to urge that articulate sounds are fitted for the purpose of communicating thought than visible gesticulation; for, though this may be true, it is a truth which could readily occur to savages, who had never experienced the fitness of either; and if, to counterbalance the superior fitness of articulation, its extreme difficulty be taken into view, it must appear little less than miraculous that every savage tribe should think of it rather than the easier method of artificial gesticulation. Savages, it is well known, are remarkable for their indolence, and for always preferring ease to utility; but their modes of life give such facility to their bodies, that they could with very little trouble bend their limbs and members into any positions agreed upon as the signs of ideas. This is so far from being the case with respect to the organs of articulation, that it is with extreme difficulty, if at all, that a man advanced in life can be taught to articulate any sound which he has not been accustomed to hear. No foreigner who comes to England after the age of thirty ever pronounces the language tolerably well; an Englishman of that age can hardly be taught to utter the guttural sound which a Scotman gives to the Greek χ, or even the French sound of the vowel ο; and of the solitary savages who have been caught in different forests, we know not that there has been one who, after the age of manhood, learned to articulate any language so as to make himself readily understood. The present age has indeed furnished many instances of deaf persons being taught to speak intelligibly by skilful masters moulding the organs of the mouth into the positions proper for articulating the voice; but who was to perform this task amongst the inventors of language, when all mankind were equally ignorant of the means by which articulation is effected? In a word, daily experience informs us, that men who have not learned to articulate in their childhood, never afterwards acquire the faculty of speech but by such helps as savages cannot obtain; and therefore, if speech was invented at all, it must have been either by children who were incapable of invention, or by men who were incapable of speech. A thousand, nay, a million, of children could not think of inventing a language. While the organs are pliable, there is not understanding enough to frame the conception of a language; and by the time that there is understanding, the organs are become too stiff for the task, and therefore, say the advocates for the divine origin of language, reason as well as history intimates, that mankind in all ages must have been speaking animals—the young having constantly acquired this art by imitating those who were older; and we may warrantably conclude, that our first parents received it by immediate inspiration.

To this account of the origin of language an objection readily offers itself. If the first language was communicated by inspiration, it must have been perfect, and held in reverence by those who spoke it, in other words, by all mankind. But a great variety of languages have prevailed in the world; and some of these which remain are known to be very imperfect, whilst there is reason to believe that many others are lost. If different languages were originally invented by different nations, all this would naturally follow from the mixture of these nations; but what could induce men possessed of one perfect language of divine original, to forsake it for barbarous jargons of their own invention, and in every respect inferior to that with which their forefathers or themselves had been inspired?

In answer to this objection, it is said, that nothing was given by inspiration but the faculty of speech and the elements of language; for, when once men had language, it is easy to conceive how they might have modified it by their natural powers, as thousands can improve what they could not invent. The first language, if given by inspiration, must, in its principles, have had all the perfection of which language is susceptible; but from the nature of things it could not possibly be very copious. The words of language are either proper names or the signs of ideas and relations; but it cannot be supposed that the All-wise Instructor would load the memories of men with words to denote things then unknown, or with the signs of ideas which they had not then acquired. It was sufficient that a foundation was laid of such a nature as would support the largest superstructure which they might ever after have occasion to raise upon it, and that they were taught the method of building by composition and derivation. This would long preserve the language radically the same, though it could not prevent the introduction of different dialects in the different countries over which men spread themselves. In whatever region we suppose the human race to have been originally placed, the increase of their numbers would, in process of time, either disperse them into different nations, or extend the one nation to a vast distance on all sides from what we may call the seat of government. In either case they would everywhere meet with new objects, which would occasion the invention of new names; and as the difference of climate and other natural causes would compel those who removed eastward or northward to adopt modes of life in many respects different from the modes of those who travelled towards the west or the south, a vast number of words would in one country be fabricated to denote complex conceptions, which must necessarily be unintelligible to the body of the people inhabiting countries where those conceptions had never been formed. Thus would various dialects be unavoidably introduced into the original language, even whilst all mankind remained in one society and under one government. But after separate and independent societies were formed, these variations would become more numerous, and the several dialects would deviate farther and farther from each other, as well as from the idiom and genius of the parent tongue, in proportion to the distance of the tribes by whom they were spoken.

This is not sufficient. Languages ought, by hypothesis, to graduate into each other more than they do. In order to account for the existing lines of demarcation, which are broad and definite, we must bear in mind a fresh phenomenon, viz., the spread of one dialect at the expense of others, a fact which obliterates intermediate forms, and brings extreme ones into geographical juxtaposition.

And here it may be noticed, that the distribution of languages over the earth's surface is irregular. In all the quarters of the world we find the two kinds of areas contrasted with one another. In the one, a single language, with a Language. minimum amount of dialectical difference, is spread over a vast extent of country; in the other, a multiplicity of mutually unintelligible forms of speech is found within a small compass.

In Europe, the Slavonic tongues have a large, the Biscayan a small area.

In Asia, the Turk area is large, that of the languages of Caucasus, Nepali, &c., small.

In Africa, the Berber is large, and certain Abyssinian and other areas small.

In Polynesia, the Malay and Polynesian are large, the Papuan area small.

In America there are numberless small areas; but the Algonkin, Guarani, and a few others, are large.—[x. o. t.]

If we suppose a few people either to have been banished together from the society of their brethren, or to have wandered of their own accord to a distance, from which, through trackless forests, they could not return (and such migrations have often taken place), it is easy to see how the most copious language must in their mouths have soon become narrow, and how the offspring of inspiration must have in time become so deformed as hardly to retain a feature of the ancestor whence it originally sprung. Men do not lose entirely a practical skill in those arts which they never exercise; and there are abundance of tasks to prove, that a single man cast upon a desert island, and having to provide the necessaries of life by his own ingenuity, would soon lose the art of speaking with fluency his mother tongue. A small number of men cast away together, would indeed retain that art somewhat longer; but in a space of time not very long, it would in a great measure be lost by them or their posterity. In this state of banishment, as their time would be almost wholly occupied in hunting, fishing, and other means within their reach, to support a wretched existence, they would have very little leisure, and perhaps less desire, to preserve by conversation the remembrance of that ease and those comforts of which they now found themselves for ever deprived; and they would of course soon forget all the words which in their native language had been used to denote the accommodations and elegancies of polished life. This at least seems to be certain, that they would not attempt to teach their children a part of language which in their circumstances could be of no use to them, and of which it would be impossible to make them comprehend the meaning; for where there are no ideas, the signs of ideas cannot be made intelligible. From such colonies as this, dispersed over the earth, it is probable that all those nations of savages have arisen, which have induced so many philosophers to imagine that the state of the savage was the original state of man; and if so, we see that from the language of inspiration must have unavoidably sprung a number of different dialects, all extremely rude and narrow, and retaining nothing of the parent tongue, except perhaps the names of the most conspicuous objects of nature, and of those wants and enjoyments which are inseparable from humanity. The savage state has no artificial wants, and furnishes few ideas that require terms to express them. The Language-habits of solitude and silence incline a savage rarely to speak; and when he speaks, he uses the same terms to denote different ideas. Speech, therefore, in this rare condition of man, must be extremely narrow, and extremely various. Every new region and every new climate, suggests different ideas, and creates different wants, which must be expressed either by terms entirely new, or by old terms used with a new signification. Hence must originate great diversity, even in the first elements of speech, among all savage nations, the words retained of the original language being used in various senses, and pronounced, as we may believe, with various accents.

When any of those savage tribes emerged from their barbarism, whether by their own efforts or by the aid of people more enlightened than themselves, it is obvious that the improvement and copiousness of their language would keep pace with their own progress in knowledge, and in the arts of civil life; but, in the infinite multitude of words which civilization and refinement add to language, it would be little less than miraculous were any two nations to agree upon the same sounds to represent the same ideas. Superior refinement, indeed, may induce imitation, conquests may impose a language, and extension of empires may melt down different nations and different dialects into one mass; but independent tribes naturally give rise to diversity of tongues; nor does it seem possible that they should retain more of the original language than the words expressive of those objects with which all men are at all times equally concerned. The variety of tongues, therefore, the copiousness of some, and the narrowness of others, furnish no good objection to the divine origin of language in general; for, whether language was at first revealed from heaven, or in a course of ages invented by men, a multitude of dialects would inevitably arise as soon as the human race had separated into a number of distinct and independent nations.

As the knowledge of languages constitutes a great part of Idiom and erudition, as their beauties and deformities furnish employ, genius of mind to taste, and as these depend much upon the idioms of language, the different tongues, we shall proceed to make a few remarks upon the advantages and defects of some of those idioms of language with which we are best acquainted. As the words idiom and genius of a language are often confounded, it will be necessary to inform the reader, that by idioms we would here be understood to mean that general mode of arranging words into sentences which is peculiar in any particular language; and by the genius of a language, we mean to express the particular set of ideas which the words of any language, either from their formation or multiplicity, are most naturally apt to excite in the mind of any one who hears it properly uttered. Thus, although the English, French, Italian, and Spanish languages nearly agree in the same general idioms, yet the particular genius of each is remarkably different. The English is naturally bold, nervous, and strongly articulated; the French is weaker, and more flowing; the Italian

---

1 In the foregoing view of this subject, the argument for the supernatural origin of language is evidently that which the author wishes to favour. But though we have, with some slight alterations, reprinted this part of the article, which was written for the earlier editions of this work, we are far from adopting the very unphilosophical opinion which it supports. We do not assert, because we could have nothing like evidence for such an assertion, that the Deity did not originally bestow on man the gift of speech; but we think, with Lord Monboddo, and many others, that if such a book ever was written, it must be in the revolutions and calamities that have befallen the human race, inevitably have been lost; and, therefore, that as multitudes of languages exist and have ever existed, the art of speech is one which man is capable of attaining to, independently of any supernatural aid. This is the view taken by our best philosophers. Thus, Mr Stewart, notwithstanding his usual caution in forming conclusions on momentous subjects, unhesitatingly avows his conviction, that "the human faculties are competent to the formation of language." (Philosophy of the Mind, vol. III., chap. I.) In another place, the same admirable writer makes the following philosophical observations on this subject:—“The steps in the formation of language must probably be determined with certainty; yet if we can show, from the known principles of human nature, how the various parts may naturally have arisen, the mind is not only to a certain degree satisfied, but a check is given to that species of philosophy which refers to a miracle whatever appearances, either in the natural or moral world, it is unable to explain.” One of the most ingenious attempts that ever has been made to show by what steps the human mind could naturally proceed in the acquisition of speech, is that contained in Dr Adam Smith’s “Considerations on the Formation of Languages.” His theory may perhaps be liable to objection in some of its details; but it is developed with singular elegance and simplicity; and it must, in its general scope and design, be allowed by all candid and competent judges to be distinctly philosophical. This essay was annexed by its author to the first edition of his Theory of Moral Sentiments; and we are told by Mr Stewart, that he always regarded it with great partiality, and that he reprinted it, without a single alteration, in the last edition of that work which he himself revised. Mr Stewart’s commentary upon it (Philosophy of the Mind, vol. sup.) ought to be perused along with it. [The phenomena of languages are those of Growth and Development, rather than those of an organic whole invented at once. The process by which these phenomena are studied is all that can here be noticed. We begin with language as it exists at present, and eliminate whatever we can trace to a recent origin, e.g., abstract terms, conjunctions, inflections, &c. Doing this, though we never come to the original nucleus of language, we soon become satisfied that that was exceedingly small.—x. o. t.] Language, more soothing and harmonious; and the Spanish more grave, somberous, and stately. Now, when we examine the several languages which have been most esteemed in Europe, we find that there are only two idioms among them which are essentially distinguished from one another; and all those languages are divided between these two idioms, following sometimes the one and sometimes the other, either wholly or in part. The languages which may be said to adhere to the first idiom, are those which, in their construction, follow the order of nature; that is, express their ideas in the natural order in which they occur to the mind—the subject which occasions the action appearing first; then the action accompanied with its several modifications; and, last of all, the object to which it has reference. These may properly be called analogous languages; and of this kind are the English, French, and most of the modern languages in Europe. The languages which may be referred to the other idiom, are those which follow no other order in their construction than what the taste or fancy of the composer may suggest; sometimes making the object, sometimes the action, and sometimes the modification of the action, to precede or follow the other parts. The confusion which this might occasion is avoided by the particular manner of inflecting their words, by which they are made to refer to the others with which they ought to be connected, in whatever part of the sentence they occur, the mind being left at liberty to connect the several parts with one another after the whole sentence is concluded; and as the words may be here transposed at pleasure, those languages may be called transpositive languages. To this class we must, in an especial manner, refer the Latin and Greek languages. As each of these idioms has several advantages and defects peculiar to itself, we shall endeavour to point out the most considerable of them, in order to ascertain with greater precision the particular character and excellence of each of those languages now principally spoken or studied in Europe.

The partiality which our forefathers, at the revival of letters in Europe, naturally entertained for the Greek and Roman languages, made them look upon every distinguishing peculiarity belonging to them as one of the many causes of the amazing superiority which those languages evidently enjoyed above every other at that time spoken in Europe. This blind deference still continues to be paid to them, as our minds are early prepossessed with these ideas, and as we are taught in our earliest infancy to believe, that to entertain the least idea of our own language being equal to the Greek or Latin in any particular whatever, would be a certain mark of ignorance or want of taste. Their rights, therefore, like those of the church in former ages, remain still to be examined; and we, without exerting our reason to discover truth from falsehood, tamely sit down satisfied with the idea of their undoubted pre-eminence in every respect. But if we look around us for a moment, and observe the many excellent productions which are to be met with in almost every language of Europe, we must be satisfied that even those are now possessed of some powers which might afford at least a presumption, that if they were cultivated with a proper degree of attention, they might, in some respects, be made to rival, if not to excel, those beautiful and justly admired remains of antiquity. Without endeavouring to derogate from their merit, let us, with the cool eye of philosophic reasoning, endeavour to bring before the sacred tribunal of truth some of those opinions which have been most generally received upon this subject, and rest the determination of the cause on her impartial decision.

The learned reader well knows, that the several changes which take place in the arrangement of the words in every transpositive language, could not be admitted without occasioning great confusion, unless certain classes of words were endowed with particular variations, by means of which they might be able to refer to the other words with which they ought naturally to be connected. From this cause proceeds the necessity of several relations of verbs, nouns, and adjectives; which are not in the least essentially necessary in the analogous languages, as we have pretty fully explained under the article Grammar, to which we refer for illustration on this head. We shall, in this place, consider whether these variations are an advantage or a disadvantage to language.

As it is generally supposed that every language, the verbs of which admit of inflection, is on that account much more perfect than one where they are varied by auxiliaries, we shall, in the first place, examine this with some degree of attention; and, that what is said on this head may be the more intelligible, we shall give examples from the Latin and English languages. We make choice of these languages because the Latin is more purely transpositive than the Greek, and the English admits of less inflection than any other language that we are acquainted with.

If any preference be due to a language from the one or the other method of conjugating verbs, it must in a great measure be owing to one or other of these three causes: Either it must admit of a greater variety of sounds, and consequently afford more scope for harmonious diversity of tones in the language; or a greater freedom of expression is allowed in uttering any simple idea, by the easy admitting of a greater variety in the arrangement of the words which are necessary to express that idea than the other does; or, lastly, a greater precision and accuracy in fixing the meaning of the person who uses the language, arise from the use of one of these forms rather than from the use of the other. For, as all the other circumstances which may serve to give a diversity to language, such as the general and most prevalent sounds, the frequent repetition of any one particular letter, and a variety of other circumstances of that nature, which may serve to diversify a particular language, are not influenced in the least by the different methods of varying the verbs, they cannot be here considered. We shall therefore proceed to make a comparison of the advantages or disadvantages which may accrue to a language by inflecting its verbs, with regard to the variety of sound, variety of arrangement, and accuracy of meaning.

The first particular that we have to examine is, whether the one method of expressing the variations of a verb admits of a greater variety of sounds. In this respect the Latin seems, at first view, to have a great advantage over the English; since the words, amo, amabam, amaveram, amavero, amas, &c., seem to be more different from one another than the English translations of these, I love, I was loving, I had loved, I shall have loved, I may love, &c.; for although the syllable am is repeated in every one of the first, yet, as the last syllable usually strikes the ear with greater force, and leaves a greater impression than the first, it is very probable that many will think the frequent repetition of the word love, in the last instance, more striking to the ear than the repetition of am in the former. We will therefore allow this its full weight, and grant that there is as great, or even a greater, difference between the sounds of the different tenses of a Latin verb, than there is between the words that are equivalent to them in English. But as we here consider the variety of sounds of the language in general, before any just conclusion can be drawn, we must not only compare the different parts of the same verb, but also compare the different verbs with one another in each of these languages. And here, at first view, we perceive a most striking distinction in favour of the analogous language over the inflected: for, as it would be impossible to form a particular set of inflections, different from one another, for each particular verb, all those languages which have adopted this method have been obliged to reduce their verbs into a small number of classes, all the words of each of which classes, commonly called conjugations, have the several variations of the moods, tenses, and persons, expressed exactly in the same manner, which must of necessity introduce a similarity of sounds into the language in general, much greater than where every particular verb always retains its own distinguishing sound. To be convinced of this, we need only repeat any number of verbs in Latin and English, and observe on which side the preference with respect to variety of sounds must fall.

| Latin | English | |-------|---------| | Pono, | I put. | | Dono, | I give. | | Cano, | I sing. | | Sono, | I sound. | | Orno, | I adorn. | | Pugno, | I fight. | | Lego, | I read. | | Scribo, | I write. | | Puto, | I think. | | Vivo, | I love. | | Ambulo, | I walk. |

| Latin | English | |-------|---------| | Movero, | I move. | | Doceo, | I tell. | | Luggeo, | I mourn. | | Obeo, | I love. | | Gaudeo, | I rejoice. | | Incipio, | I begin. | | Facio, | I make. | | Fedio, | I dig. | | Ridéo, | I laugh. | | Impléo, | I fill. | | Abstineo, | I forbear. | The similarity of sounds is here so obvious in the Latin, as to be perceived at the first glance; nor can we be surprised to find it when we consider that all their regular verbs, amounting to near a thousand or upwards, must be reduced to four conjugations, and even these differing but little from one another, which must of necessity produce the sameness of sounds which we here perceive; whereas, every language that follows the natural order, like the English, instead of this small number of uniform terminations, has almost as many distinct sounds as original verbs.

But if, instead of the present of the indicative mood, we should take almost any other tense of the Latin verb, the similarity of sounds would still be more perceptible, as many of these tenses have the same termination in all the four conjugations, particularly in the imperfect of the indicative, as below:

| Latin Verb | English Translation | |------------|---------------------| | Pono-bam | I did put, | | Donna-bam | I did give, | | Cano-bam | I did sing, | | Sono-bam | I did sound, | | Orna-bam | I did adorn, | | Pugna-bam | I did fight, | | Lege-bam | I did read, | | Scribe-bam | I did write, | | Puta-bam | I did think, | | Vive-bam | I did live, | | Ambula-bam | I did walk, | | Move-bam | I did move, | | Delibe-bam | I did die, | | Luge-bam | I did mourn, | | Obi-bam | I did die, | | Gaudie-bam | I did rejoice, | | Incipie-bam| I did begin, | | Facie-bam | I did make, | | Fodle-bam | I did dig, | | Ride-bam | I did laugh, | | Impie-bam | I did fill, | | Absine-bam | I did forbear, |

It is unnecessary to make any remarks on the Latin words in this example. But in the English translation we have carefully marked in the first column the words without any inflection; and in the second, have put down the same meaning by an inflection of our verb; which we have been enabled to do, from a peculiar excellency in our own language unknown to any other, either ancient or modern. Were it necessary to pursue this subject farther, we might observe, that the perfect tense in all the conjugations ends universally in _i_, the pluperfect in _eaxm_, and the future in _am_ or _no_; in the subjunctive mood the imperfect universally in _em_, the perfect in _eaxm_, the pluperfect in _eaxm_, and the future in _eax_; and as a still greater sameness is observable in the different variations for the persons in these tenses, seeing the first person plural in all tenses ends in _mus_, and the second person in _vis_, with little variation in the other persons, it is evident that, in respect to diversity of sounds, this method of conjugating verbs by inflection is greatly inferior to the more natural method of expressing the various connections and relations of the verbal attribute by different words, usually called _analogies_.

The second particular by which the different methods of marking the relation of the verbal attribute can affect language, arises from the variety of expressions which either of these may admit of in uttering the same sentiment. In this respect likewise, the method of conjugation by inflection seems to be deficient. Thus the present of the indicative mood in Latin can at most be expressed only in two ways, viz., _senmo_ and _noo senmo_, which ought, perhaps, in strictness to be admitted only as one; whereas, in English, we can vary it in four different ways, viz.—_let, I write_; _2dly_, _I do write_; _3dly_, _Write I do_; _4thly_, _Write do I_: And if we consider Language, the further variation which these receive in power as well as in sound, by having the emphasis placed on the different words, instead of four we will find eleven different variations. Thus, _I write_, with the emphasis upon the _i_; _write_, with the emphasis upon the word _write_. Let any one pronounce these with the different emphasis necessary, and he will be immediately satisfied that they are not only distinct from each other with respect to meaning, but also with regard to sound. The same must be understood of all the other parts of this example:

| Latin Verb | English Translation | |------------|---------------------| | I do write.| Write I do. | | I do write.| Write do I. | | I do write.| Write I do. | | Write I do.| Write do I. | | Write I do.| Write I do. |

None of the Latin tenses admit of more variations than the two above mentioned: nor do almost any of the English admit of fewer than in the above example; and several of these phrases, which must be considered as exact translations of some of the tenses of the Latin verb, admit of many more. Thus the imperfect of the subjunctive mood, which in Latin admits of the above two variations, admits in English of the following:

| Latin Verb | English Translation | |------------|---------------------| | I might have written.| Written might have I. | | Written I might have.| I written might have. | | Have written I might.| Have written might I. |

And if we likewise consider the variations which may be produced by a variation of the emphasis, they will be as under:

| Latin Verb | English Translation | |------------|---------------------| | I might have written.| Written might have I. | | Written I might have.| I written might have. | | Have written I might.| Have written might I. |

In all, twenty-four variations, instead of two. If we likewise consider that the Latins were obliged to employ the same word, not only to express "I might have written," but also, "I could, I would, or I should have written," each of which would admit of the same variations as the word _might_, we have in all ninety-six different expressions in English for the same phrase, which in Latin admits only of two, unless they have recourse to other forced terms of expression, which the defects of their verbs in this particular has compelled them to invent.

But if it should be objected that the last circumstance we have taken notice of as a defect can only be considered as a defect of the Latin language, and is not to be attributed to the inflection of their verbs, seeing they might have had a particular tense for each of these different words _might_, _could_, _would_, and _should_; we answer, that, even admitting this excuse to be valid, the superiority of the analogous language as such still remains in this respect as twelve to one. Yet even this concession is greater than ought to have been made. For as the difficulty of forming a sufficient variety of words for all the different modifications which a verb may be made to undergo is too great for any rude people to overcome, we find, that every nation which has adopted this mode of inflection, not excepting the Greeks themselves, has been obliged to remain satisfied with fewer words than would have been necessary even to effect this purpose, and make the same word serve a double, triple,

---

1 We are sufficiently aware that the last variation cannot in strictness be considered as good language, although many examples of this manner of using it in serious composition, both in poetry and prose, might easily be produced from the best authors in the English language. But, however unjustifiable it may be to use it in serious composition, yet, when judiciously employed in works of humour, this and other forced expressions of the like nature produce a fine effect, by giving a burlesque air to the language, and beautifully contrasting it with the purer diction of solid reasoning. Shakespeare has on many occasions showed how successfully these may be employed in composition, particularly in drawing the character of Ancient Pistol in Henry V. Without this liberty, Butler would have found greater difficulty in drawing the inimitable character of Hudibras. Let this apology suffice for having inserted this and other variations of the same kind, which, although they may be often improper for serious composition, have still their use in language. Language, or even quadruple office, as in the Latin tense which gave rise to these observations. So that, however, in physical necessity this may not be chargeable upon the particular mode of construction, yet in moral certainty it must always be the case; and, therefore, we may safely conclude, that the mode of varying verbs by inflection affords less variety in the arrangement of the words of the particular phrases, than the method of varying them by the help of auxiliaries.

But if there should still remain any shadow of doubt in the mind of the reader, whether the method of varying the verbs by inflection is inferior to that by auxiliaries, with regard to diversity of sounds or variety of expression, there cannot be the least doubt that, with respect to precision, distinctness, and accuracy in expressing any idea, the latter enjoys a superiority beyond all comparison. Thus, the Latin verb amo may be Englished either by the words I love or I do love, and the emphasis placed upon any of the words that the circumstances may require; by means of which the meaning is pointed out with a force and energy which it is altogether impossible to produce by the use of any single word. The following line from Shakespeare's Othello may serve as an example:

"Excellent wretch! Perdition catch my soul, but I love thee."

Here the strong emphasis on the word no gives it a force and energy which conveys, in an irresistible manner, a most perfect knowledge of the situation of the mind of the speaker at the time. That the whole energy of the expression depends upon this seemingly insignificant word, we may be at once satisfied of by keeping it away, in this manner:

"Excellent wretch! Perdition catch my soul, but I love thee."

How poor, how tame, how insignificant is this, when compared with the other. Here nothing remains but a tame assertion, ushered in with a pompous exclamation, which could not here be introduced with any degree of propriety. Whereas, in the way that Shakespeare has left it to us, it has an energy which nothing can surpass; for, overpowered with the irresistible force of Desdemona's charms, this strong exclamation is extorted from the soul of Othello in spite of himself. Surprised at this tender emotion, which brings to his mind all those amiable qualities for which he had so much esteemed her, and at the same time fully impressed with the firm persuasion of her guilt, he bursts out into that seemingly inconsistent exclamation,—"Excellent wretch!" and then he adds in the warmth of his surprise, thinking it a thing most astonishing that any warmth of affection should still remain in his breast,—he even confirms it with an oath.—"Perdition catch my soul, but I love thee." "In spite of all the falsehoods with which I know thou hast deceived me, in spite of all the crimes of which I know thee guilty, in spite of all those reasons for which I ought to hate thee, in spite of myself, still I find that I love thee, yes, I love thee." We look upon it as a thing altogether impossible to transmute the energy of this expression into any language whose verbs are regularly inflected.

In the same manner we might go through all the other tenses, and show that the same superiority is to be found in each. Thus, in the perfect tense of the Latins, instead of the simple AMAVI, we say I HAVE LOVED; and by the liberty we have of putting the emphasis upon any of the words which compose this phrase, we can in the most adequate manner fix the precise idea which we mean to excite; for, if we say, I have loved, with the emphasis upon the word I, it at once points out the person as the principal object in that phrase, and makes us naturally look for a contrast in some other person, and the other parts of the phrase become subordinate to it: "He has loved thee much, but I have loved thee infinitely more." The Latins, too, as they were not prohibited from joining the pronoun with their verb, were also acquainted with this excellence, which Virgil has beautifully used in this verse:

"Nos patriam fugimus; Tu, Titre, lentus in umbra," &c.

We are not only enabled thus to distinguish the person in as powerful a manner as the Latins, but can also with the same facility point out any of the other circumstances as principals; for if we say, with the emphasis upon the word have, "I have loved," it as naturally points out the time as the principal object, and makes us to look for a contrast in that peculiarity, I have: "I have loved indeed; my imagination has been led astray, my reason has been perplexed; but, since that time has opened my eyes, I can smile at those imaginary distresses which once perplexed me." In the same manner we can put the emphasis upon the other word of the phrase, "I have been loved." Here the passion is exhibited as the principal circumstance; and as this can never be excited without some object, we naturally wish to know the object of that passion. "Who! what have you loved?" are the natural questions we would put in this case. "I have loved Eliza." In this manner we are, on all occasions, enabled to express, with the utmost precision, that particular idea which we would wish to excite, so as to give an energy and perpetuity to the language, which can never be attained by those languages whose verbs are conjugated by inflection; and if to this we add the inconvenience which all inflected languages are subject to, by having too small a number of tenses, so as to be compelled to make one word on many occasions supply the place of two, three, or even four, the balance is turned still more in our favour. Thus, in Latin, the same word, AMABO, stands for shall or will love, so that the reader is left to guess from the context which of the two meanings it was most likely the writer had in view. In the same manner may or can love are expressed by the same word, AMAR; as are also might, could, would, or should love, by the single word AMARE, as we have already observed, so that the reader is left to guess which of these four meanings the writer intended to express; an ambiguity which occasions a perplexity very different from that clear precision which our language allows of, by not only pointing out the different words, but also by allowing us to put the emphasis upon any of them we please, which superadds energy and force to the proposition we would have had without that assistance.

Upon the whole, therefore, after the most candid examination, we must conclude, that the method of conjugating verbs by inflection is inferior to that which is performed by the help of auxiliaries; because it does not afford such a diversity of sounds, nor allow such variety in the arrangement of expression for the same thought, nor give such great distinction and precision in the meaning. It is, however, attended with one considerable advantage above the other method; for, as the words of which it is formed are necessarily of great length, and more sonorous than in the analogous languages, it admits of a more flowing harmony of expression; for the number of monosyllables in this last greatly checks that pompous dignity which naturally results from longer words. Whether this single advantage is sufficient to counterbalance all the other defects with which it is attended, is left to the judgment of the reader to determine. But we may remark before we quit the subject, that even this excellence is attended with some peculiar inconveniences, which shall be more particularly pointed out in the sequel.

But perhaps it might still be objected, that although the comparisons we have made above may be fair, and the conclusion just, with regard to the Latin and English languages, yet it does not appear clear, that on that account the method of conjugating verbs by inflection is inferior to that by auxiliaries; for, although it be allowed that the Latin language is defective in point of tenses, yet, if a language were formed which had a sufficient number of inflected tenses to answer every purpose; if it had, for instance, a word properly formed for every variation of each tense; one for I love, another for I do love; one for I shall, another for I will love; one for I might, another for I could, and would, and should love; and so on through all the other tenses; that this language would not be liable to the objections we have brought against the inflections of verbs; and that of course the objections we have brought would be invalid against those languages which have followed that mode, and adopted it imperfectly. We answer, that although this would in some measure remedy the evil, yet it would not remove it entirely. For, in the first place, unless every verb, or every small number of verbs, were conjugated in one way, having the sound of the word in each tense, and division of tenses, as we may say, different from all the other conjugations, it would always occasion a sameness of sound, which would in some measure prevent that variety of sounds Language so proper for a language. And even if this could be effected, it would not give such a latitude to the expression as auxiliaries allow; for although there should be two words, one for I might, and another for I could love, yet as these are single words, they cannot be varied; whereas, by auxiliaries, either of these can be varied twenty-four different ways, as has been shown above. In the last place, no single word can ever express all that variety of meaning which we can do by the help of our auxiliaries and the emphasis. I have loved, if expressed by any one word, could only denote at all times one distinct meaning; so that to give it the power of ours, three distinct words at least would be necessary. However, if all this were done; that is, if there were a distinct conjugation formed for every forty or fifty verbs; if each of the tenses had proper forms, and all of them different from every other tense, as well as every other verb; and these all carried through each of the different persons, so as to be all different from one another; and if likewise there were a distinct word to mark each of the separate meanings, which the same tense could be made to assume by means of the emphasis; and if all this infinite variety of words could be formed in a distinct manner, different from each other, and harmonious; this language would have powers greater than any that could be formed by auxiliaries, if it were possible for the human powers to acquire such a degree of knowledge as to be able to employ it with facility. But how could this be attained, since upwards of ten thousand words would be necessary to form the variations of any one verb, and a hundred times that number would not include the knowledge of the verbs alone of such a language? How much, therefore, ought we to admire the simple perspicuity of our language, which enables us, by the proper application of ten or twelve seemingly trifling words, the meaning and use of which can be attained with the utmost ease, to express all that could be expressed by this unwieldy apparatus? What can equal the simplicity or the power of the one method, but the well-known powers of the twenty-four letters, the knowledge of which can be obtained with so much ease, whilst their powers know no limits? or, what can be compared to the fancied perfection of the other, but the transcript of it which the Chinese seem to have formed in their unintelligible language?

Having thus considered pretty fully the advantages and defects of each of these two methods of varying verbs, we cannot help feeling a secret wish arise in our mind, that there had been a people sagacious enough to have united the powers of the one method with those of the other; nor can we help being surprised, that among the changes which took place in the several languages of Europe after the downfall of the Roman monarchy, some of them did not accidentally stumble on the method of doing it. From many concerning circumstances, it seems probable that the greatest part, if not all, of the Gothic nations that overrun Italy at that time, had their verbs varied by the help of auxiliaries; and many of the modern European languages which have sprung from them, have so far borrowed from the Latin, as to have some of the tenses of their verbs inflected. Yet the English alone have in any instance combined the joint powers of the two, which could only be done by forming inflections for the different tenses in the same manner as the Latins, and at the same time retaining the original method of varying them by auxiliaries; by which means either the one or the other method could have been employed to occasion reduction. We have luckily two tenses formed in that way, the present of the indicative, and the aorist of the past. In almost all our verbs these can be declined either with or without auxiliaries. Thus the present, without an auxiliary, is, I love, I write, I speak; with an auxiliary, I do write, I do love, I do speak. In the same manner, the past tense, by inflection, is, I loved, I wrote, I spoke; by auxiliaries, I did love, I did speak, I did write.

Every author who knows anything of the power of the English language, knows the use which may be made of this distinction. What a pity it is that we should have stooped short so soon. How blind was it in many other nations to imitate the defects without making a proper use of that beautiful language Language, which is now numbered among the dead.

After the verbs, the next most considerable variation we find between the analogous and transpositive languages is in the nouns; the latter varying the different cases of these by inflection; whereas the former express all the different variations of them by the help of other words prefixed, called prepositions. Now, if we consider the advantages or disadvantages of each of these methods under the same circumstances, we have done the verbs, we shall find, that with regard to the first particular, viz., variety of sounds, since the same remarks may be made upon the verbs, for if we compare any particular noun by itself, the variety of sound appears much greater between the different cases in the transpositive, than between the translation of these in the analogous language. Thus rex, regis, regni, regem, &c., are more distinct from one another in point of sound, than the translation of these, a king, of a king, to a king, a king, &c. But if we proceed one step further, and consider the variety which is produced in the language in general by the one or the other of these methods, the case is entirely reversed. For, as it would have been impossible to form distinct variations, different from one another, for each case of every noun, they have been obliged to reduce all their nouns into a few general classes, called declensions, and to give to all those included under each class the same termination in every case, which produces a like similarity of sound with what we already observed was occasioned to the verbs from the same cause; whereas, in the analogous languages, as there is no necessity for any constraint, there is almost as great a variety of sounds as there are of nouns. The Latins have only five different declensions; so that all the great number of words of this general order must be reduced to the very small diversity of sounds which these few classes admit of; and even the sounds of these few classes are not so much diversified as they might have been, as many of the different cases in the different declensions have exactly the same sounds, as we shall have occasion to remark more fully hereafter. We might here produce examples to show the great similarity of sounds between different nouns in the Latin language, and variety in the English, in the same way as we did of the verbs; but as every reader in the least acquainted with these two languages, can satisfy himself in this particular, without any further trouble than by marking down any number of Latin nouns, with their translations into English, we think it unnecessary to dwell longer on this particular.

But if the inflection of nouns is a disadvantage to a language in point of diversity of sounds, it is very much the reverse with regard to the variety it allows in the arranging the words of the phrase. Here, indeed, the transpositive language shines most in all its glory; and the analogous must yield the palm without occasioning the smallest confusion; whereas, in the analogous languages, as these two different states of the noun are expressed by the same word, they cannot be distinguished but by their position alone; so that the noun, which is the essential cause, must always precede the verb, and that which is the passive subject must follow, which greatly cramps the harmonious flow of composition. Thus the Latins, without the smallest perplexity in the meaning, could say either Brutum amavit Cassius, or Cassius amavit Brutum, or Brutus Cassius amavit, or Cassius Brutum amavit. As the termination of the word Cassius always points out that it is in the nominative case, and therefore that he is the person from whom the energy proceeds; and in the same manner, as the termination of the word Brutum points out that it is in the accusative case, and consequently that he is the object upon whom the energy is exerted, the meaning continues still distinct and clear, notwithstanding all these several variations; whereas, in the Eng-

---

1 This assertion may perhaps appear to many very much exaggerated; but if any one should think so, we only beg the favour that he will set himself to mark all the variations of tenses, mode, person, and number, which an English verb can be made to assume, varying each of these in every way that it will admit, both as to the diversity of expression and the emphasis, and he will soon be convinced that we have here said nothing more than enough. Language. In language, we could only say, Cassius loved Brutus, or, by a more forced phraseology, Cassius Brutus loved. Were we to reverse the case, as in the Latin, the meaning also would be reversed; for if we say Brutus loved Cassius, it is evident, that instead of being the person beloved, as before, Brutus now becomes the person from whom the energy proceeds, and Cassius becomes the object beloved. In this respect, therefore, the analogous languages are greatly inferior to the transpositive; and, indeed, it is from this single circumstance alone that they derive their chief excellence.

But although it thus appears evident that any language which has a particular variation of its nouns to distinguish the accusative from the nominative case, has an advantage over those languages which have none; yet it does not appear that any other of their cases adds to this variety, but rather the reverse; for in Latin we can only say Amor Dei; in English the same phrase may be rendered either the love of God, of God the love, or, by a more forced arrangement, God the love of. And these oblique cases, as the Latins called them, except the accusative, are clearly distinguished from one another and from the nominative, by the preposition which accompanies them, we are not confined to any particular arrangement with regard to these, as with the accusative, but may place them in what order we please, as in Milton's elegant invocation at the beginning of *Paradise Lost*:

"Of man's first disobedience, and the fruit Of that forbidden tree, whose mortal taste Brought death into the world, and all our woe, With loss of Eden, till one greater man Restores us, and regain the blissful seat, Sing, heavenly muse."

In this sentence the transposition is almost as great as the Latin language would admit of, and the meaning as distinct as if Milton had begun with the plain language of prose.

Before we leave this head, we may remark, that the little attention which seems to have been paid to this peculiar advantage, derived from the use of an accusative case different from the nominative, is somewhat surprising. The Latins, who had occasion to attend to this with care than any other nation, and even the Greeks themselves have in many cases overlooked it; it is evident from the various instances we meet with in their languages where this is not distinguished.

For all nouns of the neuter gender, both in Greek and Latin, have in every declension their nominative and accusative singular alike. Nor in the plural of such nouns is there any distinction between these two cases; and in Latin all nouns whatever of the third, fourth, and fifth declensions, of which the number is very considerable, have their nominative and accusative plural alike. So that their languages have no advantage in this respect from almost one-half of their nouns. Nor have any of the modern languages in Europe, however much they may have borrowed from the ancient languages in other respects, attempted to copy from them in this particular; from which perhaps more advantage would have been gained than from copying all the other supposed excellencies of their language. But to return to our object.

It remains that we consider whether the inflection of nouns gives any advantage over the method of defining them by prepositions, in point of distinctness and precision of meaning? But in this respect, too, the analogous languages must come off victorious. Indeed, this is the particular in which their greatest excellence consists; nor was it, we believe, ever disputed but that, in point of accuracy and precision, this method must excel all others, however it may be defective in other respects. We observed under this head, when speaking of verbs, that it might perhaps be possible to form a language by inflection which should be capable of as great accuracy as in the more simple order of auxiliaries; but this would have been such an infinite labour that it was not to be expected that ever human powers would have been able to accomplish it. More easy would it have been to have formed the several inflections of the nouns so different from one another as to have rendered it impossible ever to mistake the meaning. Yet even this has not been attempted. And as we find that those languages, which have adopted the method of inflecting their verbs, are more imperfect in point of precision than the other, so the same may be said of inflecting the nouns; for, not to mention the energy which the analogous languages acquire by putting the accent upon the noun, or its proposition, when in an oblique case, according as the subject may require,—to express which variation of meaning no particular variety of words have been invented in any inflected language,—they are not even complete in other respects. The Latin, in particular, is in many cases defective, the same termination being employed in many instances for different cases of the same noun. Thus the genitive and dative singular, and nominative and vocative plural, of the first declension, are all exactly alike, and can only be distinguished from one another by the formation of the sentence; as are also the nominative, vocative, and ablative singular, and the dative and ablative plural. In the second, the genitive singular and nominative and vocative plural are the same, as are also the dative and ablative singular, and dative and ablative plural, except those in -um, which are nominative, accusative, and vocative singular, and nominative, accusative, and vocative plural, alike. The other three declensions agree in as many of their cases as those do, which evidently tend to perplex the meaning, unless the hearer is particularly attentive to, and well acquainted with, the particular construction of the other parts of the sentence; all of which is totally removed, and the clearest certainty exhibited at once, by the help of prepositions in the analogous languages.

It will hardly be necessary to enter into such a minute examination of the advantages or disadvantages attending the variation of adjectives, as it will appear evident, from what has been already said, that the endowing them with terminations similar to, and corresponding with, substantives, must tend still more to increase the similarity of sounds in any language, than any of those particulars we have already taken notice of; and were it not for the liberty which they have, in transpositive languages, of separating the adjective from the substantive, this must have occasioned such a jingle of similar sounds as could not fail to have been most disgusting to the ear; but as it would have been impossible in many cases, in those languages where the verbs and nouns are inflected, to have pronounced the words which ought to have followed each other, unless their adjectives could have been separated from the substantives, therefore, to remedy this inconvenience, they were forced to devise this unnatural method of inflecting them also; by which means it is easy to recognise to what substantive any adjective has a reference, in whatever part of the sentence it may be placed. In these languages, therefore, this inflection, both as to gender, number, and case, becomes absolutely necessary; and, by the diversity which it admitted in the arranging the words of the several phrases, might counterbalance the jingle of similar sounds which it introduced into the language.

Having thus examined the most striking particulars in which the transpositive and analogous languages differ, and endeavoured to show the general tendency of every one of the particulars separately, it would not be fair to dismiss the subject without considering each of these as a whole, and pointing out their general tendency in that light: for, we all know, that it often happens in human inventions, that every part which composes the whole, taken separately, may appear extremely fine; and yet, when all these parts are put together, they may not agree, but produce a jarring and confusion very different from what we might have expected. We, therefore, imagine that a few remarks upon the genius of each of these two distinct idioms of language, considered as a whole, will not be deemed useless.

Although all languages agree in this respect, that they are the means of conveying the ideas of one man to another, yet as there is an infinite variety of ways in which we might wish to convey these ideas, sometimes by the easy and familiar mode of conversation, and at other times by more solemn addresses to the understanding, by pompous declamation, &c., it may so happen that the genius of one language may be more properly adapted to the one of these than the other, while another language may excel in the opposite particular. This is exactly the case in the two general idioms of which we now treat. Every particular in a transpositive language is peculiarly calculated for that solemn dignity which is necessary for pompous orations. Long-sounding words, formed by the inflection of the different parts of speech; flowing periods, in which the Language. attention is kept awake by the harmony of the sounds, and in expectation of that word which is to unravel the whole; if composed by a skilful artist, are admirably suited to that solemn dignity and awful grace which constitute the essence of a public harangue. On the contrary, in private conversation, where the mind wishes to bend itself with ease, these become so many slops which embarrass and perplex. At these moments we wish to transmute our thoughts with ease and facility, and be tired with every unnecessary syllable, and wish to be freed from the trouble of attention as much as may be. Like our state robes, we would wish to lay aside our pompous language, and enjoy ourselves at home with freedom and ease. Here the solemnity and windings of the transpositive language are burdensome, while the facility with which a sentiment can be expressed in the analogous language is the thing that we wish to acquire. Accordingly, in Terence and Plautus, where the beauties of dialogue are most charmingly displayed, transposition is sparingly used. In this humble, though most engaging sphere, the analogous language moves unrivalled; in this it wishes to indulge, and never tires. But it vain attempts to rival the transpositive in dignity and pomp: the number of monosyllables interrupts the flow of harmony; and, although they may give a greater variety of sounds, yet they do not naturally possess that dignified gravity which suits the other language. This, then, must be considered as the striking particular in the genius of these two different idioms which marks their characters.

If we consider the effects which these two different characters of language must naturally produce upon the people who employ them, we will soon perceive that the genius of the analogous language is much more favourable for the most engaging purposes of life—the civilizing the human mind by mutual intercourse of thought—than the transpositive. For as it is chiefly by the use of speech that man is raised above the brute creation—as it is by this means he improves every faculty of his mind, and, to the observations which he may himself have made, has the additional advantage of the experience of those with whom he may converse, as well as the knowledge which the human race have acquired by the accumulated experience of all preceding ages—as it is by the enlivening glow of conversation that kindred souls catch fire from one another, that thought produces thought, and each improves upon the other, till they soar beyond the bounds which human reason, if left alone, could ever have aspired to—we must surely consider that language as the most beneficial to society which most effectually removes those bars that obstruct its progress. Now, the genius of the analogous languages is so easy, so simple and plain, as to be within the reach of every one who is born in the kingdom where it is used to speak it with facility; even the rudest among the vulgar can hardly fall into any considerable grammatical errors; whereas, in the transpositive languages, so many rules are necessary to be attended to, and so much variation is produced in the meaning, that the slightest variations in the sound, that it requires to study far above the reach of the illiterate mechanic over to attain. So that, how perfect soever the language may be when spoken with purity, the bulk of the nation may ever labour under the inconvenience of rudeness and inaccuracy of speech, and all the evils which this naturally produces. Accordingly, we find that in Rome a man, even in the highest rank, received as much honour, and was as much distinguished among his equals, for being able to converse with ease, as a modern author would be for writing in an easy and elegant style; and Caesar, among his contemporaries, was as much esteemed for his superiority in speaking the language in ordinary conversation with ease and elegance, as for his powers of oratory, his skill in arms, or his excellence in literary composition. It is needless to point out the many inconveniences which this must unavoidably produce in a state. It is sufficient to observe, that it naturally tends to introduce a vast distinction between the orders of men—to set an impenetrable barrier between those born in a high and those born in a low station—to keep the latter in ignorance and barbarity, while it elevates the former to such a height as must subject the other to be easily led by every popular demagogue. How far the history of the nations who have followed this mode of language confirms this observation, every one is left to judge for himself.

The first stage of language exhibits single words, chiefly nouns or verbs, in a short form, and with a minimum amount of inflection, some subordinate to the others, but still separate words. The Chinese is usually considered to be the type of language in this state.

In the second stage the subordinate words coalesce with the main ones, but not so as to wholly conceal their originally separate existence. Languages in this stage are called agglutinative. This is the state in which most of the languages of the world exist. The Manchu and Mongol are the usual examples of this condition. Most other tongues, however, would serve as well.

The coalition of the subordinate with the main word having become so perfect as for the former to look like a part of the latter rather than a word originally separate, the combination becomes amalgamate instead of agglutinate, and the language inflectional. The Greek and Latin are the types of this form.

Inflections fall off, and get displaced by separate words, these words being of a peculiar kind—prepositions with substantives, auxiliaries with verbs. The English is in this stage, and, doubtless, it is destined to further changes.

These, however, will not be in a cycle. It is only in appearance that the fourth stage is in the same predicament as the first. It is only in the comparative absence of inflections, and the separate existence of the subordinate parts of sentences, that they agree. The subordinate words themselves differ.

An English auxiliary, or proposition, in its more typical form, is meaningless when taken by itself. In I have ridden a horse, the word have originally meant, I possess a horse as a ridden thing; where, however, is the idea of possession in I have been, a sentence exhibiting language in the most abstract or supra-sensible form in which it has ever existed?

In the difference of character between the subordinate words of sentences, we have a means of ascertaining whether languages in the same non-inflectional state have lost their inflection, or have yet to develop them. The question, however, is not asked so often as it ought to be.

In the actual field of language the lines of demarcation are less definitely marked than in the preceding sketch. The phenomena of growth, however, are, upon the whole, what it suggests.

That languages like the English, French, &c., are more advanced, grown, or developed, than the classical tongues, is certain, whatever may be the interpretation of the fact in a question as to the relative values of the different stages.—[n. o. L.]

II.—AFFINITIES OF LANGUAGE.

The study of the affinities of various languages is so far one of the most important of all branches of human knowledge, as it affords, when properly applied, an unerring test of the truth or falsehood of historical evidence without which it would sometimes be impossible to unravel the mysteries of contradictory testimonies respecting the relations of the different races of mankind. We have, for example, no traditional evidence in support of any connection between the ancient Egyptians and the Indians; whilst, on the other hand, a number of persons who came with the English army from the East Indies into Egypt, were so strongly impressed with the resemblance of the Egyptian and Indian temples, which appeared even to excite the religious feelings of many of the natives who were amongst the troops employed, that a very general inclination has arisen from these circumstances to consider the Egyptian mythology as merely a branch of the Indian. But if the Egyptian people had really been of Indian origin—that is, if the Egyptians and Indians had really been one people, at any later period than that at which the whole of the Indian and European races were separated from their common stock—the languages of India and of Egypt could not but have exhibited some features of resemblance, which would have preserved the traces of the connection; whilst, in fact, there is much less similarity between the Egyptian and the Indian, than between the Indian and the Greek, or the English and the Persian; so that etymology may here be adduced as confirming the evidence, or as justifying the silence, of history; and the resemblance of the mythological representations must be considered as in great measure accidental. Language.

[Similarity of language is strong proof of evidence in favor of a common origin; but it is nothing more. The strongest instances of a mother-tongue having been forgotten or unlearned, and a new language adopted in its stead, can be found amongst the negroes of the New World. In St Domingo, the languages are French and Spanish; in the United States, English; and in South America, Dutch, Spanish, and Portuguese. Here, however, the conditions were peculiar. The native language was no longer connected with the soil to which it was indigenous, but transplanted to a new area. Neither was it any single homogeneous form of speech that was obliterated, but, on the contrary, a multitude of mutually unintelligible tongues, which, under any circumstances, would have ended in the establishment of a Lingua Franca.

With the negro languages of the New World we have the maximum amount of change in speech with a minimum in the way of intermixture of blood. With the native American tribes, the phenomena of change are somewhat different. As a general rule, the number of individuals who speak one and the same language is remarkably small; every now and then, however, in contact with these small patches of speech, is to be found some language spread over a considerable area, and spoken by several tribes. These the missionaries have converted into Lingua Franca; it being a matter of observation that an American Indian learns an American language, no matter how unlike his own, easier than one from Europe. Many of the minor languages of the South American republics and Brazil have thus been replaced by the Guarani.

In each of the previous cases, there is the actual replacement of one language by another. In the Lingua Franca of Europe this is not always the case. In the Levant, the Italian (the Lingua Franca, or its basis) is spoken by numerous Arabs, Greeks, &c. The native Arabic, however, and the native Greek, co-exist by the side of it. There is no extinction as yet. Nevertheless, the tendencies towards it have set in, as much as where two languages have to be learned, the less useful is the weakest, and has a chance of going to the wall.

Sometimes with two languages thus brought in contact with each other we have the phenomenon of intermixture, rather than obliteration—an intermediate tongue being formed out of the fusion of two. A priori, it seems likely that such should be the case often. In reality, however, the development of such a language as (say) C, out of languages A and B, is very rare indeed. The ordinary phenomenon is A with a certain amount of B, or B with a certain amount of A—the original character of the fundamental language being preserved. The English (for instance), for all its Latin elements, is German; the French, for all its German elements, Latin.

This suggests, that in language we have every degree of change, from simple intermixture to absolute obliteration and replacement. It also suggests, that similarity of language is a matter of degree. There may be absolutely community of tongue, or there may be an admixture of say one per cent. of foreign terms. Language is one of those signs of community of origin which is slow to be abolished—slower than most others—slower, perhaps (on the whole), than any other; nevertheless, it is only a sign, and a sign capable of obliteration. Its relative permanence, when compared with other criteria, is a matter upon which there is a wide discrepancy of opinion; the facts upon which our hypotheses must rest being by no means easily ascertained. It is only certain that the questions involved in it are far too complicated to be disposed of by the application of any general rule. As new ideas are introduced, language changes. As new physical influences are brought into action, the anatomical conformation of the human body becomes modified. That these latter forces have some influence is universally admitted; though many competent authorities put a close limit on its extent. It is clear, however, that, within certain limits, both language and physical conformation may change.

They may change at different rates—i.e., in a given period (say ten generations), the speech may be considerably modified, whilst the anatomy of the speakers remains the same. And, vice versé, the physiognomy may alter, whilst language remains fixed. Every comparison of the difference of rate between such changes should be made on the merits of the particular question under notice, no general rule being sufficient.

Next comes the question of race. Here we may safely say, that the range of change in language is wider than that of language, which physical form is susceptible. It is, clearly, easier for a negro to be converted into a Frenchman in the matter of language, than in that of colour. Extreme forms of language may easily be converted into each other, even extreme forms of physical conformation; and, in fact, all that can fairly be said, is that it is by no means certain that a population of negroes, transplanted from a low alluvial swamp to an elevated mountain range, would not retain their language without alteration, than they would their physical form—within certain limits.

The contact of two languages has a greater tendency to effect the obliteration of one of them than the development of a tertium quid out of their fusion.

The contact of different stocks in the way of physical union has a greater tendency to effect the intermixture of blood than the obliteration of one of the constituent elements.

From this it follows, that languages are much more either one thing or another than stocks, races, or families. The language of Radnorshire and Cornwall is much more English (as opposed to Welsh) than the blood or pedigree of its speakers is English; indeed, as a general rule, the blood of a given population is more mixed than its language. This is because, whilst A and B, in the way of stock, blood, or pedigree, will give C (a true tertium quid, or a near approach to it), A and B, in the way of language, will only give themselves,—i.e., they will give no true tertium quid, nor any very close approach to it. These, however, are matters that belong to the question of man in general rather than to language. Language (as an instrument of criticism in ethnology) is the most permanent of the criteria of human relationship derivable from our moral constitution, and, in some cases, equally permanent with physical form, though, in the case of extreme changes, less so.

The particular illustration of the original text to which this notice is a comment, drawn from the languages of India and Egypt, is exceptionable. The languages of India are two in number. Of these it is only one, the Sanscrit, that is more Persian, &c., than Egyptian. Of the other, the Tamil, &c., no such statement can be made. It is, probably, more Egyptian than Persian—certainly more Egyptian than Greek or English—pop. c. l.

It is, however, only with regard to the languages of the ancient world that we can feel much interest in such an investigation. The American dialects might afford equally extensive subjects of speculation in a metaphysical and critical point of view; but the concerns of barbarians, unconnected and remote from all contact with literature or civilisation, and destitute of all historical records, will scarcely be thought to require any great portion of attention from a philosophical inquirer; and there is ample scope for the employment of all our faculties in the analysis and comparison of the various languages of Europe, Asia, and Africa. If, indeed, an extraordinary exertion of enterprise and industry, which can be expected from a few distinguished individuals only in the course of as many centuries, should make known relations, such as Alexander von Humboldt has appeared to discover, between the American and Asiatic nations, a new field would be opened for the gratification of our curiosity; but it can scarcely be expected that these points of resemblance can be sufficiently numerous to afford any thing like demonstrative evidence, until the whole subject has been much more deeply and repeatedly discussed. In the mean time, a very brief enumeration of the names of the American languages is all that can be required, on an occasion like the present; except the insulated though interesting remark, that the countries separated by Behring's Straits exhibit, as might indeed be expected, strong resemblances in some of their languages.

[It is in the ruder languages that the important phenomena of development and growth—the laws of language—are best studied. The following extracts are from an opuscule of the writer on the subject:

"A little consideration will show, that that difference between the study of a given subject in its general and abstract, and the study of one in its applied or concrete form, which finds place in so many departments of human knowledge, finds place in respect to language and languages. It finds place in the subject before us as truly as it does in the science of the laws of life—physiology or biology. Just as there is therein a certain series of laws relating to life and organization, which Language would command our attention, if the whole animal and vegetable world consisted of but a single species, so the study of speech would find scope in a well-defined system of education, even if the tongues of the whole wide world were reduced to a single language, and that language to a single dialect. This is because the science of life is one thing, the science of the forms under which the phenomena of life are manifested, another. And just as physiology, or biology, is, more or less, anterior to, and independent of, such departments of study as botany and zoology, so, in the subject under notice, there is the double division of the study of languages in respect to structure and development, and the study of languages as instances of the variety of form in which the phenomenon of human speech exhibits, or has exhibited, itself. Thus—

"When (as I believe once to have been the case) there was but a single language on the face of the earth, the former of these divisions had its subject-matter. And—

"When (as is by no means improbable) one paramount and exclusive tongue, developed at first, rapidly and at the expense of the smaller languages of the world, and, subsequently, slowly and at that of the more widely-diffused ones, shall have replaced the still numerous tongues of the nineteenth century, and when all the dialects of the world shall be merged into one universal language, the same subject-matter for the study of the structure of language, its growth and changes, will still exist.

"So that the study of language is one thing; the study of languages another.

"One main distinction between the study of language and the study of languages lies in the fact of the value of the former being constant, that of the latter fluctuating. The relative importance of any two languages, as objects of special attention, scarcely ever remains steady. The value, for instance, of the German—to look amongst the contemporary forms of speech—has notably risen within the present century. And why? Because the literature in which it is embodied has improved. Because the scientific knowledge which, to all who want the key, is (so to say) locked up in it, has increased some hundred per cent.

"But it may go down again. Suppose, for instance, that new writers of pre-eminent merit ennoble some of the minor languages of Europe—the Danish, Swedish, Dutch, &c. Such a fact would divide the attention of serious—attention which can only be bestowed upon some second, at the expense of some first object. In such a case, the extent to which the German language got studied would be affected much in the same way as that of the French was by the development of the literature of Germany.

"Or the area over which a language is spoken may increase, as it may also diminish.

"Or the number of individuals that speak it may multiply—the area being the same.

"Or the special application of the language, whether for the purpose of commerce, literature, science, or politics, may become changed. In this way, as well as in others, the English is becoming, day by day, more important.

"There are other influences.

"High as is the value of the great classical languages of Greece and Rome, we can easily conceive how that value might be enhanced. Let a manuscript containing the works of some of the lost, or imperfectly preserved, writers of antiquity be discovered. Let, for instance, Gibbon's desiderata—the lost Decads of Livy, the Orations of Hyperides, or the Dramas of Menander—be made good: the percentage of classical scholars would increase, little or much.

"Some years back it was announced that the Armenian language contained translations, made during the earlier centuries of our era, of certain classical and ecclesiastical writings, of which the originals had been lost—lost in the interval. This did not exactly make the Armenian, with its alphabet of six-and-thirty letters, a popular tongue; but it made it, by a fraction, more popular than it was in the days of Whiston and La Croze, when those two alone, of all the learned men of Europe, could read it.

"Translations tell in another way. Whatever is worth reading in the Danish and Swedish is forthwith translated into German. E.g., Professor Retzius of Stockholm wrote a good Manual of Anatomy. He had the satisfaction of seeing it translated into German. He had the further satisfaction of hearing that the translation ran through five editions in less time than the original did through one.

"Now, if the Germans were to leave off translating, the value of the language in which Professor Retzius wrote his Anatomy would die.

"Upon the whole, the French is, perhaps, the most important language of the nineteenth century; yet it is only where we take into consideration the whole of its elements of value.

"To certain special purposes, the German is worth more; to the artist, the Italian; to the American, the Spanish. It fell, too, in value, when nations like our own insisted upon the use of their native tongues in diplomacy. It fell in value because it became less indispensable; and another cause, now in operation, affects the same element of indispensability. The French are beginning to learn the languages of other nations. Their own literature will certainly be none the worse for their so doing. But it by no means follows that that literature will be any the more studied; on the contrary, Frenchmen will learn English more, and, pro tanto, Englishmen learn French less.

"What is the import of such sounds as that of the letter s in the word father-s? It is the sign of the plural number.

"Such is the question—such the answer. But is the answer a real one? Is it an answer at all? How come such things as plural numbers, and signs of plural numbers, into language? How the particular plural before us came into being, I cannot say; but I can show how some plurals have. Let us explain the following—

| Ngi = I. | Ngi-n-de = We. | | Ngo = Thou. | Ngo-n-da = Ye. | | Ngu = He. | Nge-n-da = They. |

The da (or de) in the second column is the sign of the plural number in a language which shall at present be nameless. It is also the preposition with. Now with denotes association; association, plurality. Hence

| Ngi - n - de = I + = We. | | Ngo - n - da = Thou + = Ye. | | Ngu - n - da = He + = They. |

This is just as if the Latins, instead of nos and vos, said mecum and te-cum.

"Such is the history of one mode of expressing the idea of plurality; we can scarcely say of a plural number. The words plural number suggest the idea of a single word, like fathers, where the s is inseparably connected with the root, at least so far inseparably connected as to have no independent existence of its own. Ngi-n-de, however, is no single word at all, but a pair of words in juxtaposition, each with a separate existence of its own. But what if this juxtaposition turns into amalgamation? What if the word da changes? What if it becomes t or z, or th, or ef? What if, meanwhile, the separate preposition do change in form also; in form or meaning, or perhaps in both? In such a case a true plural form is evolved, the history of its evolution being a mystery.

"So much for one of the inflections of a noun. The remaining words illustrate one of a verb.

"Hundreds of grammarians have suggested that the signs of the persons in the verb might be neither more nor less than the personal pronouns appended, in the first instance, to the verb, but afterwards amalgamated or incorporated with it. If so, the -m in ingue-ns is the m in me, &c. The late Mr Garrett, a comparative philologist whose reputation is far below his merits, saw that this was not exactly the case. He observed that the appended pronoun was not so much Personal Possessive; that the analysis of a word like ingue-ns was not so much, say-I, as saying + my; in short, that the verb was a noun, and the pronoun either an adjective (like meus) or an oblique case (like mel), agreeing with or governed by it.

"It is certainly so in some cases. In a language which, at present, shall be nameless, instead of saying my apple, thy apple, they say what is equivalent to apple-me, apple-th, &c.—i.e., they append the possessive pronoun to the substantive, and by modifying its form, partially incorporate or amalgamate it. They do more than this. They do precisely the same with the verbs in their personal as they do with the nouns in their..." Language, possessive relations. Hence, olvas-om, &c., is less I read than my-reading; less read+I, than reading+my.

Olvas—om = I read. od = Thou readest. uk = We read. atok = Ye read.

Alma—m = My apple. d = Thy apple. nk = Our apple. tok = Your apple.

"I submit, that facts of this kind are of some value, great or small. But the facts themselves are not all. How were they got at? They were got at by dealing with the phenomena of language as we found them, by an induction of no ordinary width and compass; and many forms of speech had to be investigated before the facts came out in their best and most satisfactory form.

"The illustration of the verb (olvasom, and almaom, &c.) is from the Hungarian; that of the plural number (ngidje, &c.), from the Tumali—the Tumali being a language no nearer than the negro districts to the South of Cordova, between Sennar and Darfur, and (as such) not exactly in the highway of literature and philology." (Lecture on the Importance of the Study of Language as a Branch of Education for all Classes, by Robert Gordon Latham, M.D., F.R.S.)

Of language in general we do not here intend to treat, but merely of languages as they are distinct from each other. It is not, however, very easy to say what the definition ought to be that should constitute a separate language; but it seems most natural to call those languages distinct, of which the one cannot be understood by common persons in the habit of speaking the other, so that an interpreter would be required for communication between persons of the respective nations. Still, however, it may remain doubtful whether the Danes and the Swedes could not, in general, understand each other tolerably well, and whether the Scottish Highlanders and the Irish would be able to drink their whisky together without an interpreter; nor is it possible to say, if the twenty ways of pronouncing the sounds belonging to the Chinese characters, ought or ought not to be considered as so many languages or dialects, though they would render all oral intercourse between the persons so speaking the language actually impracticable. But, whether we call such variations different languages, or different dialects, or merely different pronunciations of the same dialect, it is obvious that they ought all to be noticed in a complete history of languages; and, at the same time, that the languages so nearly allied must stand next to each other in a symmetrical order; the perfection of which would be, to place those languages nearest together in which the number of coincidences in the signification of words throughout the language are the most numerous.

It has sometimes been imagined, that all languages in existence present something like a trace of having been deduced from a common origin; and it would be difficult to confute this opinion by negative evidence, unless every separate language had been very completely analysed and examined by a person well acquainted with a variety of other languages, with which it might be compared. But, without such an examination, the opinion must remain conjectural only, and no more admissible as demonstrated, than the opinions of some empirics, that there is only one disease, and that the only remedy for it is brandy. In an essay on probabilities, lately published in the Philosophical Transactions, Dr Young has remarked, that "nothing whatever could be inferred, with respect to the relation of two languages, from the coincidence of the sense of any single word in both of them," that is, supposing the same simple and limited combinations of sounds to occur in both, but to be applied accidentally to the same number of objects, without any common links of connection; "and that the odds would only be three to one against the agreement of two words, but if three words appeared to be identical, it would be more than ten to one that they must be derived, in both cases, from some parent language, or introduced in some other manner," from a common source; whilst "six words would give near 1700 chances to one, and eight near 100,000; so that, in these last cases, the evidence would be little short of absolute certainty."

[Few applications of mathematical reasoning to questions beyond its usual field, have commanded more attention than this calculation of Dr Young's, under whose name it is quoted in most works that support the doctrine of the fundamental unity of languages. Its scientific appearance has, doubtless, recommended it. It is exceptionable, however, on the score of its only meeting one out of the several causes of similarity of meaning accompanied with similarity of form in different languages. It only deals with the coincidences referable to chance or accident; of these, however, no advanced philologue takes any account either way; indeed, it is probable that an advocate of the unity of language might allow his opponent some dozen or scores of similar words in different languages, without either drawing any inference in favour of his own views himself, or expecting from his adversary any anticipation of any argument whatever founded upon them. The simple question of character language has ceased to be one of any importance. The really important question is one of the kind to which the writer in Mill's Logic, on what the author calls the Collocation of Causes, has given prominence. Here a variety of causes may end in the same effect. Nowhere is this commoner than in language. Let two sounds have a tendency to change into a third—say m and b into v, and the combination vav may come out of base and man equally. The field of language is full of instances of this kind. Then there are the words that resemble one another, independent of imitation, either from being onomatopoeic, or imitative of certain sounds, or organically connected, i.e., referable to a tendency to express certain ideas by certain combinations. Certain ideas come earlier than others; so does the capability of pronouncing certain sounds. The former will naturally be represented by the latter. The sound of r, which comes late, is not likely to enter into the child's name of his papa or mama—ideas expressed, half the world over, by the same sounds. In the investigation of the laws by which the articulate sounds are distributed over the different languages of the world little has been done. In many cases, however, the sounds (r, th, &c.) that children learn last, are those that the rudier languages most want.—n. o. l.]

The author of the article in the Quarterly Review, on Adelung's Mithridates, observes, that, setting out from the establishment of a certain number of separate languages as species, "we may proceed to comprehend, in the description of one family, such as have more coincidences with each other than diversities, and to refer to the same class such families as exhibit any coincidences at all that are not fortuitous, imitative," that is, from onomatopoeia, "or adoptive. In order, however, to avoid too great a number of classes, which would arise from an inadequate comparison of languages imperfectly known, it may be proper, in some cases, to adopt a geographical distinction, as sufficient to define the limits of a class, or to assist in its subdivision into orders. We are thus obliged to employ an arrangement of a mixed nature;" and, in fact, the tests of affinity here proposed depend so much on the progress of our knowledge in the study of each language, that the results must unavoidably be liable to great uncertainty and fluctuation, of that we can reasonably expect nothing more than an approximation to an arrangement completely methodical.

"If," continues the reviewer, "the resemblance or identity of a single word in two languages, supposed to be exempt from the effects of all later intercourse, were to be esteemed a sufficient proof of their having been derived from a common stock, it would follow, that more than half the languages of the universe would exhibit traces of such connection, in whatever order we might pursue the comparison. Thus we find in a very great number, and perhaps in a majority of known languages, that the sound of the vowel o, with a labial consonant, is employed for the name of Father; and if this be supposed to be something like an onomatopoeia, or an application of the first sounds which an infant naturally utters, the same reason cannot possibly be assigned for the still more general occurrence of the combination sm in the term smee, which is by no means likely to have originated from any natural association of this kind. But neither these points of resemblance, Language, nor any other that can be assigned, are absolutely universal; for, besides the numberless varieties referrible more or less immediately to Abba, Father, we have at least twenty different and independent terms for the same relation in the old world;" Tia, Iusa, Flor, Haur, Rama, Dian, Gira, Kuten, Assinalagi, Medua, Thanes, Siah, Tot, Muthen, Meese, Indus, Nau, Nome, Manung, Danegale, Ray, Tikib, and Oa; and about as many for Name, besides those languages in which the version of an abstract term of this kind is less likely to have been ascertained; Ming, Trea, Diant, Shu, Hosura, Shem, Tiarchip, Ad, Nipta, Lion, Sucheli, Asein, Wasta, Nopla, Taire, Sunna, Ran, Hilli, Ding, Dhai, and Amphara." At the same time, therefore, that we venerate the traces of our common descent from a single pair, wherever they are still perceptible, we must not expect to find them in all existing languages without exception; and an Etymological Universal, considered as intended to establish such a perfect community of derivation, can only be regarded as a visionary undertaking. Nor must we neglect to unite, in some common arrangement of classification, those languages which have the words here specified, or any other radical words, in common, as incomparably more related to each other than the Chinese to the Cantabrian, or the Irish to the Hottentot.

"The gradations by which a language is likely to vary in a given time, seem to be in some measure dependent on the degree of cultivation of the language, and of the civilization of the people employing it. From Homer to the Byzantine historians, the Greek language remained essentially the same for 2000 years; the German has varied but little in 1500; and even the English, notwithstanding its mixture with French and Latin, has altered but three radical words out of the fifty-four which constitute the Lord's Prayer, in the same period. On the other hand, a few barbarians in the neighbourhood of Mount Caucasus and of the Caspian Sea, of modern origin, and ignorant of the art of writing, are divided into more nations speaking peculiar languages, radically different from each other, than the whole of civilized Europe. In such cases, little light can be thrown upon history by etymological researches; while, with regard to more cultivated nations, we obtain, from the examination of their languages, historical evidence of such a nature as it is scarcely possible for either accident or design to have falsified."

[The study of the rate at which languages change must begin with the history of those tongues which we know in more stages than one. These are few in number; being only those which are cultivated and written. In the fact of their being known in one form only (combined with the absence of a literature) lies the true reason for comparative unimportance of the rudest languages. As languages—i.e., as data for the natural history of speech—they are valuable; as they are as instruments of etymological criticism.

Amongst the languages known to us in more stages than one, the classical languages are the chief; next to these the Hebrew, Sanskrit, Arabic, Old Slavonic, Anglo-Saxon, &c. Up to the present time they have told us, over and above the fact, that languages change at different rates, the more important one, that the change after the same fashion or upon the same principle. The Roman differs from the ancient Greek; the Italian, &c., from the Latin; the English from the Anglo-Saxon; the Danish from the Icelandic; all in the same way—i.e., in the use of prepositions and auxiliary verbs instead of cases and moods and tenses.

This means that certain parts have a tendency to change in a definite manner. Further researches will show that they generally change in the same order, e.g., the sign of a dative case will go before that of a genitive, &c.

Hence, there is not only a question of parts changed, but also that of the order and rate of change; and, lastly, that of the forces that regulate this rate. They are somewhat different from those indicated in the text, the influence of writing and cultivation being less than is usually imagined. What they are in detail, however, has yet to be investigated.

The best exhibition of the transition of a language from one stage to another, is to be found in Petersen's History of the break-up of the Icelandic, or old Scandinavian language, into the modern tongues of Denmark, Norway, and Sweden. His facts appertaining to rate are interesting. The Danish is generally about a century in advance of the Swedish—i.e., the Swedish of a.d. 1300, or a.d. 1600, is as archaic as the Danish Language of a.d. 1200, or a.d. 1700; the Danish of 1600 being, there or thereabouts, of the form of the Swedish of 1700.

So much for the regularity of relation between these two. Turn from them to the Norwegian. Up to the time of the Reformation it hardly changed at all; afterwards it changes in about one century as much as the other two tongues had in four centuries.]

According to the exposition of Professor Adelung, it seems not improbable that Tibbet, on the E. of Cashmere, may "have been the habitation of Adam immediately after his fall, and the country occupied by the descendants of Cain. In Tibbet, and in the countries immediately beyond it, the languages of at least a hundred and fifty millions of people are still principally monosyllabic; and from this peculiarity, as well as from the singular simplicity of their structure, they are supposed to constitute the most ancient class of existing languages, though it must be confessed that much of Adelung's reasoning on this subject is extremely inconclusive." Mr Townsend remarks very judiciously, that one of the canons of Rudbeck is by no means admissible. He states, that a language which has numerous monosyllabic expressions is a parent language. The English has more than 3700 monosyllabic expressions, and the Chinese has none but such; yet neither of them is, for that reason, to be considered as a parent language. Certain it is, that all languages, by abbreviations, have a tendency to become monosyllabic, and therefore a language which abounds in monosyllables is ancient, and these commonly are the most antiquated parts of every language. New compounds are incessantly created. These are abbreviated, and in process of time become monosyllabic. In deriving, therefore, a word in one language from its correspondent expression in some other language, we must ever bear in mind, that, unless in the formation of new compounds, the least abbreviated is commonly the parent, and the most abbreviated its offspring. Would it be possible for any one to persuade us that Colophons was derived from Cuff, or Blasphemy from Blame? "A similar instance," says the reviewer, "might be found in Trachelos and Hals of the Greeks and Germans; for certainly Hals is more like Trachelos than like Colophon."

(Monosyllabicism is one thing: the want of inflection is another. Wherever there is a want of inflection, there is the word in its simpler form—i.e., a root minus any adjuncts. Without this absence of adjuncts, there is but little chance of words being monosyllabic. It may occur, however, and yet leave the roots disyllabic or even polysyllabic—since there is no necessity for its being limited to one syllable; though such is, generally, the case. On the contrary, inflections may coalesce with the root, and so exist without increasing the number of syllables. Logic is as much a monosyllable as dog.

Referring to what has been said about loss and non-development in the way of inflection, let us ask whether, assuming a development to be in its first stage, it is therefore a parent language? No. It has simply changed more slowly than others.

This leads to see that the word old has two meanings.

The English of America is, in one sense, older than the English of Australia. It separated earlier from the parent tongue.

Yet it may not be the oldest one thousand years hence. Suppose the Australian, by changing the slower, preserves more old forms. In that sense it will be the older language.

—A. L.] Language modifications which the character has conventionally undergone.

And in this point of view the Chinese would require to be classed with the old Egyptian only, since we know of no other language which was habitually expressed in hieroglyphics and their immediate derivatives. It is not at all uncommon for the same sound in Coptic, as in Chinese, to have four or five senses all essentially different, as may easily be observed in turning over a dictionary; more, for instance, means Bach, and Thomas and a Sheep in two verses of St Matthew (v. 45, 46), and perhaps several other things.

This is chiefly a question of writing, not one of speaking; a question concerning the representation of language, not language itself. In the importance of the accent we recognize a point of speech; and it indicates poverty of vocabulary; but the expression of it is wholly a matter of orthography. In the history of the alphabet, the peculiarities of Egypt and China find place; not in the history of language.

Another ancient and extensive class of languages, united by a greater number of resemblances than can well be altogether accidental, may be denominated the Indo-European, comprehending the Indian, the West Asiatic, and almost all the European languages. If we choose to assign a geographical situation to the common parent of this class, we should place it to the south and west of the supposed origin of the human race; leaving the north for our third class, which we can only define as including all the Asiatic and European languages not belonging to the two former; which may be called Asiatic, or, perhaps, without much impropriety, Tataric; and which may be subdivided into five orders, Sporadic, Caucasian, Tartarian, Siberian, and Insular. The African and American languages will constitute a fourth and fifth class, sufficiently distinct from all the rest, but not intended to be considered as any otherwise united among themselves, than by their geographical situation. There is, indeed, little doubt, that some of the languages here called Tataric are essentially allied to others, which are referred to the Indo-European class; but they have been too little investigated to allow us to make the selection that would be required for completing the classification. (The extent to which this classification is accurate or inaccurate will be seen in the sequel.—z. o. l.)

The following tables are copied, with considerable additions, from Adelung's Mithridates. The words Heaven and Earth are chosen as specimens, because they seem to be known in a greater number of languages than any others, except the name of Father, which is supposed to exhibit, in some cases, a fallacious similarity. The German orthography has been principally employed, except in such languages as are usually written in the Roman characters, the pronunciation of the consonants being more uniform than in English, and that of the vowels differing little from the Italian.

The selection of words representative of rude and unknown languages, has commanded no small amount of the attention of philologists and ethnologists; and many lists, ready prepared, are in circulation—in India and North America more especially. They are none of them unexaminable; and the reason for their being so lies in the fact of the choice being made on a priori views of what words are fundamental and what not. Actual observation tells us, that the most permanent parts of languages are by no means the words that a priori speculations indicate.

Among the best words for specimens are the names for fire, water, sun, moon, and star. Man and woman are generally ambiguous, the names for them being often the names for man and wife as well. Hence, unless we have the terms for all four (man, husband, woman, wife), we have but half the requisite information. It may also be added, that along with the name for sun, the names for sky (heaven), light and day, should be given.

The numerals are always of value, because, whether they be of little or great use for the purposes of comparative philology, they have always a value in the history of the arithmetic. Their pure and proper philological importance, however, is inconstant. Sometimes the numerals of two or more languages shall be alike whilst the rest of the vocabulary differs. Sometimes the similitude between the words other than numeral shall be great, the numerals themselves being unlike. In many languages it happens, that if some of the numerals are alike, the others will be so also. In others, on the contrary, it by no means follows that because (say) one and two are alike, three Language, and four, &c., should be equally so. As a class, the names of the parts of the human body are, perhaps, the best.—z. o. l.

Classes, Orders, and Families of Languages.

I. MONOSYLLABIC. 1. Chinese 2. Siamese 3. Avanese 4. Tibetan

II. INDO-EUROPEAN. 5. Sanskrit 6. Median 7. Arabian 8. Lycean 9. Phrygian 10. Greek 11. German 12. Celtic 13. Etruscan 14. Latin 15. Cantabrian 16. Slavonic

III. TATARIC. (1.) Sporadic. 17. Tchadish 18. Hungarian 19. Albanian (2.) Caucasian. 20. Armenian 21. Georgian 22. Abassan 23. Circassian 24. Ossetish 25. Kistic 26. Lesgian

IV. AFRICAN.

V. AMERICAN.

Families, Species, or Distinct Languages, and Varieties or Dialects, with Specimens.

I. MONOSYLLABIC CLASS.

Heaven, Sky. Earth.

1. CHINESE Tien, Li Ti, To Fo Kien Tshio Tshio Tonquinese Thien, Blod Dat, Dis Los

2. SIAMESE Sa Wang Din (Maan, Pho chal) (Hood, Mu)

3. AVANESE Mo kams, Nip ban Lu pu, Milé Pyama Mo kams La pri f Tre Rubheng

4. TIBETAN Nam khei (Dak) Dahk ten, Sa

II. INDO-EUROPEAN CLASS.

5. SANSKRIT Paramandale, Vana, Bumi, Stiria Aagasa, Svarga, Veigunda, Arthalogia, Nibu (Maa, Parusha) Prabrit Saggio (Maa, Parilo) Bali Saggio (Maa, Burut-sa) Devanagri Ardwa, Arthalogia Buma Nepal Assam Assam Tiperah Kassal Bengaloo Hindoo Urda Brijbasa Jyura Hindustane Moors Asmaan, Mukti Sijmien, Dakhmin, Dania | Language | Heaven, Sky | Earth | Language | |----------|------------|-------|----------| | Udiapura | Ashaman | Terti | Ge | | Renares | | | Airtha | | Manipura | | | Erdu | | Gondoo | | | Erde | | Orissa | Paramandal | Bumi, Bumilo | Jorda | | Telugu | (King, Raja) | Nacla | | | Carnatic | | | Hardi | | Mārava | Wana, Mama, Paramandala (pam) | Pumili, Nawarg | Eere, Erda | | Tamul | | | Jerda, Yrteik | | Malelam | Wana (Bread, Ap.) | Bhumi, Samina | Eerde, Wrroll | | Malabar | Asmanu, Agasha | Paramandala, Pama | Aerde, Eertryke, 1270 | | Kanara | Welkuntha, Agasha | Purtimir, Soumsar, Zimmin | Jord, Jera | | Decan | Sorgi | | Yurn | | Kunkuna | | | Jord | | Mahratta | Welkuntham, Saargi, Agasha | Pumandi, Saummar, Puma | Jord, Jordriki | | Garment | | | Jord | | Belochee (Afghan) | Paramanda | Bumi | Jord | | Biharee | | | Jord | | Sindh | | | Jord | | Multan | Oshman | Dhemi | Eortha | | Gipsy | Amengi, Teheros | Pu, Phu, Pube | Birth, Yearth | | Wuch | | | | | Sikh | | | | | Cashmir | | | | | Kuch | (1. Katka; 3. Tuhmka; Mom, Mipa; Father, P'ha) | | | | Mahārājām | Ouddou, Uda | Bin | Talu | | Cingalas | Swarga, Abasa | Bumidahe, Bami | Talanch, Dtalmhuin | | Malayam | Surga | | Talu | | Sumatran | | | Talushin | | Bataran | (Day, Terangharik) | Tana | | | Rejang | (Day, Illeytung) | Pihta | | | Lampung | (Day, Rannih) | Tanno | | | Achém | (Day, Ural) | Tano | | | Néos | | | | | Popgy | | | | | Javanese | Surga, Dilangin (Night, Malang) | Lemma, Darat | Talambh, Thillamb, Talu | | Borneo | | | Talanch, Dtalmhuin | | Andaman | Madamo (Head, Ta-bal) | Totonguandshi | Talu | | Zendish | Tahkhre, Sakhter, Za, Zao, Zemo, Zemeno | | Talushin | | Pehlviš | Taherk, Shmehs, Seper | Zivanand, Arta, Damik, Leksa, Bamih | | | Persian | Asmon | | | | (Bucharian) | Kurdish | Asman, Banta | Ard, Sigit, Chnak, Chbel | Terra, Tellus | | Kurdish | Asmo, Asman | Smak, Smilge, Zmuku | Terra | | Afghan | | | Terra | | Syriac | Shemalo | Aro, Areto | Terra | | Aramaic | Simmi | Diali | Terra | | Phoenician | Punet | | Terra | | Hebrew | Shamaim | Arez | Terra | | Chaldee | Shemaiia | Ara, Arga | Terra | | Samaritan| Sumia | Arca | Terra | | Arabic | Semavati | Ardi | Terra | | Modern Arabic | Semavat, Shema, Tel-ek | Arz, Ardib, Auf, Tura, Aalem | Terra | | Moroccan | | | Terra | | Ethiopic | | | Terra | | Greek | Samainat | Mydrai | Terra | | Tygri | Samal, Sämäle | Mydrm, Medre, Medere | Terra | | Armenian | | | Terra | | Hunsan | Szemney | Middrih | Terra | | Maltese | Sema, Smeviet, Smeviet | Art | Terra | | 8. Lycean | (Sea, Tidalmi; And, Athi) | | Terra | | 9. Phrygian | (Bread, Bek; Water, Bedi) | | Terra | | 10. Greek | Ouranos | Ge | Terra |

| Romanic | Heaven, Sky | Earth | Language | |---------|-------------|-------|----------| | 11. Germanic, 360 | Himins, Himins | Himil | Ge | | Alemannish, 720 | Himmel | Himmel | Airtha | | Classical German | Trompfenhaus | Himmel | Erdu | | German | Jewish German | Himmel | Erde | | Low Saxon | Himmel, Himmel | Himmel | Jerda | | Frisian | North Frisland | Hemmel | Hardi | | Dutch | Hezel | Hezel | Eere, Erda | | Danish | Himmel | Himmel | Jerda, Yrteik | | Norwegian | Orkney | Chimrie | Eerde, Wrroll | | Icelandic | Himne | Himne | Aerde, Eertryke, 1270 | | Swedish | Himil, Himirik | Himilom | Jord | | Dalecarlian | Gotlandisch | Hymblum | Jordriki | | Danish Saxon, 880 | Heofena | Heaven | Jord | | English, 1160 | | | Eortha | | 12. Celtic | Irish | Neamb, Nau | Birth, Yearth | | Gaelic | Neamb | Neamb | Talu | | Monks | Nian | Nian | Talanch, Dtalmhuin | | Walden | Neamb | Neamb | Talushin | | Cimbric | | | | | Welsh | | | | | Cornish | | | | | Brittonish | | | | | (Brand, Puni, Urta) | | | | | 13. Etruscan | | | | | 14. Latin | Italian | Cielo | Terra | | Piedmontese | | Siel | Terra | | Waldensian P. | | Cel | Terra | | Genoese | | Ze | Terra | | Occitane | | Ciel | Terra | | Venetian | | Zielo | Terra | | Friulian | | Cil, Cilli | Tiarra | | Valais | | Cil | Terra | | Bolognese | | Cil | Terra | | Sicilian | | Cil | Terra | | Old Milan | | Cil, Chelo, Quela | Terra | | Spanish | | Cielo | Terra | | Castilian | | Cil | Terra | | Gallican | | Cei | Terra | | Portuguese | | Cei | Terra | | Romanish | | Ciel, Tehiel | Terra | | Provençal | | Cil | Terra | | French | | Ciel | Terra | | Bearinish | | Cion | Terra | | Rouergue | | Col | Terra | | Flemers | | Ciel | Terra | | Walloon | | Cir | Terra | | Wallachian | | | Terra | | Dacian | | Taheri, Cselburg | Pâmentiv | | Cuzco-Wallachian | | Cercio | Pîmchita | | 15. Cantabrian | | Saera | Lurre | | 16. Slavonic | | | | | Slavonian | Russian Church | Nebesi | Semli | | Common Russian | Nebö | Semla | Semli | | Malo-Russian | Nebo | Zemla | Zemla | | Sandzian | | Nebo | Zemla | | Servian | | Nebeso | Semli | | Ukok | | Nebbu | Semli | | Rajasan | | Nebbu | Semli | | Transylvanian | | Nibe (Brand, Liab) | Semli | | Sol | | Nebi, Nebiesi | Semli | | Croatian | | Nebi | Semli | | South Wendish | | Nebi | Semli | | Hungary-Wendish | | Nebi | Semi | | Polish | | Nebie | Semi | | Kazubish | | Nebo, Wnebi | Semi | | Bohemian | | Nebi | Semi | | Serbian, Upper | | Nebiu | Semi | | Lusatia | | | Semi | | Serbian, Lower | | Nebu | Semi | | Lusatia | | | Semi | | Polishish, 1691 | Nibis, Nebul | Semi | Semi | ### III. TATARIC CLASS.

#### (1.) SPORADIC ORDER.

| Language | Heaven, Sky | Earth | |------------------|-------------|-------| | Lithuanian | Delbes | Semiš, Worsizny | | Old Prussian | Dangon | Zemes | | Pruss.-Lithuanian| Debuissa | Ziemas | | Polish-Lithuanian| Danguje | Ziemas | | Russian | Danguse | Ziemas | | Livonian | Dangus | Ziemas | | Livonian Proper | Debbae | Zemmo, Zemmo |

#### (2.) CAUCASIAN ORDER.

| Language | Heaven, Sky | Earth | |------------------|-------------|-------| | Armenian | Hleškma, Girskin | Eršekir, Gerkril | | Georgian | Tzin, Zata | Kir, Tap | | Ingrian | Tshabash | Sze, Miza | | Kabardian | (Bread, Tshkom) | Kwekara | | Karachay | Tshah | Dicha | | Khasi | Makha | Gim | | Kabardinian | Agagahan, Ashan | Jobste | | Kabardinian | (Bread, Mak) | Astula, Tshillah | | Kabardinian | (Bread, Mikel) | Tola | | Kabardinian | Wurda | Truli | | Kabardinian | Phamah | Tshich | | Kabardinian | Arv, Arwi | Tahigit, Segh, Still | | Kabardinian | Arf | Gukh | | Kabardinian | Sigellich | Late, Mezha, Ghumm |

#### (3.) TARTARIC ORDER.

| Language | Heaven, Sky | Earth | |------------------|-------------|-------| | Turkish | Gug, Kickler | Jer, Gyr | | Turkish | (Chicol) | Kher, Der | | Turkish | Telek, Asman | Kuk | | Turkish | (Zamán, Chak) | Jurd | | Turkish | Gug, Ghiogh, Chok | Gier | | Turkish | Kok | Kok | | Turkish | Ruk, Heda | Er, Toprak | | Turkish | Ruk, Kek, Kik | Jer | | Turkish | Ruk, Tengeri, Samoh | Jer | | Turkish | Ruk, Tengeri | Jer | | Turkish | Ruk | Jer | | Turkish | Ruk, Ava | Jer, Dzhir | | Turkish | Arva, Asman | Jer, Dzhir | | Turkish | Kuk, Asman | Jer | | Turkish | Ruk, Asman | Jer | | Turkish | Oklok, Gioch | Jer | | Turkish | Tengri | Tobruk, Dzhir | | Turkish | Tengeri | Jer, Tser, Toprak | | Turkish | Asman, Hava | Jer, Toprak | | Turkish | Asman | Clz, Jer, Toprak |

#### (4.) SIBERIAN ORDER.

| Language | Heaven, Sky | Earth | |------------------|-------------|-------| | Olanūn, Kumar | Ma, Mu | | Jen-esh, Nebus | Mu | | Wilma | Mu | | Kuldjenja-muu | Mu | | Imun, Immun | Mu | | Eterdaman | Maanku | | Tarom, Nair | Ma | | Numma | Ma | | Tui | Ma | | Soom | Mag | | Nopkon | Jogot | | Nomen, Numtorem | Mig, Mü | | Saika | Müg | | Nusunde | Müg | | Ninnak | Müc | | Num torem | Müc | | Tegem, Jon | Müc | | Lom | Tavesh | | Kišušušte, Kišaha | Ijuklin, Melenteesta | | junja, Pil | Rok, Milande | | jona, Tuhja | Jens, Zamislek | | Manel | Mastor, Moda | | Manen | Mastor, Moda | | Shkai | Mastor, Moda |

#### (5.) TURCO-TAR-TARIAN ORDER.

| Language | Heaven, Sky | Earth | |------------------|-------------|-------| | Turkish | Turuncanalb | Na, Telga | | Turkish | Tomskic S. | Nom fende, Lom | | Turkish | Narymyic S. | Tit | | Turkish | Ketiah | Tita | | Turkish | Timakio | Tit | | Turkish | Carusgase | Tit | | Turkish | (Dakhshin) | Tit | | Turkish | Taigish | Tit | | Turkish | Kolbalie | Tit | | Turkish | Motoric | Tit | | Turkish | (Jenešiostia) | Aršič | | Turkish | (Jenešiostia) | Kotovic | | Turkish | Assanic | Oesh | | Turkish | Inbatshlie | Es | | Turkish | Lumpokolic J. | Es | | Turkish | (Jekadshin) | Dzhenga, Zjugo | | Turkish | Khabulka | Kh'igan, Cherwel | | Turkish | (Jenešiostia) | Chain | | Turkish | (Jenešiostia) | Kuka | | Turkish | (Jenešiostia) | Shilchen | | Turkish | (Jenešiostia) | Keh'quin, Cherwel | | Turkish | (Jenešiostia) | Chiternik, Killak | | Turkish | (Jenešiostia) | Ging, Kellak |

---

**Note:** The table provides a comprehensive list of languages classified into different orders within the Tataric class, detailing their specific terms for "Heaven/Sky" and "Earth." Each language is listed under its respective category with its unique terms for each category. ### Language

#### (Greenland and Eskinian)

41. **Kamtchatkan** - Kochan, Hai - Keis - Kochal, Kollas - Kogal

#### (5.) Insular Order

42. **Curillean** - Nis - Inak

43. **Eastern Islands**

44. **Japanese** - Ten - Dianni - Vallum, Tounnoun

45. **Liu Cheu**

46. **Formosan**

47. **Philippine** - Moluccan - Magindano - Tagalog - Bissayish - Bugis - Manggarai - Pelew - Mariana - Friendly Islands - Coco Islands - Savu - Pampang - New Guinea - New Britain - Blina - Seramboe - Van Diemen's Land - New Caledonian - New Zealand - Otago - Marquesas - Sandwich Islands - Easter Islands

#### IV. African Class.

**Egyptian** - Coptic, Memphite - Sahidic, Thebaic - Baramic - Coptic - Barabirish - Kenya, Burckhardt - Nouba, B. - Ishareen, B. - Adareb, Salt - Arguesa - Massowari - Akkerlo, Salt - Saurin - Shino, Salt - Takie, Salt - Barra, Salt - Mutshana, Salt - Briqua, Salt - Shangalla, Salt

**Heaven, Sky** - Gora, Sky (O. Woka; English) - P. Beja - Qesegh, Sky (O. Hogg) - Wah; (P. Terah) - (O., E-200-ah; ) - (O., D'yoora; J., Moore) - "A catch or click"

**Earth** - Dermitchequa - Tachess - Makooa, Salt - Monjou, Salt - Sowalli, Salt - Somaull, Salt - Hurub, Salt - Galla - Adaib, Salt - Dananii, Salt - Dungolish - Bomhou, Burckhardt - Bohno, Burckhardt - Darfur - (Amharic) Salt - (Tigre) Salt - Aoow, Salt - Tibern A. - Damot A. - Gafat - Falasha - Soudan - Berima - Fulah - Phillata - Yalofs - Berber - Canary - Tibbo - Shillah - Suleh - Serrawallis - Mandingo - Yallonka - Sokro - Felups - Timmaney, Winterb. - Bullam - Sueu - Kanga - Mangree - Gien - Quoja - Fante - Akirpon - Amina - Aklim - Akra - Tamai - Whydah - Papna - Watje

**Language** - English - Hogga - Ira-poo - Moore - Chemo-je; 3, - Ma-da-boo - Char-rah; J., - Tal-ya; I, Kow; - S, Sudu - (O., Eer; J., Di-che - Werke; I, Ahad) - (O., Afro; J., Alaa) - Am-be-re, Sky - Szamma (O., Ayero; J., Alaa, Berna) - Perig - Sema (Day, Dealka) - Szamma (O., Dule; J., Doal, Salt) - (O., Tsai; J., Tucker) - (O., Tsai; J., Werke) - (O., Quo-rah; J., Ziv-va - (King, Negumani; - Song, Mossagam) - (King, Negus; Song, - Mazona) - (King, Negus; Song, - Andje) - (O., Koma; Song, - Baro) - (I, Oku; 3, Oku) - (I, Kilde; 3, Metta) - Hyalla - Szemma - Assama - Giana, Tiget, Igna Doomit, Akai - Ataman - (King, Monsey) - (I, Trono; 3, Agueso) - (O., Afreet) - (O., Ifust) - Rona - (I, Bani; 3, Sicco) - Santo - Margetangala (I, Kidding) - Bandee (I, Kulle) - (I, Esory; 3, Simos) - (I, Pin; 3, Pinas) - (I, Bui, Nimbul; Upock Leh - (I, Rah, Nura, - Wind.) - Aralani - Aralani - Neza (I, Aniandu) - Tata (Head, Tel) - Lam (I, Do) - (King, Dondag - (Head, Hundu) - Niame - Adaukam (I, Ehoo) - Jamkombum (I, Ak- - kum) - Jahime (Head, Me- - tih) - Ngol, Jamkombum I - Glom (Father, - Tabiah; Head, II) - (I, De; 3, Otton) - Jiwel (I, Depoo) - (I, De; 3, Etong) The Hottentots have three particular clicking sounds, made by withdrawing the tongue from the teeth, the fore part, and the back part of the palate; they are respectively denoted by T', T'', and T'''. The first two appear to resemble the sounds sometimes used to express a trifling vexation, and to make a horse go on, or to call to poultry.

V. AMERICAN CLASS.

(1.) SOUTH AMERICAN.

A. Southern Extremity

1. Terra del Fuego (A. Penguin, Com-pogre)

2. Patagonia, Chili

Mokhams, Aran-

kan

Tehuahet

Puelche

B. East from R. Plata

to Marathon

3. Charrua

4. Yaro

5. Bohane

6. Chana

7. Minumae

8. Guenea

9. Pauqua

10. Guarany

Ibog, (O. Currari) Ibi

South West | Language | Heaven, Sky | Earth | Language | |----------|-------------|-------|----------| | 65. Carib Yoi | Onbecou Capou | Menha Soye | 10. Ecclemach D. About Nootka | | Islands | (Men Women) | Nonum Monha | 11. Nootka Sound | | L. Mountains in the N.W. | (O Sun; Moon, Muysea) | | | | 66. Mayaca | | | 12. Atnah | | 67. Kinimaze | | | 13. Friendly Village | | 68. Pepaya | (O Nie; Cepego; 3, Pasquah) | | (Water, Ulkan; Fire, Scapacny) | | 69. Darlen | | | 14. Queen Charlotte's Island | | | | | (Fire, Tesh; 1, Souchou; 3, Slöönia) | | (2.) MIDDLE AMERICAN. | | | 15. Colushan | | A. Islands | | | Ki, Kau, Kiwa, Kitani, Kiggon, Chaaz | | 1. St Domingo | (Field, Conuco; Meadow, Savans; House, Boa; Bread, Casab) | | (Stone, Te; Mex. Teel) | | B. Darlen to Guatemala | | | Kosa (Throat, Katkatl; Mex. Cocotl; "Bot, Coatl; Mex. Coxita") | | 2. Kiche, Utlateco | | | 17. Tehlakitaney | | 3. Pocochic | Taxah (O, Quilh; Acal, Vlen Head, Na; Head, Cam, Bird, Teel) | Nuualhui, Nufai Nitiet | (O, Krane; 1, Clerg. Kalke; 3, Notakh, Nez) | | 4. Yucatan, Maya, Camaó | (O, Kin; Head, Cab) | Luam | 18. Kinalzi | | C. Table Land of Mexico | | | Jijuan, Juon, Jugan Altzen, Aleleh An | | 5. Mixtecan | Andihul, Andi | Tialli | E. West of Mississippi | | 6. Totonacan | Tiayan, Acapoon, Acapalan (1, Tom; 3, Toto) | | (1, Tokes-cum; 3, Nobokescum) | | 7. Mexican, Aztecan | Ikuacat (O, Tomation; 1, Ce; 3, Yel) | | 19. Blackfooted Indian Blood Indian, Pegon | | 8. Huastecan | Tisheh (O, Aquicha; Trabal Head, Na) | Mahédel | 20. Tall Indian | | 9. Othomi | | | (1, Karel; 3, Narce; 4, Nean) | | 10. Mechoacan | | | 21. Saseco | | 11. Pirindan | Pininte | Chimohbl, Hoy | (1, Ut-te-gar; 3, Taukey; 4, Tobo) | | 12. Tarascan | Avandaro (1, Ma; 3, Tanimo) | | 22. Snake Indian | | D. California to Rio del Norte | | | Uchta tibi (O, Poeta; 3, Oweeh) | | 13. Coran | Tahapoa | Chéhtli | (Dog, Shong; 4, Tops) | | 14. Tepehuanas, Topia | | | 23. Nadowessian | | 15. Tubar | Tegmecarichui | Nunguatae | Asinopetuc Sioux | | 16. Tamhumaran | Guami (Bad, Tseti; Gob, Jog, Cocotshi) | | 24. Saki, Ottogami Menomene | | 17. Zuququan, Yaqui | Tercepo | Bayapo | 25. Osage | | 18. Pima | Titauacatum (I, Ani; 1, Mato; 3, Waik) | Inatubarch | (Wind, Tattaunggy; Elekton, Tinal-tauna II) | | 19. Esdeeve | Tevictse | Yuhpetzs | Wimshag, Maha Minocori, Oto Arkanso, Kunze | | 20. Opata | Tequina | Terepa | 26. Pani | | (3.) NORTH AMERICAN. | | | 27. Cadde, Natahi-totache | | A. N.W. of New Mexico | | | 28. Adaire, Attahapa | | 1. Jetan, Apache | Tekeridcademba (O, Datamba Ibo, Ibunga; 3, Gomma, Ganehma) | | F. West of Mississippi, to Ohio | | 2. Keres, Moqui | | | (Agreeable, Hitanchili, Pronto, Jasilia) | | B. About California | | | (My, Na; Elder Brother, Niva; 1, Micocotamano; 3, Nahapumime) | | 3. Pericu | | | 31. Natches | | 4. Walcurio | | | 32. Muskohge, Creek | | 5. Cochimi, Laymon | Tekeridcademba (O, Datamba Ibo, Ibunga; 3, Gomma, Ganehma) | Keammcel, Amet, Ametetenang | 33. Chikkaaw Choktaw | | C. N. of California | | | 34. Cherokee | | 6. St Barbara | (Head, Necchu; 1, Paca; 3, Mapa) | | | | 7. Estene | Imita (O, Tomanis ashli; 1, Pek; 3, Julep) | | 35. Woccon | | 8. Ransien | Terray (O, Orpetnel istmen; 1, Enjala; 3, Rappes) | | 36. Katahba | | 9. Achaelien | (1, Mookola; 3, Capes) | | 37. Six Nations | | | | | | | | | | Mohawk Seneca | | | | | (Fire, Occheleb) |

---

**Note:** The table appears to be a list of indigenous languages with their corresponding names in different regions of North America. It includes various tribes and their languages, such as Carib, Mixtecan, and Osage, among others. have been principally, if not exclusively traced; thus the Egyptian Language had \( \odot \) for the sun and moon, and \( \oplus \) for a country or field, and the Chinese have still \( \Box \), \( \checkmark \), \( \Theta \) for these objects respectively, the characters having been made square instead of round, which some of them were in their more ancient forms. The Egyptians represented a man by a figure kneeling, and stretching out his hand, or, in the hieroglyphic character, thus \( P \). The Chinese figure may originally have been of the same form; but at present is more like a pair of legs only, \( \Lambda \), whilst a dog seems to have three or four legs, \( \Phi \) or \( \Omega \). A thousand, according to Mr Jomard's ingenious conjecture, was copied from the lotus, with its seed vessel, having a great multitude of seeds, and the Chinese \( \Omega \) is certainly not altogether unlike the Egyptian \( \Phi \); nor is the character for light \( \Phi \), which seems intended to represent a radiant body, altogether different from the \( \Phi \) or \( \Omega \) so often found among the hieroglyphics of Egypt, although it is not easy to believe, with Mr Palin, that the manuscripts found with the mummies agree precisely with a Chinese version of the Psalms of David, character for character. The successive introduction of figurative expressions and characters may easily be imagined; but it would be useless to enter at present into further details of this kind, on grounds almost entirely speculative. The Chinese are said to have been, in the ninth century, a race of people resembling the Arabs; their physiognomy was contaminated in the thirteenth and fourteenth by a mixture with their conquerors, the Mongols; but their language remained unaltered. The dialect of Tonkin is sometimes called the language of Annam, and the Guara; on occasions of state they use the Chinese character, but more commonly a character of their own, probably resembling that of the Siamese. Dr Leyden observes, that at least twenty different nations employ the Chinese characters, though they read them quite differently; and he considers the Cochin-Chinese, the Cantonese, and the Japanese, as all essentially different from the Mandarin Chinese, though they have all some words in common. He gives us as the names of the dialects of Chinese, constituting almost as many separate languages.—1. Kong, spoken at Canton; 2. Way; 3. Naka; 4. Chew; 5. See; 6. Lai; 7. Linan; 8. Khams, or Mandarin; 9. Sin; 10. Kuan; 11. Hyong san, spoken at Macao; 12. San takh; 13. Nam kai; 14. Pin ngi; 15. Tong khin; 16. Fo khun, or Chin-chow. There is also a language spoken by the Quen to, between Tonkin and China, a people who consider themselves as more ancient than their neighbours. Notwithstanding, however, all this supposed diversity, we may trace a considerable resemblance in the spoken language, even as far as Korea. In all these dialects, the conversation is a sort of recitative, and the different notes give distinct meanings to the words; as, in fact, we distinguish in English the senses of \( M \) from \( M! \) or simply \( M \); tones perfectly understood, though never written. The Chinese are without the sound of the letter \( r \), and several other sounds common in Europe; they only express in which they express foreign words by putting together the characters of the nearest likeness, with a symbol of pronunciation annexed to them; thus, for Christian, and Cardinalis, they are obliged to write \( Ki \) (for \( eu \) tu \( su \)), and \( Kia \) \( u \) \( f \) \( a \) \( n \) \( i \) \( s \) \( u \), with a mouth annexed to them. The names of places are generally distinguished by a square including the characters which express them; and the names of men, in some books, by a line drawn on one side of the characters only. In this there seems to be a distant analogy to the ring which incloses proper names in the Egyptian inscriptions, but the names of places were not distinguished in this manner by the Egyptians. The dialects of Cambodia and Laos have received some mixture of Malayan from their neighbours; in writing the former of these, sometimes called \( K \) \( h \) \( a \) \( h \) \( a \) \( n \) \( e \), according to Dr Leyden, the Bali, or old Sanscrit character, is employed; and the latter has some analogy with the Siamese; indeed, both the Siamese and the

The tables will at least serve, notwithstanding some imperfections and uncertainties, as a convenient synopsis for facilitating the reference to a brief sketch of the history of the different families of languages.

1. The strongest proof of the great antiquity of the Chinese language appears to be the extreme simplicity of its structure, and the want of those abbreviations and conventional implications which have been sometimes called the wings of languages. It is natural that, in attempting to express ideas at once by characters, the rude pictures of material objects should first Avaneso are disposed to derive themselves from Laos. It may be seen, from the specimens exhibited in the article Philology, that at least some of the Chinese dialects have sounds agreeing in several instances with European words of the same import; but the agreement is scarcely precise enough to justify our inferring from it an original connection between the languages.

2. The language of Siam resembles the Chinese in its simplicity and metaphorical structure, though not so decidedly monosyllabic. It is obvious, however, that the distinction of monosyllabic and polysyllabic could not, in very ancient times, have been so positively laid down as at present, since it was usual, in almost all countries, to write the words contiguous to each other in a continued series, without any divisions between them; and, even in modern printing, there is a happy invention, which often restores this agreeable obscurity, under the name of a hyphen, by the use of which we avoid the difficulty of determining whether we wish to employ one word or several.

The Siamese call themselves T'houy; but a part of their country is distinguished by the appellation Tai hai, or Great T'houy.

The Siamese resemble the Mandarin Chinese; several words of the language are borrowed from the Bali; it is written in an alphabetical character, which is said to be complicated and refined.

3. The Avaneso, or Burmanitish, has also borrowed some polysyllabic words from the Bali, and is written in a peculiar alphabetical character. It must be considered as an era in the history of this country, that its emperor has employed Mr Felix Carey, at his own expense, to establish a printing press at Ava, his metropolis, for printing a translation of the Scriptures in Burmanitish. A dialect, spoken in the district called Tanegnagri, is said to be of greater antiquity. The Môn or Poguon is called by Dr Leyden a distinct original language; but it is written in the Aranese character; and Adelung's specimen scarcely differs at all from the Burmanitish. The language of Arakan and Rakheng is called Rakheng; it contains a number of words from the Bali, many of them converted into monosyllables by an imperfect pronunciation. Dr Leyden considers it as the connecting link between the monosyllabic and the polysyllabic languages; and he calls it an original language, notwithstanding its acknowledged derivation from its neighbours. It employs the Devanagari alphabet, including the letter n. Out of fifty words of Rakheng, quoted by Buchanan, the seven which are not Burman are only varieties of pronunciation. The Kipun or Koïn, and the Kukis, N.E. of Chaitong, are mentioned as neighbouring tribes, speaking languages almost entirely different from the Rakheng. We find, in Mr Buchanan's paper, some specimens of the languages of the Burma empire, which it is difficult to distribute methodically without a further knowledge of their peculiar characters, but some of which may, without impropriety, be introduced here.

4. The language of Thibet, or the Tanguish, has some words in common with the Chinese, but is less simple in its structure. It is at least as ancient as the religion of the country, which is nearly coeval with Christianity. Its character is well known to be alphabetical, from the title of the learned Language-work of Father Georgi on the subject.

(The Anamite of Tonquin and Cochin-China is a separate language, with Chinese affinities.)

The Môn of Pegu is (what Dr Leyden makes it) a separate language from the Burmese.

The Rakheng is Burmese.

Nothing has to be subtracted from this class. The additions, however, are numerous.

1. Numerous forms of the speech of the ruder branches of the Burmese—viz., Karans, tribes of the Yoma Mountains, tribes of the interior of Aracan.

2. The same in respect to the Siamese—Khamti, Laos, &c.

3. Forms of speech, with Burmese affinities—more or less close—from Syriek, Tippera, Chittagong, Manipur, and the southern frontier of Assam—Naga, and other dialects.

4. Forms of speech from the northern and eastern frontiers of Assam—more or less Burmese, Siamese, and Tibetan—Jili, Singpho, Mishmi, Abor, Aka, Dofa, &c.

5. Sub-Himalayan forms of speech—Garo, Kooch, Bodo, Dhimal, &c.

6. Sikkim and Nepal forms of speech—Lepcha, Limbu, Murmi, Newar, &c.

7. The Chepang of Kunzum.

8. The Andaman.

9. The language of the Nicobar Islands.

10. The languages (Sifan) of the ruder parts of China.—[O.O.]

The Indo-European languages have been referred to a single class, because every one of them has too great a number of coincidences with some of the others to be considered as merely accidental, and many of them in terms relating to objects of such a nature that they must necessarily have been, in both of the languages compared, rather original than adoptive.

The Sanscrit, which is confessedly the parent language of India, may easily be shown to be intimately connected with the Greek, the Latin, and the German, although it is a great exaggeration to assert anything like its complete identity with either of these languages. Thus, we find, within the compass of the Lord's Prayer only, Pido, Pitir, among the Sanscrit terms for Father; Gr. Pater, Nama, or Namadhyana, for Name; Gr. Onoma, Onomati, Radhsham, Kingdom; Lat. Regnum from Rego, Monasum, Will, like the Gr. Meno, and the Latin Mens, Terra, Earth; Gr. Era, whence perhaps the Latin Terra; and Dasum, or Devanagri Das, Day; Lat. Dies. There are also some singular resemblances of declension and conjugation between the Sanscrit and the Greek, Dodanit, Dodanit, Dodatit; in old Greek, Didouit, Didouit, Didouit. In a tablet of the date 23 B.C. we find Kritico for a Judge, Gr. Krites, Criticus. In Mr Townshend's work we also find some well-selected instances of resemblance between the Sanscrit and other languages; thus, Bhar, is Brow; Pota, a Boat; Bath, a Bath, Germ. Bad; Dhara, Terra; Nama, Nama, Night; Night; Pad, Poet, Poeta, Poetica, or Proetica. First, whence we derive both the Greek protos and the Latin primum, and Upadesa, Didasana, Docos, and Disco. We have also Vayana, Wind; in Russian, Vyevane; and Vyevane, Whirl; Latin, Vides; German, Witter; Russian, Edova. The number of the plural, which is found in the Sanscrit Bhavanati, They are; Dodanit, They give. Sir William Jones and many others have attributed to some of the works which are still extant in Sanscrit an antiquity of 4000 or 3000 years; but Professor Adelung denies the validity of any of the arguments which have been advanced in favour of a date at all approaching to this.

The Sanscrit, even in its earliest state, can scarcely have been altogether uniform throughout all the countries in which it was spoken, and it has degenerated by degrees into a great diversity of modern dialects; the term signifies learned or polished. Beyond the Ganges, it is called Bali or Magudha, which, the missionaries say, "scarcely" differs from Sanscrit; the term Magudha is said to mean mixed or irregular. In Siam the Sanscrit is still the language of elegant literature; and it is often employed throughout India, with some little difference of construction, under the name of Devanagari, the divine language.

The Prakrit is rather a vague term, meaning, according to Mr Colebrooke, common or vulgar; but it is also applied to the language of the sacred books of the "Jainas." We find in a little publication, entitled a Brief View of the Baptist Missions Language and Translations; some useful information respecting the Indian languages and dialects, into a great number of which these laborious and disinterested persons have made or procured translations of the whole of the Scriptures, which they have printed at Serampore, near Calcutta. The dialects which they enumerate are principally arranged in a geographical order; and, beginning with those which are nearest towards the middle of India, as the pure Sanscrit and its least modified dialects, we may place next to them the languages of the countries bordering on the monosyllabic nations, toward the N. and E. We have here the dialects of Nepal, Assam, or Ummiya, Tipperah, and Kasai, of which little more is known than that translations into the first two have been already executed. The Bengali is spoken in and about Calcutta. The Hindoo or Hindustanee is spoken about Agra; it is printed in the Devanagari character, the font of which contains more than 800 varieties of letters and their combinations. The Urdu or Oordoo is a sub-dialect of the Hindoo, as well as the Brijbhasha, which is nearer to the Sanscrit than some other dialects. The Jyapura is mentioned as another language belonging to the same neighbourhood. The Hindustanee is spoken in Hindustan Proper, or Lower Hindustan; the missionaries say it is "diametrically different" from the Hindoo. The Moor or Mongol Indostanee seems to belong to this country, being mixed with a good deal of Persian and Arabic, unless it be rather referable to the Hindoo. The dialects of Udaipur, Benares, and Manipur, are also called separate languages. The Gondacee is spoken at Nagpore, in the Mahatta country. Further east is Orissa or Oriya, the language of which is printed in a character requiring 390 different types. The Telug or Warung is spoken at Cuddalore and Madras; the Telugues further west. The Canarese has a peculiar language, besides the Tamul, which is spoken from Palenque, near Madras, to Cape Comorin, and the Malay, which appears to belong to a part of this country. About Cochim in Travancore we have the Maltese; further north, the languages of Melabar, Konnara, and of the Deccan. The dialect of Malabar is of considerable antiquity, being found in two copper tablets, one as the eighth or ninth century. Then comes the Kunkuma, above Bombay. The Mahatta is further inland; the Guzerat on the coast, and beyond the Indus, the Balochee in Beluchistan. North of this we find the Afghana or Pushtee language, which contains many Hebrew words than any of its neighbours; the people are said to have come from the N., about 2000 years ago, and, according to a Persian tradition, to be descended from King Saul; indeed, the language stands somewhat more correctly under the Median family in the Mithridates, but since it forms the connecting link between the two families, it might perhaps be as conveniently arranged among the more numerous species of the Sanscrit; it is written in the Arabic character, with some additional letters for expressing the Sanscrit sounds. The language of Multan, N. of Sindh, has about one-tenth of Persian mixed with it. The Gipsies were certainly expelled from some part of India by the cruelties of Timur Leng, about the year 1400; and there were probably some of the Zingoes in the neighbourhood of Multan, their language having a great number of coincidences with that of Multan, and being still more manifestly a dialect of the Sanscrit, although they have adopted many European, and especially Slavonian words. When they first appeared in Europe, they were supposed to amount to about half a million; at present they are less numerous.

The Mahatta is peculiar to the group of small islands from which it is named; the Baptists have already printed some books in it. The people are said greatly to resemble those of Ceylon. The Cingalese, which is spoken in great parts of Ceylon, is a mixture of several of the continental dialects, and it has been observed, that the proper names in Ceylon mentioned by Ptolemy are of Sanscrit origin.

(The Sanscrit can scarcely be called "confessedly" the parent language of India. That the languages of the Dekhan are Sanscrit, only in the sense that English is Latin (i.e., in respect to some incorporated elements), is generally admitted. That the northern forms of speech are Sanscrit is beginning to be doubted. That their vocabulary is largely Sanscrit is certain. Not so, however, their grammar. The forms of speech that, at one and the same time, are most generally treated as separate languages, and, at the same time, most especially believed to be of Sanscrit origin, are—

1. The Hindoo, i.e., the Brijbhasha of the text. 2. The Gujarati of Gujarat. 3. The Udhiya of Orissa. 4. The Bengali. 5. The Mahatta.

The Multan, Punjab, Seinde, Rajasthan, Cashmerean, and other less important forms of speech, belong to this division, and chiefly approach the Hindoo section of it.

The Hindoo is a Lingua-Franca, rather than a true native form of speech.

On the other hand, the Telugua, Telunga, Canarese (or Carnatic), Malayalam, and Kunkuma (of the Concan), are all Tamil; as are all the Todah, Ghoad, Khound, and Kol dialects; as well as that of the Rajmahali mountaineers, on the very verge of the Ganges, and within 100 miles of the southern boundary of the sub-Himalayan monosyllabic tongues.

The basis of the Mahatta and Cingalese is Tamil; the place of many of the Bhil dialects being, if not in the Tamil class, at least intermediate.

In Beluchistan the Biluch is not the only language. The Brahui of the mountains is spoken concurrently with it, and that is Tamil, or South Indian.

Of Noqul, Assam, Tipperah, and the Kasia country, the native dialects are all monosyllabic. It is only the intrusive forms of speech that are Indian.

The Afghan and Bilach are Persian, rather than either Tamul or Northern Indian.—m. o. l.]

Dr Leyden gives, as a proof of the antiquity of the Malayans, that the Temata of Ptolemy is derived from Tema, tin. The connection of this language with the Sanscrit has not been very universally admitted; and some of those who have studied it most are disposed to consider it as wholly original; but, in the purest part of the language, Dr Leyden confesses that there is a considerable resemblance to the Aramee and the Siamese; the words derived from the Sanscrit he considers as somewhat less numerous, amounting, however, to about 5000; they are generally less like the Bali than the Sanscrit, and a still smaller number are borrowed from the Arabic. The character of the monosyllabic languages is in some measure retained. Sir William Jones considered the Malayan as a derivative of the Sanscrit; Mr Marsden supposes it to have received its Sanscrit words through Gujarat; Dr Leyden rather from Kalinga or Telunga; and it exhibits some traces of the dialects of Tasul and Maleian. Besides these various sources, it is said to have borrowed some of its simplest words from the Javaans and the Bugis; and it has become more nearly monosyllabic by suppressing the first syllables of some of the words which it has adopted. The Javanese is said to be more ancient than the Malayans; the empire of Java was formerly powerful and flourishing; the ancient language was much like the Sanscrit, more so than the Malayans, but was written in a peculiar character. Dialects of this language are still spoken in Bali and in Madura. Leyden thinks the Malayans were derived from Java; Marsden rather from Sumatra, though he allows that there are some reasons for conjecturing that an old Sanscrit colony may have settled many hundreds of years ago in Java, and mixed its language with a supposed mother tongue of that Asiatic race.

Of the Sumatraan dialects, the principal, according to Dr Leyden, is the Batta, spoken by a people who occupy the centre of the island, and who still, like some other Indian nations, retain the custom of eating their old relations. The language seems to be partly original, and partly connected with the Malayans; and other dialects of the neighbouring islands. The Rejang is chiefly a mixture of Batna and Malayans; in the Lampuha or Lampung there is also some Javanese. The Achi has admitted a still further influx of words belonging to all the Musulman jargons of the neighbourhood, especially to that of the Mapulas of Malabar. There are other dialects of less note in Neas and the Poggy Islands, most resembling the Batna. This language is provided with a peculiar alphabet, which is remarkable for being written from the bottom of the paper upwards, like the Mexican hieroglyphics; though the Batnas, as well as the Chinese, sometimes hold their books so as to read horizontally. In Borneo there appear to be several dialects, or rather separate languages; two of them, according to Dr Leyden, are the Blajou and the Tisun.

[The Malay is nearest the southern members of the mono- syllable stock. In Malacca and Sumatra it is spoken with little variation, with more in Borneo, more still in Java and Celebes, and the Philippines. It then passes through the intermediate forms of Lord North's Island, the Pelvus, &c., into the Micronesian of the Carolines and Marianaes; thence on to the Polynesian of the Navigators' Isles and all the islands between New Zealand S., and the Sandwich Isles N.; the Fijis W., and Easter Island E. The islands between Java and Timor (the latter included) are Malay. So is, to a great extent, the language of Madagascar.

The numerous mutually unintelligible languages of Australia are connected with some of the ruder forms of speech of Timor, Ombay, &c., being also radically connected with each other. The extent to which they are other than Malay being greatly exaggerated.

In New Guinea, the Louisiade, New Ireland, &c., the languages are Papuan, mutually unintelligible to a great extent, yet related, and touching the Malay tongues in the Arru Isles.

The Tasmanian of Van Diemen's Land is perhaps more akin to the New Caledonian than to the Australian, as if the stream of population had gone round Australia rather than across it.

A Malay form of speech is spoken in Formosa. The languages of all the wilder tribes of the Peninsula of Malacca and the larger isles is, even when they are more or less black, Malay.—n. o. l.

The Andamanese is nearest the languages of the opposite coast, i.e., the Burmese.—n. o. l.

Besides the numerous translations into languages of the Indo-European class, the Baptist missionaries have also printed some Armenian and Persian works at the indefatigable press of Serampore, which is supplied by a letter-foundry and a paper-mill, belonging to the same establishment, enabling them to execute the whole business at less than half the expense of European books of the same magnitude. The little pamphlet already quoted contains also specimens of the characters of the Sanscrit, Assam, Bengalee, Mahratta, Sikh, and Cashmirian, which somewhat resemble each other in the square form of their characters; as well as of the Barman, Orissa, Telenga, and Cingalese, which have a more rounded and flourished appearance; of the Tamil, which looks a little like Armenian; of the Afghan and the Persian used in India; and of the Chinese, both as printed from blocks, and from the moveable metal types which have been cast at Serampore.

6. The connection of the Median family with the Sanscrit on one side, and with the Greek and German on the other, is sufficiently proved by the words Abetop, Zend, Sun; Sansc. Ablutea, Dar, Ter, Pers. Door; Sansc. Dure, Tusura; Javanese, Turri; Gr. Thira; Germ. Thir, Thor, Dip, Pers. Land or Island; Sansc. Dilp, Dochtar, Pars. Potrhi; Zend. Daughter; Gr. Thapater; Germ. Tochter; Sansc. Patril, Jar, Zend. Year; Sansc. Jahron; Germ. Jahr; and Isth, Zend. Love; Sansc. Isthia. To this list we may add, from Dr Leyden, Stre, Zend. Woman; Sansc. Stri, Ants, Zend. He is; Sansc. Asti; Gr. Esti, Hopt, Zend. Seven; Sansc. Sep-tah; Gr. Hepta. There are also some coincidences with the Chaldee, but the Median is certainly not a dialect of the Chaldee. Sir W. Jones and others have said that the Zendish was nearest to the Sanscrit, and the Pehlvi to the Chaldee or Arabic. In ancient Media, the Zendish was the language of the northern, and the Pehlvi, or Parthian, of the southern parts; the word Pehlvi, or Pahalevi, is supposed by Leyden to have been nearly synonymous with Pali or Bali, though this is said to be derived from Bahliki, an Indo-European country. The Zendish was more particularly appropriated to religious purposes, and the Pehlvi had in a great measure superseded it for common use at a very early period. Under the Sassanides, again, from the third to the seventh century, the use of the Pehlvi was discouraged, and the old Persian substituted for it. It is said, however, that in the remote parts of the country, about Shirwan, some traces of the Pehlvi may still be found in existence. The Zendavesta of Zoroaster, which is still extant in Zendish, is said to have been written 320 years B.C., and Adelung follows Ampeillet in asserting its authenticity, even in opposition to the opinion of Jones and Richardson. These languages have little or no connection with the Georgian and Armenian, which have succeeded them in some of the same countries. The old Persian, which seems to be much connected with the Pehlvi, has remained in use, either as a living or as a learned language, ever since the time of the Sassanides; it was current among the Persians when they were conquered by the Arabs in the seventh century; and it is the language of the Shah Namah of Ferdusi, written in the tenth century, as well as of the Ayen Akbery, of which the date is about 1600. The modern Persian became a calibrated language about the year 1000, having received a considerable mixture of Arabic and Turkish words. The term Parsi is commonly applied to a corrupt Pehlvi, spoken by the refugee fire-worshippers in Bombay.

The Goths are said to have inhabited, for some centuries, the countries about the Black Sea, and may originally have bordered on Persia. From this circumstance, and probably also from the effects of a later irruption of the Goths into Persia, which is recorded in history, we may easily explain the occurrence of many Persian words in German, and in the other languages of Northern Europe. Professor Adelung has enumerated more than two hundred cases of such resemblances, and has found only one sixth part of them in Ampeillet's vocabularies of the more ancient dialects. He has, however, omitted to state what proportion the whole magnitude of these vocabularies bears to that of a complete dictionary of the language. It is well known that an essay was published a few years since in London, On the Similarity of the Persian and English Languages; and a more elaborate work on the relations of the Persian languages, by M. Lepicur, has since appeared in Holland. M. Lepicur attempts to explain the is or s of the genitive of the northern languages by the Persian preposition xx, which seems to be synonymous with the Greek and Latin xx; but he has not shown that this xx ever follows the noun to which it relates.

The Kurds speak a corrupt dialect of the Persian; they are probably derived from the Cardushi of the Greeks, who inhabited the Gordian Hills. They spread into Persia about the year 1000, and are now situated on the borders of the Persian and Turkish dominions. The language of the Afghans, about Candahar, is said to contain about one-fourth of Persian, and some Tartarian, besides the Sanscrit which abounds in it.

(The term Medea appears to mean the languages allied to either the Zend or the modern Persian; and as few writers (of which, however, the present is one) hesitate to derive the latter from the former, this meaning may stand; subject, however, to future correction.

Zend is so akin to the Sanscrit, that a class of any convenient magnitude must contain both. If so, their descendants should be in the same class also, viz., the modern Persian and the North Indian dialects. The complications and difficulties that arise out of this are considerable.

The Zend is the language of the Zendavesta; and closely allied to it is that of the Cuneiform inscriptions.

The Pehlvi is the language of the Sassanian inscriptions.

It is safer to write thus than to distribute the languages thus named geographically.

There are difficulties in deriving the modern Persian from the Zend, and there are difficulties in not doing so. In the mind of the present writer, the former outweigh the latter. Suppose that, after the Macedonian kingdoms of Bactria was formed, the successors of Alexander had extended their conquests into Tibet; that they filled the country with inscriptions in Macedonian Greek, and that a Greek literature was developed. Time goes on, and the Greek language changes. It takes in some words from the original Tibetan, it gives more, say thirty, forty, fifty per cent. Meanwhile, the original monosyllabic and uninflected character of the native language is preserved. Modern scholars investigate it; they note the number of Greek words; they note, too, the absence of the Greek inflections; they infer that the Tibetan is a language of Greek origin, being once, but now no longer, inflectional; just like the English All this is wrong. The non-inflectional character of the Tibetan arises, not out of the facts of inflections having been lost, but out of the fact of their never having been developed. Nevertheless, a historical series of changes that never existed is inferred.

The real relation of the Greek to the Tibetan is that of the Latin (not the Anglo-Saxon) to the English. The real history of the want of inflections is the opposite of the one assumed. A change from the inflected to the uninflated stage of a single language is simulated. The actual fact is the incorporation of a large (say an inordinately large) percentage of words from a polysyllabic to a monosyllabic tongue.

Write Zend for Greek, and Persian for Tibetan, and the modern language of Isfahan and Teheran appear as descendants from the Zendavesta and arrow-headed forms of speech. Apply the criticism suggested by the erroneous character of the preceding inference, and this affiliation becomes doubtful. Such criticism, however, has yet to be applied.

Mutatis mutandis, the same applies to relations of the Sanscrit to the North Indian dialects.

The Bihuc is Persian.

The Afghan belongs to the same class.

Again, an important group, that by slightly raising the value of this class can be referred to it; or, if this be inconvenient may stand as a closely allied order, is one for which our data are quite recent. It contains the languages of some of the northern tribes of Afghanistan, the Kohistan of Kabul, Kaferistan (Siafoosh), the head waters of the Oxus, Chitral and Hunz-Nagar. The present writer has called this the Paropamisan group.

The present writer, holding that the comparative absence of inflections of the Persian arises not from loss, and from non-development (the same applying to the North Indian forms of speech), stands prepared for placing it, with its congeners, in immediate connection with the monosyllabic languages,—the Zend and Sanscrit (the apparent mother-tongues being widely separated from them, though closely allied to each other.

If this be true, the Indian tongues are those where the basis is Tamil rather than monosyllabic; the Persian those where it is monosyllabic rather than Tamil.

More, however, will be said about this in the sequel; also something about the Goths.—n. o. l.

7. The Arabian family is called by the German critics Semitic, from Shem the son of Noah, as having been principally spoken by his descendants. Though not intimately connected with the European languages, it is well known to have afforded some words to the Greek and Latin; it has also some in common with the Sanscrit, though apparently fewer than either the Greek or the German. Thus we have Acer, Heb. a Husbandman; Agor, Lat. a Field; Asther, a Star; Gr. Aster; Bara, Duri, Germ. Burg; Ben, Heb. Son; Same, Bana, Child; Esh, Heb. Eshta; Chald, Fire; Sanse, Aster; and Ish, Heb. Man; Same, Isha, Man or Lord. The Hebrew Ani, Ameli, I, has been noticed by Townsend and others as affording an etymology for Ego as well as Ni or Mi of verbs, for the Anok of the Egyptians, and even for the Ngoo of the Chinese.

The northern nations of this family have sometimes been comprehended under the name Aramaic, in contradistinction to the middle or Canaanitish, and the southern or Arabian. The Eastern Aramaic, or old Chaldee, is very little known; it was the language of a people situated in the north of Mesopotamia, which is now the south of Armenia; a part of them extended themselves further south, and became Babylonians, of whose dialect some traces are said still to exist about Mosul and Diarbeker. The old Assyrians, between the Tigris and Medea, were a colony of the Babylonians, and spoke a language unintelligible to the Jews (2 Kings xviii.). The Western Aramaic has become known, since the Christian era, as the Syriac, in which there is an ancient and valuable translation of the New Testament. It is still spoken about Edessa and Harran. The Palmyrene was one of its dialects; the modern Assyrian or the Russian vocabularies appears to be another.

The language of the Canaanites is said by St Jerome to have been intermediate between the Hebrew and the Egyptian. The people are supposed to have come originally from the Persian Gulf; the Philistines who were found among them to have emigrated from the Delta to Cyprus, to have been thence expelled by the Phoenicians, to have adopted the language of the Canaanites when they settled among them. The book of Job is considered as affording some idea of the dialect of Edom; it is well known to contain many Arabisms, besides some other peculiarities. The Phoenician is only known from a few coins and inscriptions found chiefly in Cyprus and in Malta, and not yet very satisfactorily deciphered; though Akerblad is convinced, by some of them, that it varied but very little from the Hebrew; of its descendant, the Punic, or Carthaginian, a specimen is preserved in the speech of Hannu in Plautus, as happily arranged by Bochart; the objection of Adelung, respecting the want of a proper name, appearing to have arisen from a mistake. The last six lines of the text are probably either a repetition of the same speech in the old Libyan of the neighbourhood, or a jargon intended to imitate it.

The Hebrews originated among the Chaldeans, Terah, the father of Abraham, having been a native of Ur, or Edessa, beyond the Euphrates. They adopted the language of the Canaanites, among whom they led a nomadic life, till their residence in Egypt, which must probably have had some effect in modifying their language. After that time, however, it appears to have varied but little in a period of 1000 years, from Moses to Malachi; and this circumstance Adelung considers as so uncommon and improbable, that he is disposed to believe that the writings of Moses must have been modernized at least as late as the time of Samuel. The old Hebrew became extinct, as a living language, about 500 n.c.; 1000 years afterwards the Masoretic points were added, to assist in its pronunciation; and this was done in some measure upon the model of the Syro-Chaldaic, which at that time was still spoken. The Septuagint version, which is much older, supports, in the instances of many of the proper names, the reading indicated by the points; but in about as many others it appears to deviate from that system, and to agree with a mode of pronunciation founded upon the text or principal characters alone. The reading in Greek letters of Origen, in his Hexapla, tends, on the whole, very strongly to support the points. The Chaldee has superseded the Hebrew at the time of the captivity, and was generally connected with Syro-Chaldaic, which is called Hebrew in the New Testament. The Targums, and the Talmud of Babylon, are in the older Chaldee; and a Syro-Chaldaic translation of the New Testament has been discovered to be still in existence.

The Samaritan somewhat resembles the Chaldee. It was formed among the Phenicians and others, who occupied the habitations of the ten tribes when they were carried into captivity by Salmanasar and Barhadaden. Its peculiar alphabet is well known as a mere variation of the Hebrew.

The Rabbinical dialect was principally formed in the middle ages, among the Spanish Jews, who were chiefly descended from the inhabitants of Jerusalem; while those of Germany and Poland were generally Galileans, and spoke a ruder dialect of the Hebrew than the fugitives from the metropolis.

The Arabs have been a distinct, and, in a great measure, an independent nation, for more than 3000 years. Some of them were descended from Shem; others, as the Cushites, Canaanites, and Amalekites, from his brother Ham. Their language, as it is found in the Koran, contains some mixture of Indian, Persian, and Abyssinian words. Its grammar was little cultivated until a century or two after the time of Mohammed. It is certainly copious; but its copiousness has been ridiculously exaggerated and absurdly admired. The best Arabic is spoken by the upper classes in Yemen; in Mecca it is more mixed; in Syria corrupt, and still more so in some parts of Africa. There are dialects which require the assistance of an interpreter to make them intelligible; at the same time, it has been maintained by Arya, a learned Arab of Syria, in contradiction to Niebuhr, that the Arabic of the Koran is still employed in conversation among the best educated of the people, as well as in correct writing. The Arabs living in houses are called Moors, and those of Africa are the best known under this name. The Mapuls or Mapalets of Malabar and Coromandel are a numerous colony of Arabs, who have been settled there above a thousand years. The Ethiopians are descended from the Cushite Arabs. In the time of Nimrod they conquered Babylon; before that of Moses they emigrated into Africa, and settled in and about Tigri; in Isaiah's time they seem to have extended to Fez; and at present they occupy Tigri, Amhara, and some neighbouring countries. They became Christians in 325, but retained the initiatory ceremony of the Jews and Mussulmans.

The pure or literary Ethiopic is called Ge'ez, or Axumitic, in contradistinction to the Amharic, by which it was superseded as the language of common life in Amhara about the fourteenth century, although it is still spoken, without much alteration, in some parts of Tigri; while in others, as in Hamasor, a different dialect is spoken. The Ethiopic was first particularly made known in Europe by the elaborate publications of Ludolf. Mr Asselin has since procured a translation of the whole of the Bible into the Amharic, as it is now spoken at Gondar; it was executed by the old Abyssinian traveller who was known to Bruce and to Sir William Jones, and was afterwards put to press at the expense of some of the British societies.

The Maltese is immediately derived from the modern Arabic, without any intervention of the Punic. The island, having been successively subject to the Phoenicians, Phoenicians, Greeks, Carthaginians, Romans, and Goths, was subdued by the Arabians in the ninth century; in the eleventh the Normans conquered it; and it remained united with Sicily until it became in some measure independent, under the Knights of St John.

[There is nothing to be subtracted from this group. In the way of addition, it contains the Gafat of Ethiopia.]

The Semitic tongues graduate into the other African tongues by means of a class for which the name Sub-Semitic has been suggested, containing—

1. The Coptic of Egypt, which has affinities with the Bisharaye, the Nubian, and the Galla.

2. The Amazigh or Berber tongue, leading to the Berou, Hausa, and other languages of Central Africa.

3. The Agyes and other allied tongues of Abyssinia. These are, perhaps, absolutely Semitic.—x. g. l.

8. The Lycean is only known from a few short inscriptions copied by Mr R. Cockerell, and published in Mr Walpole's collection, together with two or three longer ones, which have been lately brought from Antiphollos by the enterprising and indefatigable Mr W. J. Banks. By means of a proper name in one of Mr Cockerell's inscriptions we obtain a part of the alphabet; thus i is a; Δ, δ; η, η; ρ, ρ; Σ, s; and probably Λ, η; and i, l. A further comparison of the different parts of the other inscriptions with the Greek phrases that almost uniformly accompany them, implying "for himself and his wife and children," gives us the words a, or ἐμοῦ, myself; sa, his, or for his; here, or perhaps ἡμῶν, wife; τιδαῖμι, son; τιδαιμονί, children; and athi, child.

It does not appear that any of these words would authorize us to place the Lycean language as a member of the great Indo-European class; but it is reported to have been much mixed with Greek, and, on account of its geographical situation, it may be allowed to occupy a temporary rank between the principal oriental and European languages. If it has a shadow of likeness to any other language, it is perhaps to the Cimbrie; and Tidaiami may also possibly be allied to the Greek Titheos, to nurse.

9. Respecting the ancient Phrygian, we have a few traditions only, which at least agree in giving it a high antiquity, as the source of several Greek words. Thus, Plato observes, in his Cratyless, that the terms denoting fire and water are not derived from any other Greek words, but are Phrygian primitives. It seems, however, that water was called Idaia by the Phrygians, and the word resembles the Bada, Bath, of the northern nations, as well as the Vota, water, of the Swedes. Moirini, the fates, derived from the Phrygian, is compared to Meyer, virgin, of the Gothic; and Bek, bread, is as much like our Bake, as like the Albanian Bok, bread.

10. The German has no very intimate or general connection with any of the older languages which have been preserved entire, although there are a number of particular instances of its resemblance to the Sanscrit, some of which have been already mentioned; it has also many German and Celtic words, some Slavonian, and, as it is said, a few Finnish. It can only Language have been immediately derived from the language of the neighbouring Thracians and Pelasgians, who seem to have come originally from the middle of Asia, through the countries N. of the Black Sea, and to have occupied not only Greece and Thrace, but also the neighbouring parts of Asia Minor, where they probably retained their ancient dialect to a later period than elsewhere. The whole of the Thracian states were greatly deranged by the expedition of the Celts in 278 B.C., which terminated in their settling the colony of Galatia. The Dacians, or Getes, who principally occupied Bulgaria, extended themselves further northwards, and afterwards constituted the Roman provinces of Moesia and Dacia, which were conquered by the Goths in the third century. The Macedonians, in the time of Alexander, spoke a language which was nearly unintelligible to the Greeks in general; even the Pelasgi in Epirus and Thessaly long retained a dialect materially different from their neighbours, and in Arcadia still longer. The Hellenes, who emigrated from Asia Minor into Greece, were not sufficiently numerous to carry their own dialect with them, although the language assumed their name. The Greek in Italy were Pelasgians, although Dionysius of Halicarnassus includes them in the denomination Hellenic; their language must have been Æolo-Doric, and it was in this form that the Latin received its mixture of Greek; the Laconianomian also retained it till a late period, writing, for instance, instead of Patr, Poitr, as in Latin Puer. The Æolic appears once to have extended over Attica, and to have left some Æolians in the old Attic dialect. This dialect was the principal basis of the common language of Greece at a later period, which must have been the most cultivated under the protection of the court of Alexandria, and which continued to be spoken and written in the highest circles of Constantinople throughout the middle ages. By degrees it degenerated into the modern Romance, having received a mixture of Turkish and Italian, and perhaps of some other neighboring languages.

11. The German family is sufficiently connected with a variety of others belonging to the Indo-European class to be admitted into it upon a very short investigation. Its resemblances to the Greek, within the compass of the Lord's Prayer, besides Father and Name, are Wille, Wollen, Gr.; Boule; perhaps Brat or Prout, bread, like Artos; and Freyge or Löch, like Rhiaucia and Luscin. Dr Jamieson has shown very clearly, in his Horace Syllectus, how immediately the structure of the Gothic languages is derived from that of the Greeks. Thus the ear of the Greek infinitive became in the Mosso-Gothic an or tan, in German en. The root of the adjectives, Mosso-Gothic, ans, ans, or xios, as mahtoigs, mighty, Germ. machtig; the Slavonians have ski, the Swedes ska. The inos, Lat. ensis, Anglo-Saxon en. The leos, Latin lis, German lich, English like; thus pelice is what like, at least in Scotland; the Mosso-Gothic sveleks is our such; somelteks is similis. Los, lis, lion, of diminutives, in Latin ius, becomes in Mosso-Gothic ilo, as barilo, a little child; in German maunt is a little man. Among the pronouns we have roo in Greek and Latin, Mosso-Gothic ir, Icelandic eo, Swedish jao; emot, movt, Gr., Latin mihi, Mosso-Gothic meina, German meinen; emoi, moi, Latin mini, Mosso-Gothic mio, Swedish mio, Dutch my; eme, me, Latin me, Mosso-Gothic mie, Anglo-Saxon me. Dutch my, Sc. Doric tu, Latin tu, Mosso-Gothic thu. Is in Latin, Mosso-Gothic is; ieu, Mosso-Gothic is, izos; io, Mosso-Gothic ita, English it. Quis, culum, Mosso-Gothic quais, quis, quae, quiana, the last having the n, as the Greek hon. Uther, whether, alter, other, seem to be derived from another, esthera, meaning one of them, so that in this instance the Gothic has the appearance of the greater antiquity, while the Greek affords, on the other hand, an etymology for ekinos, from ekki, there, which is wanting to the Mosso-Gothic gainis or jains, the Alemannic genesi, the German jenes, and the English yonder or yon. Again, among the numerals, deka has been derived from duo, as if both hands were tied together; and pentu has a strong resemblance to panta, as if all five fingers were reckoned; and, on the other hand, da cuig in Gaelic, meaning twice five, has been considered as the original of deka. But none of these etymologies seems to be so decisive of originality as that of esthera, which is evidently related to turba or turma, while the first syllable remains unexplained in Latin; but in the Language. Celtic we have cad tarf, or cat tarf, a war troop, agreeing undeniably with the sense. For another example, we may take ventus and wind, for which we find no Latin etymology, while the German furnishes us with wehen, to blow, and thence wehend and wehend. The words nodus and knot afford also a similar instance; nodus having nothing nearer to it in Latin than nec, to spin, necte, to unite; but in German we have knüten, to join, and in English knit and knead from the same root. The degrees of comparison are expressed in Greek by ἐκοσμος and ἐκοσμος; in Anglo-Saxon by ἐκοσμος or ἐκοσμος, and ἐκοσμος of art. ἐκοσμος seems to mean before, as well as the Latin on. The Coptic has no comparative; but for better than I, the Egyptians said very good before me. It would seem at first sight natural to make than a preposition, as well as before, and to say better than me; but the fact is, that in English, as well as in German, it was usual of old to say than or dem in this sense; and he is wiser than I meant only, he is wise before, then I follow. The idea of time or place is now dropped as unessential to the kind of priority in question, but the ground of the grammatical construction remains unaltered. In Meso-Gothic the comparative intensification is iro or oro, the superlative ista or ista; thus the Greek μεγαλος becomes μεγαλος, and maistas is obviously maistros. The old megaloς is mikkilz, mickle or muckle; and minor, minumus, became mina-ko, minista; in Persian, mik is great, mithir, greater, mithiras, greatest; better seems to be from the old German bied or bieder, upright, honest, and resembles the Persian minith, better. The Meso-Gothic verbs have also some striking resemblances in their form to the Latin; thus the present tense of to know is hana, harnas, haratith; haram, harathit, harand; habut, habata, habandam; habentem, habandans.

The substantive verb singular in Greek is ειμι, εις, εστι; the plural in Latin sumus, estis, sunt; the Meso-Gothic has im, is, ist, sumim, sumith, sind; and sis is siaros; esse, wesan. The Meso-Gothic nouns frequently retain the resemblance of the Greek more strongly than their more modern derivatives; thus a tooth does not appear to be very immediately to denote or odonta as its source; but the older form tentho is clearly the intermediate stage of this modification; and numberless other instances of the same kind might easily be found.

The Germans were known, as early as the time of Pythias, that is, 320 B.C., as consisting of the Jutes in Denmark, the Teutones on the coast to the east of them, the Ostiones next, and lastly the Cassini, Cotini, or Goths. Professor Adelung imagines that the eastern nations, or Suevi, employed almost from the earliest times a high German dialect, and the western, or Cimbri, a low German; the Suevi he supposes to have been driven, at a remote period, into the south of Germany by the Slavonians; and some of the Goths appear to have extended as far as the Crimea. The Bible of Ulphilas, in the Gothic or Meso-Gothic of 360, is the oldest specimen in existence of the German language. Besides the Greek and Latin, which appear to prevail so much in the language, it exhibits a considerable mixture of Slavonian and Finnish; the translation is far more literal than it could be made in any of the more modern dialects of the German, and sometimes appears to follow the text with somewhat too much servility.

The modern German, founded on the higher dialects of Saxony, was fixed and made general by Martin Luther. There are many shades of dialect and pronunciation in the different parts of this diversified country, but none of them of any particular interest, or established by any literary authority. There are still some German colonies in the territories of Vicenza and Verona, called the Sette Comuni, which retain their language. The German Jews have a peculiar jargon, borrowed in some measure from their brethren in Poland, which they write in Hebrew characters; and another similar mixture of dislocated dialects is spoken by the Rothweil, a vagabond people of the south of Germany, who have sometimes been confounded with the Gipser.

The Low Saxon, or Platt Deutsch, is spoken about Halberstadt, and farther north, in the countries between the Elbe and the Weser; it seems to be intimately connected with the Frisian and Dan sh, as well as with the English. The Frislanders originally extended from the Rhine to the Ems, and the Cauchi, thence to the Elbe; these countries still retain a dialect materially varying from those of their neighbours. Language. The Brokmöke laws of the thirteenth century exhibit some remarkable differences from the German of the same date; thus we find in them Redice, a judge, or Reece, instead of Richter; Keme, kin; and sida, side, as in Swedish, instead of seite. The Batavian Frieslandish approaches very much to the English; it has several sub-dialects, as those of Molker and Hinderby. Some of the Cancelli Frieslanders remain in the territory of Drenthe; the North Frieslanders occupy Heligoland, Husum, and Amrum.

The Dutch language is a mixture of Frieslandish, Low Saxon, and German, with a little French. It appears, from Kolya's Chronicle, to have been distinctly formed as early as 1156.

The Scandinavian branch of the Germanic family is characterized by the want both of gutturals and of aspirates, which renders its pronunciation softer and less harsh; and by some peculiarities of construction, for instance, by the place of the article, which follows its noun, both in Danish and Swedish, instead of preceding it, as in most other languages. The name of Denmark is first found in the ninth century; until the sixth the people were called Jutes. Norway, in the ninth century, was termed Nordmanland. A corrupt Norwegian is still, or was lately, spoken in some of the Orkneys, which were long subject to Norway and Denmark. In the eastern parts of Iceland, the language is much like the Norwegian; but, on the coast, it is mixed with Danish. The oldest specimen of Icelandic is the Jus Ecclesiasticum of 1123. The term Runic relates to the rectilinear characters cut in wood, which were sometimes used by the Scandinavian nations. The Swedes are derived from a mixture of Scandinavians with Goths from Upper Germany, but their language does not exhibit any dialectic differences corresponding to this difference of extraction. Mr Townsend has given us a list, from Peringskiold, of 670 Swedish words resembling the Greek; but it must be confessed that the resemblance is in many cases extremely slight.

The Saxons are mentioned by Ptolemy as a small nation in Holstein, whence, in conjunction with the Frieslanders, and the Angles of South Jutland, they came over to England about the year 450. The Saxons settled principally south of the Thames, the Angles north. At the union of the Heptarchy, the Saxon dialect prevailed, and the English, which nearly resembled the Danish of that time, was less in use; but new swarms of Danes having inundated the north of England in 797, the Danish dialects were introduced by Carute and his followers; and it is about this period that our earliest specimens of the Anglo-Saxon are dated. The Saxon dialect again obtained the ascendancy under Edward the Confessor; and although some French was introduced by this prince, and still more by William the Conqueror, into the higher circles of society, the courts of law, and the schools, yet the use of the French language never became general among the lower classes; and the Saxon recovered much of its currency in the thirteenth century, when the cities and corporate towns rose into importance under Edward the First. In the fourteenth century it was permanently established, with the modifications which it had received from the French; and it may be considered as truly English from this period, or even somewhat earlier, at least if Pope Adrian's rhymes are the genuine production of 1156. It is still much more German than French; in the Lord's Prayer, the only words of Latin origin are trepasse, temptation, and deliver. Professor Adelung's remarks on the simplicity of the English language appear to be so judicious as to deserve transcribing: "The language," he observes, "only received its final cultivation at the time of the Reformation, and of the civil disturbances which followed that event; nor did it acquire its last polish till after the Revolution, when the authors who employed it elevated it to that high degree of excellence, of which, from its great copiousness, and the remarkable simplicity of its construction, it was peculiarly capable. It is the most simple of all the European languages; the terminations of its substantives being only changed in the genitive and in the plural, and the alterations of the roots of the verbs not exceeding six or seven. This simplicity depends, in some measure, on a philosophical accuracy, which is carried systematically through the whole language, so that the adjectives, participles, and article, are indeclinable, being in their nature destitute of any idea of gender, case, or number, and the form of a generic distinction." Language is [almost entirely] confined to objects which are naturally entitled to it. The pronunciation, on the other hand, is extremely intricate; and foreign proper names, in particular, are much mutilated whenever they are adopted by the English.

12. The Celtic family forms a very extensive and very interesting subdivision of the Indo-European class. It has been asserted by some writers, "That the six original European languages, the Iberian, Celtic, Germanic, Thracian, Slavonic, and Finnish, were just as distinct at the beginning of their history as they are now;" but this assertion must be subjected to considerable modification. The thing is in itself so improbable as to require far more evidence than we possess to establish it; even if that evidence were of a more decisive nature; and, in fact, it will actually be found, upon a comparison of the Gothic of Ulphilus with the modern dialects, that the Germanic of that day did approach more nearly, both to the Celtic and to the Thracian or Greek, than any of its more modern descendants do. The change of *tunxu* into *tooth*, for which the Germans have *zahn*, has already been noticed; the *atta* and *himina* of Ulphilus seem to be more like the Irish *Atáis* and *Nóimh* than the modern *Vater* and *Himmel* are; and the *Mroso-Gothic vaia*, which answers to the Cimbrian *vaia*, a man, is not at present found in German, though its traces may still be observed in the *Fritibarno* of the Franks in 1029; the antiquity of the root is shown by the Celtic names in Caesar beginning so often with *vas*, and still more strongly by the testimony of Herodotus, that the Scythian called a man *aros*. At the same time, therefore, that we admit the propriety of considering the Celtic and Germanic as families clearly distinct, with respect to any period with which we are historically acquainted, we must not forget that they exhibit undeniable traces of having been more intimately connected with each other, and with their neighbours, in the earlier stages of their existence. The resemblances of the Celtic to the Latin are too numerous to require particular notice, the immediate and extensive connection between these languages being universally admitted; but if any evidence were desired on this subject, it might be obtained in abundance, by a reference to Cour de Gebelin's *Monde Primitif*. With respect to the Greek, the terms *Hael*, sun; *Dux*, water; *Deru*, oak; *Garas*, crane; *Cruan*, sea, are among the Celtic words of the most indisputable originality; and their resemblance to *Helios*, *Hudor*, *Dras*, *Geronas*, and *Cruonas*, is equally undeniable. We find, also, in the Cimbrian, *Bas*, low, connected with *Bathis*, *Bara*, bend, perhaps with *Bora*, food; *Deprinas*, kingdom, with *Turanum*, *Dygo*, river, with *Dona*; and *Gogordant*, glory, perhaps with *Dona*, exulting. With the German it is easy to find a number of very near approaches to identity, even in that Celtic which can be proved, principally from the etymologies of proper names, to be prior to the date of any known or supposed secondary intercourse or mixture of the natives concerned. Thus we have, either accurately or very nearly in the same signification, *Ap*, *Afe*, or *Ape*; *Berru*, *Berru*; *Blume*, *Blume*; *Belgan*, *Belge*; *Brig*, *Berg*; *Brogl*, *Broil*; *Carra*, *Korra*; *Doqa*, *Teich*; *Gall*, *Klub*; *Gama*, *Kranich*; *Gnabat*, *Knabe*; *Laacca*, *Lamme*; *Mare*, *Mehe*; *Marpa*, *Miroyel*; *Reida*, *Reiten*; *Rit* or *Ret*, *Rod*; and *Ur*, *Uner*; and it is impossible to suppose that so numerous a series of coincidences can have been derived from accidental causes only.

The Celts may be imagined to have emigrated from Asia after the Hesperians or Cantabrians, and before the Thracians or Pelasgians, settling principally in Gaul, and spreading partly into Italy, under the name of Ausonians and Umbrians. In 570 B.C., they undertook expeditions for the purposes of conquest, but they were subdued by the Romans. Their language was current in Gaul till the sixth or seventh century, when it was superseded by the rustic Roman, which by degrees became French; in Ireland and Scotland it has remained with few alterations; in Wales and Brittany it has been more mixed. The Gauls must have peopled Britain at least as early as 500 B.C. The true ancient Britons are the Highlanders of Scotland only, having been driven northwards by the Cimbri; they still call their language Gaelic. The Irish are probably derived from these Highlanders; they were originally termed Scots or Scuts, that is, fugitives, from the circumstance of their expulsion from Britain; so that, where the Scots are mentioned before the tenth century, as by Porphyry in the third, we are to understand the Irish. Gildas, in 561, sometimes calls them Scotch and sometimes Irish. After the retreat of the Romans from Britain, a part of them re-entered Scotland about the year 503, and changed its name from Caledonia to Scotia Minor. In 432 St Patrick laid the foundation of the civilization of Ireland; and, in the seventh century, several Irish priests undertook missions to the continent. At the beginning of this century some Scandinavian freebooters had begun to visit Ireland; and in the year 835 they formed large colonies of emigrants, who established themselves firmly in that country, and in the Scottish islands, bringing with them many Gothic words, which became afterwards mixed with the Celtic, and which seem to constitute about one-fifth part of the modern Irish and Gaelic, 140 Gothic words being founded on the first six letters of the alphabet only. Some of these Normen remained distinct from the Irish till the year 1102. The oldest specimens of the Irish language, admitted by the continental critics to be authentic, are of the ninth century, though some of our antiquaries have imagined they have discovered records of a much earlier date. The Gaelic of the Isle of Man is mixed with Norwegian, English, and Welsh. A Gaelic colony formerly established at Walden, in Essex, has been placed by Chamberlayne in Italy, as a nation of Wal-denses.

The Cimbric or Celtico-Germanic language was remarked by Caesar as differing from the Gallic, although the distinction has not always been sufficiently observed. The Cimbrians seem to have existed as a nation 500 to 400 years B.C.; the Gauls called them Belgae; they invaded Britain a little before Caesar's time, and drove the ancient inhabitants into the Highlands and into Ireland. Having called the Saxons to their assistance against the Scots and Picts in the fifth century, they were driven by their new allies into Wales, Cornwall, and Brittany. Their language is remarkable for the frequent changes of the initial letters of its radical words in the formation of cases and numbers; thus, from *Dea*, a man, in *Bel-temish*, is derived the plural *Per*, *Veeg*, woman, *Gronges*. Almost half of the Welsh language seems to be German, and half of the remainder is perhaps Latin or Celtic; of the Britannic, almost half resembles the Latin or French; Britain was originally inhabited by the Armoricanæ; whether they were originally Belgae or Gauls is uncertain. The country was named Britannia Minor, from the emigration of the British in 449; these new comers mixed with the original inhabitants, all speaking the same language, and in a few years became so numerous as to be able to send an army of 12,000 men to the assistance of the Emperor Anthemius.

Professor Adelang is disposed to consider the German portion of the different branches of the Celtic, which varies from one-fifth to one-half of the whole language, as an accidental mixture, and derived through different channels. But we cannot in all cases find any historical evidence of the existence of these channels; it is difficult, for example, to suppose that the Scandinavian incursions were able at any early period to influence the language of the Highlands of Scotland; and wherever it happens, as it frequently does, that no term is to be found, in the Irish, the Gaelic, or the Welsh, for expressing the same idea, besides the word that they all have in common with the German, it is scarcely possible to believe that there ever was any other Celtic word which has been so uniformly superseded by independent causes. We find, for instance, under the two first letters of the alphabet only, the words *Ap* or *Apa* in Irish, *Ap* in Welsh, *Afe* or *Ape* in Gothic; again, *Abal*, *Afal*, *Apfel*; *Angar*, *Anacmy*, *Enge*; *Bacail*, *Bach*, *Bacan*; *Barrod*, *Barr*, *Bare*; *Beoir*, *Bar*, *Bier*; *Biall*, *Biall*, *Beil*; *Bocan*, *Boch*, *Bock*; *Brathair*, *Bravard*, *Bruder*; *But*, *Bucla*, *Buile*; and the same agreement is found in almost all other instances of German words that are detected in the Irish language.

The much-disputed question respecting the antiquity of the poems attributed to Ossian has an immediate reference to the history of the Celtic languages. It has been observed, with apparent justice, by Professor Adelang, who is not in general sceptical on such occasions, that if these poems were really very ancient, their language could not but exhibit marks of antiquity. There is an Irish *Levra Levain* at Paris, written in the thirteenth century, and scarcely intelligible to the best Irish scholars of the present day; the oldest Gaelic manus- Language scripts have also peculiar expressions no longer in use; while the works supposed to be the productions of a period so much more remote are found to bear excellent modern Gaelic, impressed with all the marks of the language of Christianity, and of that of the Norwegian invaders, whereas these conquerors may be supposed to have influenced the Gaelic language immediately in Scotland, or by the intervention of Ireland. It must not, however, be forgotten, that these marks of Scandinavian intercourse are somewhat more ambiguous than Professor Adelung is disposed to admit; and that a book written in the thirteenth century is more likely to have preserved the language in an antiquated form than poems so marvellously committed to memory, from continual recitation only, by people supposed to understand them, and of course imperceptibly modifying the expression without intending to alter them. But since an invasion from Lochlin, that is, Denmark or Norway, is actually mentioned in "Fingal," the author of the poem could certainly not have been older than the seventh or eighth century, if we are to credit the historical accounts of these invasions; and since, in the poems discovered by Dr Matthew Young, St Patrick is introduced discoursing with Ossian respecting the Christian religion, we have an additional argument for denying that he was contemporary with Caracalla or Carausius, those emperors having both lived in the third century, and St Patrick in the fifth.

[It was shown by Dr Pritchard that the Kelt tongues were allied to the German, Latin, Greek, Slavonic, and Sanscrit. Did this make them Indo-European? Not without raising the value of the class. How far this is legitimate has never been asked. Yet it ought to be asked wherever an addition is made. Unless it be done, classes take indefinite dimensions. B is connected with A, C with B, D with C, and so on, till classification defeats itself.—m. o.]

14. The Etruscan is only known as the immediate parent of the Latin; but it was written in a character totally different, and was read from right to left. Notwithstanding the industry and ingenuity of Lanzi, the evidence of the accuracy of his interpretations is somewhat imperfect. We should naturally have expected to find more words of a Celtic or Gothic origin, and not merely Greek or Roman words with the terminations a little varied, as Utesi for Utves, Tribe for Tribus, and Ucte or Uctes for Hecate; still less should we have expected that the same sense should be expressed sometimes by a Greek and sometimes by a Latin word, as Utres and Populus Brund, Capros and Foros for a Boar. The Etruscans and Umbrians were originally a branch of the Celts from Rhadis, as is shown by the similarity of the names of places in those countries as well as by the remains of Etruscan art found in that part of the Tyrol; they are supposed to have entered Italy through Trent, about the year 1000 n.c., and to have afterwards improved their taste and workmanship under the auspices of Demaratus of Corinth, who settled in Etruria about 660 n.c. (See Etruscans.)

(The philological place of the Etruscan, as well the Lycian and Phrygian, has yet to be determined.—m. o.]

15. The Latin language is placed at the head of a family, rather with regard to the number of its descendants than to the independence of its origin, being too evidently derived from the Celtic, mixed with Greek, to require particular comparison. The first inhabitants of Italy appear to have been Illyrians or Thracians, Cantabrians, Celts, Pelasgians, and Etruscans. Rome, from its situation, would naturally acquire much of the languages of these various nations, and at the same time much of the Greek from the colonies in the south of Italy. In the time of Cicero, the Italian songs, supposed to be about 500 years old, were no longer intelligible, even to those who sang them. We find, in an inscription still more ancient, and approaching to the time of Romulus, Latex for Larax; and for Flores, Pleores, which is somewhat nearer to the Celtic Bleum. In the time of Numa, for Hominum liberum, we have Homonem liberum; we find, also, a n added to the oblique cases, as Captivit for Captivit, which, as well as the termination ai, in the primitive audite, is almost taken immediately from the Celtic, and is even found in modern Gaelic.

The Latin remained in perfection but a few centuries. In the middle ages, a number of barbarous words were added to it, principally of Celtic origin, which are found in the glossaries of Dufrénoy and Charpentier. At the end of the seventh century it began to acquire the character of Italian, as Campo Langue division est; and, in the eighth century, in Spain, we find, as an example of its incipient conversion into Spanish, Vendant sine pocho, de nostras terras. The formation of the Italian language may be said to have been completed by Dante, in the beginning of the fourteenth century; and it was still further polished by the classical authors who immediately succeeded him. It contains many German words, derived from the different nations who occupied in succession the northern parts of Italy, and some Arabic, Norman, and Spanish, left by occasional visitors in the south. It is spoken by the common people in very different degrees of purity. Among the northern dialects, that of Friuli is mixed with French and with some Slavonian. The Sicilians, having been conquered in succession by the Greeks, Carthaginians, Romans, Byzantines, Arabs, Normans, Germans, French, and Spaniards, have retained something of the language of each. Italy has given shelter to Iberians, Libyans, Tyrrhenos, Greeks, Carthaginians, Romans, Vandals, Byzantines, Goths, Lombards, Franks, Arabs, Pisans, and Aragonians; and the proper Sardinian language is a mixture of Latin with Greek, French, German, and Castilian. Corsica has also been occupied by a similar diversity of nations; its peculiar idiom is little known, but the dialect of the upper classes is said to approach nearly to the Tuscan.

Spain, after its complete subjugation by the Romans, enjoyed some centuries of tranquillity. The Vandals and Alans retained their power in Spain but for a short time; the Suevi, on the N. coast, somewhat longer; and from these nations the rustic Roman, which had become general in Spain, received some words of German origin. It derived, however, much more from the Arabic during the domination of the Moors, which lasted from the beginning of the eighth century to the end of the fifteenth; and, at one time, the Arabic was almost universally employed throughout the country, except in the churches. The Spanish language advanced the most rapidly toward perfection during the height of the national prosperity which immediately followed the conquest of America. It was afterwards neglected, and again more particularly cultivated by the Academy of Madrid, in the eighteenth century; as far at least as an academy can be supposed to have any influence in the modifications of a language.

The Portuguese is supposed to have received a mixture of French from the followers of Count Henry of Burgundy, under whom Portugal first formed a separate state in 1109; but the language is very different from that of the confines of France and Spain; and the nasal vowels, which are remarkable in the Portuguese, differ materially from those of the French, or of any other nation. Many Latin words are retained in the Portuguese, which are not found in any other modern language; and it is remarkable that almost all the words of the language are contracted, by the omission of some of the radical letters of the originals.

The Rhadians, in the country of the Grisons, were subdued by the Romans in the time of Augustus. They became part of the Alamannish kingdom, under Theodobert, in 539; their union with Switzerland took place in the beginning of the fourteenth century. Half of the Grisons speak the Romanish language, immediately derived from the Latin, though mixed with some German, which has been particularly made known by Mr Planta's account of it in the Philosophical Transactions. One-third speak German, with some mixture of Romanish words, and the rest a bad Italian.

France in the time of the Romans, was occupied by the Gauls, together with the Aquitanians, who were probably Cantabrians, and the Cimbrians or Belgians. From the rustic Roman, mixed with the languages of these nations, the Romance was gradually formed. In the fifth century, the Franks took possession of the north-eastern part of the country. They retained their language for some centuries, but by degrees it became mixed with the Romance and formed French, of which at least one-fifth is supposed to be of German origin, though many of the German words seem to have been admitted through the medium of the Italian. In the S. of France, the language remained more exempt from the influence of the German, under the name of the Provençal; and the troubadours contributed, especially from the eleventh to the thirteenth century, to give it refinement and currency; but, in later times, the Langue Language. d'oc has prevailed over the Langue d'oc, which is now spoken by a few of the lowest class only.

The last and least genuine of the descendants of the Latin is the Wallachian, about one-half of which is borrowed from the German, Slavonian, and Turkish. The original Thracians of the country must have been in a great measure superseded by the successive settlements of various nations; in the third century, some of the Goths and Vandals; in the fourth, the Jazyges, after Attila's death; in the fifth, some Huns and Alans; about the end of the seventh, the Bulgarians, and afterwards the Petchenegs and Hungarians, established themselves in it; and, in the thirteenth century, Wallachia became an independent state. The Latin part of the language is much like the Italian form, and had even assumed it as early as the fifth century. It must have been derived from Roman colonies, and more lately, perhaps, from the missionaries sent into the country by Pope Gregory XI. The Dacian or Hungarian dialect prevails on the N. of the Danube; the Thracian, or Cutzo-Wallachian, on the S.; the latter is more mixed with Greek and Albanian. There is also a small Wallachian colony in Transylvania.

[For remarks on the Greek, Latin, and German, see below.—n. o. l.]

The Cantabrian or Basque has many words in common with the Latin, whether originally or by adoption, and was probably in some measure connected with the Celtic dialects, which were the immediate predecessors of the Latin, though still sufficiently distinct from them. The Cantabrian Ata, Father, has some resemblance to the Irish Acair, and the Mono-Gothic Atta; Germ is not wholly unlike Cædum; Ercemus, Regnum; and Borondatia, Voluntas; the coincidence of Gua, Day, with the Tartar, is perhaps more accidental. But the word Lurre, Earth, which seems at first sight so unlike any other language, is in all probability the derivative of Tellure; and this form of the word affords also a connecting link with the Irish Tula, and may have been contracted into the more common Latin word Terra; a supposition which seems to lessen the probability of the original connection of this form of the word with the Greek Era, and the Sanscrit Stira. The Basque is still spoken in the angles of France and Spain adjoining to the northern extremity of the Pyrenees. The same people were called Cantabrians in the N., and Iberians in the S., and extended between the Pyrenees and the Rhine, as Ligurians, or inhabitants of the coast. They have adopted a few German words, perhaps from the empire of the West Goths; and they have furnished the modern Spanish with more than a hundred original words of their own. The construction of the language is extremely intricate; its verbs have eleven moods, among which are a consuetudinary, a voluntary, a compulsory, and a penitential. Larramendi's Grammar, published at Salamanca in 1729, is called El Impossible Vencido. A valuable abstract of the most interesting particulars relating to the language is found in the Additions to the Mithradate by the Baron William von Humboldt, late Prussian minister to the court of Great Britain, printed at Berlin, in 1816. Dr Young has lately remarked, in the Philosophical Transactions for 1810, that at least six of the words contained in Humboldt's vocabulary coincided very accurately with the Coptic, or ancient Egyptian, though they are not found in any of the languages of the neighbouring countries; and he infers that the chances are "more than a thousand to one, that, at some very remote period, an Egyptian colony established itself in Spain." It may be observed, that one of these words, guchi, little, appears to be also Turkish or Tartarian; so that it becomes a second instance of a coincidence between this language and the Cantabrian.

[The Bask has yet to be placed. It is nearer the Celt than is generally admitted. At the same time, its African affinities are important. It is an eminently agglutinate language.—n. o. l.]

16. The connection of the Slavonian and Lithuanian, and of the other branches constituting the Slavic family, with the languages of the Indo-European class in general, is sufficiently established, without exceeding the limits of the Lord's Prayer, by the resemblance of Nebi or Nebesi to the Cimbric Nevedd, and the Greek Nephos; and of Wolja and Chljob to the Gothic Wilja and Hlaif. The Slavonians are the descendants of the ancient Sarmatians, who were situate N. of the Black Sea and of the Danube. They were conquered by the Goths, and then driven by the Tartars and Huns into the N.E. of Germany, and the neighbouring countries. Procopius calls them Spori, and divides them into the Scali and Antes, the latter, perhaps, the same as the Wends. They formed, at an early period, two principal states, Great Russia, about Novgorod, and Little Russia, on the Dnieper, its capital being Kiev. The Rusli were a Scandinavian branch, under Rurik, to whom the Slavonians of the former state submitted in 862, whence they were called Russians; and Rurik's successor, Oleg, conquered Kiev. After several vicissitudes, the Russians were liberated by Ivan Wladimiritch, at the end of the fifteenth century; and this period was the beginning of their greatness. Their language has some mixture of Greek, Finnish, Swedish, Tartarian, and Mongolian. The ecclesiastical dialect was uniformly retained in all literary works in the former part of the last century, but now the language of conversation is generally adopted in writing. This language is more immediately derived from that of Great Russia; that of the church, which is called the Slavonish, rather from Little Russia, and especially from the dialect of Servia. The Malo-Russian dialect is somewhat mixed with the Polish, and is spoken in Ukraine and Little Russia; the Sudalonia is mixed with Greek and other languages, and is spoken in Thrace.

In 640, the Slavonians took possession of Illyria, which before that time had been overrun by a variety of other nations; and they still retain it, under the names of Servians, Croatians, and Southern Wends. The Servians are supposed to have come from Great Servia, now East Gallicia, on the Upper Vistula; the Croatians, from Great Chrubota, probably situated on the Carpathian Mountains. Cyril first adapted the Greek alphabet to the Slavonian language in Pannonia; his letters were afterwards a little altered, and attributed to St Jerome, in order to reconcile the people to their use; and in this form they are termed Glagolitic characters. The Servian dialect is intermediate between the Russian and the Croatian. The Bulgarians speak a corrupt Slavonian, which Boscovich, from Ragusa, could scarcely understand. The Ukoks are a wild race of the Bulgarians, extending into Carniola, and speaking a mixed language. The dialect of Slavonia and Dalmatia is nearly the same as that of Servia and Bosnia; the churches use the ecclesiastical language of Russia. In Ragusa, the orthography approaches, in some measure, to the Italian. The Servian is also imperfectly spoken by a small colony in Transylvania. The Southern Wends were first distinguished in 630, and were probably so named, like the Veneti, from being settled on the shores of the Adriatic, the word Wend or Wond meaning Sea. They are now mixed with Germans in Carniola, Carinthia, and Lower Styria. In Hungary there is a small colony, who call themselves Slavonics, and speak the Wendish dialect of the Slavonians. The western Slavonians, or the proper Scari, write their language in the Roman characters; but the specimens copied from Adelung are accommodated in their orthography to the German mode of pronunciation.

The Polee probably came with the Russians from the Danube into the countries abandoned by the Goths; the name Pole implies an inhabitant of plains. Their language was partly superseded by the Latin in the tenth century, when they received the rites of the Latin church; but it has in later times been more cultivated. The Cassabians, or Kashubians, in Pomerania, speak a Polish mixed with a little German. In Silesia, the names of places in the plains are Slavonian; in the hills, more lately occupied, German; but German has been the language of Breslau ever since the year 1300.

The Bohemians emigrated with the Moravians and Slovaks into their present habitations about the middle of the sixth century, after the destruction of the kingdom of Thuringia by the Franks and Saxons. There is a Bohemian hymn of the date 990, and a chronicle in rhyme of 1310. One-third of the Bohemians are of German origin, and speak a corrupt German. The Serbs or Wends came about the same time into the countries between the Saal and the Oder, from the neighbourhood of the Volga or the Crimea; a few of them are still left in Lusatia, under the name of Wends or Slavonians, and some in Misnia. In Pomerania the Wendish became extinct about 1400; but the Polobes in Luneburg, on the Leyne, kept up till lately a language consisting of a mixture of Wendish Language.

The Sarabie of the Russian vocabulary seems to be the same with this Servian.

Of the Lithuanian or Lettish language, two-thirds are Slavonian, the rest is principally German. When the Goths had removed from the Baltic towards the Black Sea, their neighbours the Ests remained for some hundred years independent; till, in the sixth century, the Slavonians incorporated themselves with them, and formed the Lettish people and language. The Old Prussian was spoken, at the time of the Reformation, in Samland and its neighbourhood, but it is now lost; it contained more German than the other Lithuanian dialects. The Prussian Lithuanian is spoken from the Inster to Memel, especially in Insterberg. The Polish Lithuanian, in Samogitia, has a little mixture of Polish. The proper Lettish is current in Lettland and Courland; it is purest about Mitau and Riga; the old Courlanders having been Finns, this dialect has received a little Finnish from them. The Livonian is another dialect, spoken by the Krewins in Courland.

[Sarmatian is a convenient term for the Slavonic and Lithuanic collectively. The Sanscrit, instead of being a language of the Indo-European class, with equal affinities to the Greek, Latin, German, and Sarmatian, has a special preponderance of Sarmatian affinities.—n. c. l.]

17. The Tushnish or Finnish, the Hungarian, and the Albanian languages, have some traits of resemblance to each other: they are placed as forming the sporadic or scattered order of the great Tartaric or Asiatic class, being in some measure geographically detached from the rest, and scattered through different parts of Europe; they immediately follow the Indo-European class, as exhibiting an occasional resemblance to some of the languages contained in it, though not enough to make it certain that the resemblance is essential or original. Thus the Finnish is said to have some coincidences with the Greek, the Laplandish with the Hebrew, the Hungarian with the Finnish, and the Albanian with all its neighbours.

The term Tyrodelde is employed as comprehending the Finns, Laplanders, Esthonians, and Livonians; a race of people of unknown origin, but in all probability unconnected with the Huns or Mongols. Their languages are remarkable for the great complexity of their structure; their nouns, for example, having from ten to fifteen cases, among which are reckoned, in the Finnish, a nominative, a conditional accusative, a factitive, a mediative, a descriptive, a penetrative, a locative, a privative, and a negative. The Esthonian has less direct variety of terminations, but several intricate combinations. There is also a great multiplicity of dialects, partly from a mixture of Scandinavian, and partly from other causes; in Lapland almost every church has a peculiar version of the service kept for its use. The Finnisch is intermediate between the Laplandish and the Esthonian. The Esthonian are the Esti of the Romans, the name implying Easterly, and being appropriate to the country, and not to the people. The principal dialects of their language are those of Reval and of Dorpat; some authors also consider the dialect of the Krewins in Courland as belonging to it. The Livonian is much mixed with other languages, and has been almost superseded by the Lettish. Amongst the Laplandish words which Rudbeck has derived from the Hebrew, we find Adathem, Earth, like the Hebrew Adamah; Hadass, Now; Heb. Khadash, Hadache, the Moon; Heb. Khadash, Jed, the Hand; Heb. Id., Ise, Man; Heb. Ish, Pothi, Persuaded; Heb. Pattech, Szakke, Law; Heb. Tzedek; and Safotki, Rested; Heb. Sabbath. In the Finnish Kiruna is something like the English and German Hen.

18. The Hungarians inhabited, in the fourth century, the country of the Bashkirks, between the Tobol, the Volga, and the Jaik, perhaps as colonists, since their name signifies strangers. Their language is spoken in this neighbourhood as late as the thirteenth century; in the sixth they were conquered by some of their Turkish neighbours; in the end of the ninth they were forced by the Petchenegs, a Tartarian nation, to remove nearer to the Carpathian Mountains. They may have been engaged in the German wars, and their country having been occupied during their absence by the Bulgarians, they took possession of the Bulgarian kingdom on the Theiss, as well as of Pannonia. Their language is somewhat like the Finnish, but the people are very different from the Finns in appearance; which might indeed be the effect of a difference of climate; but, in fact, the language appears to be still more like the Slavonian with a mixture of a multitude of others; language, it has some words from various Tartarian dialects, German, French, Latin, Armenian, Hebrew, Persian, and Arabic; but it has no traces of the Mongol, nor is it possible that the people can be descendants of the Hun, whose character and most of features can never be eradicated. The word Coach, so general in Europe, is originally Hungarian, having been derived from the town of Koza, where coaches are said to have been invented. The Szeklers, in Transylvania, speak a language like the Hungarian; it is uncertain whether they are a Hungarian colony, or remains of the Petchenegs; but, however this may be, there is little doubt that the Hungarians are principally of Tartarian extraction, though much mixed with other nations.

19. The Albanians speak a language, of which a considerable portion is Greek, Latin, German, Slavonian, or Turkish; but the rest seems to be perfectly distinct from any other language with which we are acquainted. They are probably connected with the Albanians between Mount Caucasus and the River Cyrus, who are supposed to be derived from the Alani; some of them seem to have entered Bulgaria as late as 1305. In 1461 many of them fled from the Turks to Italy and Sicily, where they still exist near Reggio and Messina. The Clementines are an Albanian colony, who followed the Austrian army in 1737; such of them as escaped from the pursuit of the Turks established themselves in Syria.

[The Albanian is less, the Hungarian more, Finn than the text makes them.]

The ordinary term at present in use for the name of the class is Ugrian.

It is chiefly in the southern part of the area that the Ugrian languages are sporadic. The Hungarian has intruded into Hungary; the Russian into the governments of Permia, Viatka, Kazan, Simbirsk, and Saratov, &c., all of which were originally Ugrian. The Lap, Esthonian, Finn of Finland, &c., are all continuous.

The so-called Indo-European tongue nearest to the Finn is the Lithuanic.

The Livonians (Lives) are a small population of Courland, originally spread over Livonia, or Livland, to which they gave the name.

The locality of the Natii was the amber-country of East Prussia, not Esthonia.—n. c. l.

20. The languages referred to the Caucasian order have little to distinguish them from the rest of the class, except their geographical situation, in the immediate neighbourhood of the Caucasian Mountains. They have a general resemblance to some others of the languages of Northern Asia, and particularly to the Samoedic dialects, spoken on the mountains between Siberia and the Mongolia. Except the Armenian and Georgian, they are scarcely ever employed in writing; and principally, perhaps, from this cause, they exhibit as great a diversity, in the space of a few square miles, as those of many other nations do in as many thousands. It is only conjectured that most of the inhabitants of these countries are derived from the miscellaneous fragments of expeditions of various nations, left behind in their passage through them at different periods.

[The use of the term Caucasian in ethnological physiology (the Caucasian race of Blumenbach, containing Arabs, Jews, Europeans, &c.) as a class-name for a heterogeneous group of families, engenders ambiguities. The present writer has suggested the name Dioscurian; Dioscurias being the name of a town of Caucasus, not only mentioned by ancient writers, but mentioned with reference to the remarkable multiplicity of languages and dialects spoken in its neighbourhood—this being a notable characteristic of modern Caucasus.

The Dioscurian class he arranges as follows:—

1. Circassian or Tcherkess, falling into the Abassian, Tcherkess, and Kabardinian divisions.

2. The Lesgian, falling into: a. The Taheteki (Kirtie)—the Tahetektsi, Ingush, and Tush; b. The Lesgian Proper—the Avar, Dido, &c.

3. The Armenian—ancient and modern.

4. The Iron, or Oset, leading to the Persian, and containing numerous Indo-European words.] The nearest affinities of the most northern and most characteristic of the Caucasian dialects, the Circassian, are with the Tibetan, and the western monosyllabic tongues. It is only because the comparison has been most carefully made with the Samoyed, that they seem to be more akin to that family than to others. They have miscellaneous affinities with all the Turanian languages. There is no reason to believe that any of them are other than aboriginal to the soil.—n. o. l.

The connection of the Armenian with the Sanscrit and the Persian is just enough to make it equally possible that the coincidences may have been derived from a common parent, or that one language may have simply borrowed detached words from the other. We find, in different parts of Mr Townsend's work, about ten Armenian words resembling some other language; these are—Air, a Man; Irish Air. Atama, a Tooth; Greek Odontak. Cherek, a Horse; Sanscrit Chatur. Dor, a Door. E, I; Latin Est. Es, I; Russian Jore. Gas, a Goose; German Gans. House, a House. Lakell, to Lick; Greek Lechein; and Sirt, the Heart. Nothing is known of the history of the Armenian before the time of Moses, who translated the Bible into it in 405; the historian Moses, of Chorene, was his pupil. The language flourished till the year 800, and is still preserved in tolerable purity in the cloisters. The common people speak a dialect more corrupt and mixed. The fathers of the Armenian convent at Venice have been very laudably employed in the improvement of the literature of their nation by the publication of several very elegant editions of Armenian books, which have been executed at their press; in particular, of an Armenian translation of Eschilus, containing some passages which are not extant in Greek, and said to have been copied from a manuscript of great antiquity at Constantinople. It is, however, very remarkable, that, as they candidly confess, the copy, when first received by them, contained the corrections and additions of Seinger, in conformity with the text of the printed Greek edition; and the copyist, when questioned, asserted that he had merely transcribed the inserted passages of his own accord, and in illiteracy, in order to make the work more perfect. Such an Armenian Eschilus is a very handsome book, and every word calculated to do credit to the Venetian editors and their patrons; a Latin translation of it has only been published by Angelo Mai at Milan.

(The persons of the verbs, and the numerals are the points in which the Armenian most agrees with the so-called Indo-European languages; but these are just the points in which numerous languages, never considered Indo-European, also agree. The gossarial affinities are most with the Iron (Osset), and Persian. It graduates through these into the other allied groups. Nevertheless, many authorities make it absolutely Indo-European.—n. o. l.)

21. The Georgians are supposed to have derived their name from the River Cyrus or Gur, and to have extended formerly to Colchis, under the denomination of Iberians. Moses of Chorene, in the fifth century, mentions the Georgian translation of the Bible. The old language is still preserved in the churches, and the common dialect of the country is derived from it, together with the Kartvelish, Imirzitish, Mingrelish, and Svanetish, which are varieties of that dialect; the Tushetish is mixed with some Kistic. The Georgians have no fewer than thirty-seven letters, and among them a variety of aspirates and sibilants, of no very agreeable sounds.

[To the Georgian dialects of the text, add the Lazic of the Mohammedans of the parts between Batoum, Kara, and Trebizond. Subtract the Tushi (Tubetish) which is Tushetish. The Kartulinish is the most cultivated, the Svanic the rudest, of the Georgian forms of speech.—n. o. l.]

22. The Abassic nations seem to be the oldest inhabitants of the Caucasian country. 23. The Circassians are situate to the E. of them, on the promontory of North Caucasus. 24. The Ossetes, on the left of the Terek, N. of the mountain; the dialect of the Dugors is scarcely distinguishable from this. 25. The Kircic, spoken by the Ingushians, and their neighbours at the head of the Terek, is connected with the Tusbetish Georgian. 26. The Lesghians, E. of Caucasus, on the Caspian Sea, have a number of distinct dialects, or rather languages. Thus the Checheny and Avaric, the Dido, the Avari, Kalmak, the Andi, and the Akhshun, have little connection with each other, except that the Dido somewhat resembles the Chamsgag, of which the Avaric, the Anteg, and the Dzhur, seem to be subordinate dialects. The Kasi Kumik appears to have adopted some words of the Armenian, and the Andi and Akhshan of the Georgian. The dialect of Kubeshan resembles that of Akusha, and retains no traces of a supposed European origin.

[The Abassic is Circassian; the Circassian forms of speech being the most monosyllabic and uninflected of Caucasus—also most like the Tibetan.

The Osset (or Iron) is the most like the Persian, and has by many been considered Indo-European. But the Georgian is connected with it. What do we infer from this? That the Georgian also is Indo-European, or that the Osset is not so? The former is the inference of learned men with whom the present writer (who holds the latter view) reluctantly differs.—n. o. l.]

27. The languages of the central and elevated parts of Asia are comprehended in the order Tartarians; they extend from the Caspian Sea to the mouth of the Amur, through countries which have been in former ages the constant scenes of emigration and barbarism. The Turco-Tartarians are supposed to correspond to the descendants of the Magog of the Scriptures, and to some of the Scythians of the Greeks. The Turks of Turkestan seem to have been the Massagetae and Chorasmii of the ancients; their country extended N. of Persia and Tibet, from the Caspian to the Altai Mountains. In the twelfth century they were brilliant and victorious; at present none of their descendants only are left in the neighbourhood of the Mongols, and their language is no longer spoken; the Tartars scattered in Persia and Arabia are derived from the same race. The Demans, now commonly called Turks, left Turkestan in 915, and succeeded in the conquest of Persia. They were denominated Osmans, from one of their leaders, in the fourteenth century. Their language has been much mixed with Arabic and Persian. This language, with the neighbouring dialects, has been considered in the table as belonging to a family called Caspian, the word Tartarian having been previously applied to the whole order. Several of these dialects exhibit a mixture of words from the language of the Mongols, which, as well as the Calmusk, has a sufficient connection with them to be arranged as belonging to the same Tarco-Tartarian family. It would, perhaps, be equally correct to consider some of them rather as distinct languages than as dialects of a single one; but it is not easy to discriminate those which are entitled to this rank; and, on the other hand, some specimens have been admitted, from the Comparative Vocabularies, which scarcely deserved to be noticed as separate dialects. The Bucharianos are situate between the Oxus and Jaxartes, on the River Koly. They still retain some traces of a superior degree of civilization, by which they were once distinguished. Their language is little known, but it seems to be at least as much connected with the Median and Arabian families as with the Caspian. The Tartars were described by the terms Scythians, Bulgarians, Avari, and other appellations, before they were conquered and united by Genghis Khan the Mongol. In the year 1552 they became subject to the Russians. The most westerly are the Nagais or Nagais and Crimean Tartars; their language is much like the Turkish, but mixed with some Mongol. Those of Cumania in Hungary have now forgotten their original language, and speak the Hungarian, the last person who understood the Cumanian having died in 1770. They entered Hungary in 1086, and became Christians in 1410. The Tartarian, or rather Caspian, is spoken in great purity about Kazan; a dialect somewhat different in Orenburg; and another by the Kirghishes, who occupy part of the ancient Turkestan. Among the Siberian Tartars, the remains of the kingdom of Turan, some are Mohammedans; others, as the Tura-kinde villagers, have been made Christians by the Jesuits; the Archbishop Philopold performed the ceremony of baptizing them, by ordering his dragoons to throw them into the body into the river." The inhabitants of the banks of the Tura, a branch of the Irish, are said to be derived from the Bucharianos. The Tula-kinde farmers enjoy the same advantage as the Turalmizic, and are considered as Christians by the Russians. The Telente, in Songor, are heathens, nearly like the Shamanites of India. The Jedute extend along the Lena to the sea; their language contains some Mantshuric and some Tungusic. That of the Tshuawashes, on the Volga, is said to have been once completely distinct from the Tartarian or Caspian; and even at present, Language, though more mixed with it, may require to be classed as forming a separate species.

The Mongols are marked by their features as a race very different from the other Tartars; the character of their countenance seems to have been inherited from father to son, and never to have been completely effaced; their original habitation appears to have been in the neighbourhood of the Altai Mountains. The description of the Huns, found in Ammianus Marcellinus, and others, agrees exactly with the present Mongols, whom the Chinese still call Kiong nu; and not particularly with the Calmucks. The proper names of the Huns are also found to be explicable from the Mongol language. In the first century they were driven westwards by the Chinese; under Attila they penetrated into the middle of Europe; and they were little less successful at subsequent periods, under Genghis Khan and Timur Leng. When they were expelled from China, after having held it in subjection for more than a century, they carried back no civilization with them; nor was either of the languages permanently affected by this temporary mixture of the nations, although the physiognomy of the Chinese bears ample testimony of its having once existed. The construction of their language seems to be very indirect and figurative. Mr. Townsend has copied from General Vallancey's long list of words, in Strathenbergh's Mongol Vocabulary, which agree very remarkably with the Irish; among these we find Are and Ere, man; Irish Ar, Air, Fear, Areal, a spindle; Irish Oide, Alemamode, an apple-tree; Irish Amhalhaidhe, Aaoe, to ask; Irish Acadha, Baithu, I live long; Irish Baith, long life, Buiga, a buck; Irish Roe, a bo-goat, Choy, a ewe; Irish Choi; and Choraghe, a lamb; Irish Cioragh; without going any further in the alphabet. The last two instances are very striking, and seem to point very strongly at that part of the East from which the Celts may be supposed to have originally emigrated. The Calmuck dialect is somewhat mixed with the neighbouring Tartarian. The Tagarians, or Danrians, between the Lake Baikal and the Mongol Hills, are said to be of Manchurian origin; but their language evidently resembles the Calmuck. The Bututtiok is from the Russian vocabularies.

23. The Mantshurians are sometimes improperly called Eastern Mongols; they are subjects of the empire of China. Their language is simple, and not much like the Chinese, though evidently derived from the Manchurian class. It has some few words in common with the European languages, as Kirri, patient; Kirri, German; Clever, Latin, tanak, Farsi, Furore, Lapta, rage; Lapta, German, Sanguin blood, Sangue, Ama, a year, Anima; but, considering the remoteness of their situation, we can scarcely form any conclusion from the occurrence of these resemblances. M. Abel Rémyan hold the appointment of professor of this language at Paris; but it was found difficult for him to render its study very popular in the midst of so busy a metropolis. Whether the language of the island of Sagalass, opposite to the mouth of the Amur, is a dialect of the Mantshurian, or totally distinct, and requiring to be classified with the insular languages, appears to be not yet sufficiently ascertained. The Coreans have been supposed to be a mixture of Mantshurian and Chinese; the Coreans do not understand either of those languages when they are spoken, but this fact is perfectly compatible with the supposition.

29. The Tungusians, in the E. of Siberia, subject to the Chinese, speak a peculiar language, mixed with some Mongol. The Russian vocabularies contain specimens of a variety of their dialects, besides those of the Tsiapogirs, on the Jenisei; and the Lamuti, on the Sea of Ochotsk, none of them particularly interesting or remarkable.

(The Bucharian Proper is Persian; the Uzbek of the dominant stock only being Turk.

The Tashmush is as often considered Ugrian as Turk.

The Huns are quite (if not more) likely to have been Turks than Mongolians.

The present classification of the languages allied to the Turk is as follows:—

1. The Tartarian, or Scythic stock contains the (1.) Mongolian, (2.) Turk, Tartarian, (3.) Tungus, and (4.) Ugrian branches.

The Manchu is a section of the Tungus branch. The Corean belongs to a different division.—[c. o. L.]

30. The languages belonging to the Siberian order occupy the principal part of the N. of Asia, between the mountainous Tartarian territory and the Frozen Sea. At the commencement of this order, we find a variety of inconsiderable nations Language, in the neighbourhood of the confines of Europe and Asia, which have their distinct languages, probably formed, in times comparatively modern, out of the fragments of others. They have almost all of them some Finnish words, but none a sufficient number to justify us in considering them as dialects of the Finnish language, although the people were very probably connected with the Finns as neighbours, in the middle ages, on the banks of the Dwina and elsewhere. The Serjance, in the government of Archangel, speak nearly the same language with the Permians, who are partly in the same government, and live in that of Kasan. The Woticks, on the Wiatka, also near Kasan, have a dialect which seems to be intermediate between the Permian and the Tatarian. 31. The Woguls, situate on the Kama and the Irtysh, afford specimens of several dialects in the Russian collection; they seem to have borrowed a few words from the Hungarian, and much more from the language of the Ostiaks (32), who are also divided into several races. 33. The Tcheremissins, situate on the Volga, in Kasan, have a little mixture of Turco-Tartarian. 34. The Mordovins, on the Oka and Volga, have about one-eighth of their language Finnish, and also some Turco-Tartarian words. The Mordvin is a dialect differing but slightly from the Mordvin. 35. The Teptjeriok are people paying no taxes, who originated from the relics of the Tartaro-Kasanic kingdom in the sixteenth century, and who are said to speak a language peculiar to themselves. The arrangement of all these dialects must remain very imperfect, for want of a greater number of specimens of their peculiarities.

36. The Samojedic nations are situate N. of the Tartars, by whom they may possibly have been driven into their present habitations. Their languages seem to have some affinity with the Cassanian and Lesgian dialects, and some of them with the Wogulic and Ostiak families; the specimens in the Comparative Vocabularies seem to have been multiplied somewhat too liberally. 37. The Camasheks are situate on the right of the Jenisei; they are Shamanites or Buddhists; their language seems to be a mixture of several others, and is divided into several very distinct dialects. The Koibats have been baptized; they have borrowed some words from the Turco-Tartarian family. The Motors are situate on the Tuba. 38. The Ostiaks on the River Jenissei afford us five specimens of languages totally different from those of the Ostiaks already mentioned, but nearly connected with each other, so that they may properly be called Jenisei-Ostiaks. 39. The Jakutins, or Jakutians, are few in number; they are situate between the Jakutini and the Tchutshi; they have some Jakutian words mixed with their language, and some Tchutshian.

(The Teptjeriok are Turk rather than Ugrian.

All the other populations of this section are Finno-Ugrian, viz., Siretanian, Permian, Mordvin, Samoyed, &c. The term Siberian is not used in the sense of the text.—[c. o. L.]

40. The Korialk and the Tchutshi occupy the north-easternmost point of Siberia; the proper Korialk is spoken on the Bay of Pemshin; the Kolyma on the River Kolyma, the Tigil on the Tigil, in Kamtschatka; and the Karagin on the island Karaga; the Tchutshi has been considered as a dialect of the Korialk. 41. The Kamtschatkans are a little farther S.; the Tigilic Kamtschatkan is found, however, on the N. of the Tigil; the Srednitsch to the W., on the Bolshain, and the Jochayshic on the River Kamtschatka, and towards the South Cape. The languages of the neighbouring parts of America, according to Professor Vater, greatly resemble the Tchutshi.

The Insular order of the Tartaric or Atactic class of languages must be understood as comprehending all the Asiatic islands E. of Borneo. 42. The language of the Curiles is spoken not only in the principal of these islands, but also in Kamtschatka, about Cape Lopatka; but in some of the islands the Japanese is spoken. The Japanese derive themselves from the Chinese, but their language contradicts this opinion; they have evident traces of Mongol extraction or relationship. The amiable islanders of Loochoo will long be remembered by the British public, for the hospitality they showed to the Alceste and the Lyra; their language appears to be related to the Japanese, as might be expected from their situation.

(In the opinion of the present writer, the Korialk (to which the Kamtschadal belongs), the Curile, the Japanese, and the Corean, constitute a class, for which he recommends the name Language. Peninsular, transitional to the Ugrian tongues on the one hand, and the American on the other.—n. o. l.

Formosan was conquered by the Dutch in 1629, but in 1661 it was taken from them by a Chinese pirate; the next year some books were printed in the Formosan language in Holland, the recapture of the island not being yet known there; in 1682, it was finally given up to the Chinese government.

47. The Meluccan is considered by Dr Leyden as an original language; that of Magindanao contains some Malayan, Moroccan, Tagalogh, and Bagis. The Tagalogh, or Gala, is the principal language of the Philippines, and almost as generally understood in that neighbourhood as the Malay and Hindustani in other parts; it is allied to the Malayan and to the Javanese, and was probably derived in great measure from these languages; it also resembles in some measure the Bugis. The Bissayish is a ruder dialect of the Tagalogh. The Sulu differs but little from these dialects, being derived from the same sources. The Bugis is the language of Celebes; it is supposed to be more ancient than the Javanese; it seems to contain no Sanscrit, but much Malay, Tagalogh, and Javanese, and some of the old Ternate, or Moluccan; it is written in a peculiar character, and some good poetry is found in it. There is a dialect called the Manggarai. The Bima somewhat resembles this dialect; it is spoken in the eastern parts of Sumbawa, and the western of Ende or Flores; it is written either in the Bugis or the Malay character; it seems to have a distant resemblance to the language of Orissa. The dialect of Sumbawa exhibits some slight variations. A few single words, as Matto, the eye, and Motte, death, are found to coincide in almost all the islands of the Pacific Ocean; the languages of which, notwithstanding their immense distances, seem to differ less than those of the inhabitants of some very small continental tracts; and they might probably be divided into a few well-defined families, if our knowledge of them were more complete. The resemblance of Matto to the Arabian Mot, and the Latin Mactare, is probably accidental.

[All these languages are Malay.—n. o. l.]

The number of the African languages is supposed to amount to one hundred or one hundred and fifty, and as many as seventy or eighty of them have been distinguished with tolerable accuracy. The population of Africa seems to have been derived from Arabia, and, as some critics think, rather from the southern than from the northern part; a great number of its present inhabitants are negroes, but these cannot be distinguished from the rest by any infallible criterion. The account given by Ptolemy of the interior part of the country appears to be wonderfully accurate and extensive; although some of his measures seem to be erroneous, and not sufficiently reconcilable with the truth, even by adopting Major Kennell's hypothesis respecting them. It is, however, remarkable that Ptolemy followed Hipparchus in extending the eastern coast of Africa to the Ganges, although more correct ideas of its form had been entertained at Alexandria before his time.

The Egyptians demand the priority in treating of the inhabitants of Africa, from their early connection with ancient history, both sacred and profane. It is observable that the representations of the old Egyptians have countenances more or less approaching to the negro physiognomy, though the dry bones of the skeleton have that character somewhat less decidedly than they must have had when clothed with the thick lips and flatish noses of the generality of the representations; at the same time there are sculptures of great antiquity, which exhibit features not unlike those of correct Grecian or Roman beauty; and others have a considerable resemblance to the Arabian nation. At present the people of middle Africa in general are more or less like negroes, but they are somewhat less dark, and their noses and lips are less peculiar; the women sometimes screamed if Burckhardt made his appearance on a sudden, and called him the devil, because he was white. The Egyptians are supposed by some writers to have received their civilization from Ethiopia; but there are at present no traces of the remains of high civilization farther S. than Nubia, except a few scattered monuments about Axum, of no great antiquity. The Egyptians were at first called Copts by the Carthaginians, and their language has been commonly distinguished by the appellation Coptic, that is, as written in characters which are principally Greek, and frequently intermixed with a number of pure Greek words; but not a single fragment of Coptic has yet been discovered in this form that is earlier than the establishment of Christianity in Egypt; and it seems probable that the character was introduced by the early Christians at the time of the translation of the Scriptures into Coptic, which is certainly of very high antiquity. The Greek authors frequently mention an Egyptian alphabet of twenty-five letters; but no traces of any such alphabet are found in the multitudinous inscriptions or manuscripts that have been preserved by the exertions of the numerous and adventurous travellers who have lately visited the country. The Greek words mixed with the Coptic are not considered by the grammarians as incorporated with the language, nor are they admitted into the dictionaries. The genuine language bears very evident marks of great antiquity; its construction is simple, and often awkward, and a great number of its words are monosyllables. We have positive evidence of its having remained unaltered from the time of Herodotus, Plutarch, and other Greek authors; and it affords us the etymology of the name of Moses, and of some other words mentioned in the Scriptures. It exhibits a few coincidences with other ancient languages, but not enough to enable us to consider it either as the offspring or the parent of any of them, except that it gives us something like an explanation of the meaning of some of the Greek particles. Out of one hundred and fourteen original Egyptian words, there are fifty-two that resemble the Greek, twenty-seven the German and English, eighteen the Hebrew, three the Syrian, two the Arabic, two the Sanscrit, one the Slavonian, and one the Cantabrian. It is, however, probable, that a person more intimately acquainted with the languages of the Arabian family would have been able to find a much greater number of coincidences, since nations which had so much intercourse as the Jews and the Egyptians could scarcely fail to have many words in common, even if their languages had been at first completely different; and probably many of the Arabic roots, which are not Hebrew, may be found in the Egyptian. To the Cantabrian word included in this enumeration Dr Young added five others, the whole six being—Berria, a cow; Ora, a dog; Guchla, little Oxyrhine; Omoa, a wolf—whence the Spanish Osa; and Shashf, seven. In Coptic, Berri, new; Uhor, a dog; Kudshi, little; Olk, bread; Uomah, a wolf; and Shashf, seven. Hence he infers, that "if we consider these words as sufficiently identical to admit of our calculating upon them, the chances will be more than a thousand to one that, at some very remote period, an Egyptian colony established itself in Spain; for none of the languages of the neighbouring nations retain any traces of having been the medium through which these words have been conveyed. On the other hand," he continues, "if we adopted the opinions of a late learned antiquary, General Vallancy, the probability would be still incomparably greater that Ireland was originally peopled from the same mother country, since he has collected more than one hundred words which are certainly Egyptian, and which he considers as bearing the same sense in Irish; but the relation, which he has magnified into identity, appears in general to be that of a very faint resemblance; and this is precisely an instance of a case in which it would be deceiving ourselves to attempt to reduce the matter to a calculation." It may, indeed, be imagined that the Egyptian dominions may formerly have extended to the Straits of Gibraltar, and that Spain may have derived a part of its population from this part of Africa, which approaches so near to it; but it could scarcely have happened that no traces of Egyptian monuments should ever have been found at any distance from the Nile, if that active people had really occupied any considerable portion of the neighbouring continent. The word Chemistery, in Greek Chemia, is well known to be derived from the Egyptian; it has successively been compared, by the Quarterly Reviewer, to Chm or Chem, heat, and to Chem, secret; the latter being the more probable origin of the two; and a third etymon might be found, if it were required, in the Dahem, or Ghom, to find, or invention. The Coptic language has been nearly extinct for about two centuries; but the service has been read in Coptic much more lately in some of the churches, though it has now been almost entirely superseded by the Arabic. The proper Coptic, or Memphite, which was the dialect of Lower Egypt, is supposed, from a word quoted by Herodotus, to be the most ancient; the Sahidic or Thebaic of Upper Egypt was probably Language preserved for a longer time, especially in some of the monastic libraries; there is a separate version of the principal part of the Bible in this dialect, fragments of which have been published by Mingarelli and Wrede; a third dialect, much resembling the Thebaic, is commonly called the Bashmuric; and a fourth, the Oasisit, has been partially made known by M. Quatremère de Quincy. The Egyptians have left no traces of their language among the people who at present occupy the countries that they inhabited; the Nubian vocabularies collected by Bureckhardt contain no Coptic words; the people are of different Arab races, but have acquired peculiar dialects, probably mixed with those of the neighbouring negro nations, of several of which we find specimens in Mr Salt's Voyage to Abyssinia.

The Coptic is an eminently agglutinate language.

As far as the hieroglyphics have been deciphered, the language given by even the earliest is Coptic.

Syllable for syllable, Copt = Egy-gypt.

The Barabra is a Nubian dialect.

The nearest affinities of the Coptic are those that geography suggests, viz., the Semitic, and the other Sub-Semitic forms of speech (see p. c.)

One of the most learned, as well as the most adventurous and industrious, of modern travellers, has remarked some coincidences between the old Egyptian language and that of the Barabra, who are neighbours of the Nubians, and extend to the confluence of the Tanaeum and the Nile. The Gezir and Amharic have already been mentioned as descendants of the Arabian family; they seem to have introduced some traces of this extraction into several of the neighbouring dialects, probably by the translations of the Scriptures, or by the use of the Koran. Professor Vater has taken some pains to prove that the language of Amhara, the Camara of Agatharchides, is wholly independent of the Ethiopic and Arabian; but in this he appears to be mistaken. It exhibits some slight resemblance to the Sanscrit, in a few instances; thus Tcheguer is hair, in Sanscrit Sichicura. Macritti tells us that there are, in the whole, fifty Abyssinian dialects; but he has probably exaggerated their number. We have obtained more authentic information respecting them from the collections of Bruce, and of his editor, Murray, and still more lately from Dr Seetzen and Mr Salt. Of the Mek of Dongola, the representative of a long race of the Christian kings of Nubia, little is now known, except that he is in a great measure dependent on the king of Senaar on the one hand, and has been expelled from a part of his territories by the Manselukes on the other. Of the Agows and the Gafats, neighbours of the Abyssinians, and situated on the Bahr el Azrek, as well as the Jewish Falashas, who are scattered over the country, especially in Demena, we have read much in the historical romances of Mr Bruce, which certainly give a faithful picture of the countries to which they relate, notwithstanding some unaccountable inaccuracies with respect to the personal adventures of the author.

The N. of Africa is occupied by inhabitants not much differing in appearance from the Arabs. Its three principal divisions are—the coast, the country of wild beasts, and the desert. The latter Arabs have expelled the earlier Africans from the first division, and partly from the second; the Berbers occupy the third, inhabiting principally the oases, or cultivable islands, scattered through the desert from Mount Atlas to Egypt, and speaking, as Horneman first ascertained, the same language throughout its vast extent. They were first well described by Lee in Africa; they are probably the remains of the Mauritanians, Numidians, Garamantes, and Garamantians. There is no foundation whatever for the opinion of some modern authors of celebrity, that their language is derived from the Punic. We even find, from Sallust, that the Numidian language differed from the Carthaginian; and from Valerius Maximus, that it was written in a peculiar character, perhaps the same with that which is found in the inscriptions from Leptis, now in the court of the British Museum. The language of the Canaries considerably resembles the Berber; the milk is Acho in Herber, and Ako in the Canaries. These islands were discovered in 1330, and afterwards conquered, with some difficulty, by the Spaniards. The inhabitants were a fine race of men, and lived in comfort and tranquillity; and they still present some traces of their original character and condition.

The Gafat are Semitic; the Agows and Falasha Sub-Semitic. So are the Berber, or Amazigh tongues.

The Dongola forms of speech are Nubian. Barabra (different from Berber) and Nubian are nearly synonymous.

The country between the desert Zahara and the Niger is inhabited by a race of people who have a great resemblance to negroes, but are somewhat different from them. In the E. are those of Soudan or Afan, and Hogerman; in the W. the Foutafo; the Fellata are a branch of these, extending considerably to the N.E., with a mixture of negroes.

Of the languages of the negroes, strictly so called, many interesting specimens have been collected by the zeal of the evangelical missionaries in the Caribbee Islands, and published by Doldendorp, in his Account of the Mission; but they do not afford us sufficient materials to enable us to trace any extensive connections or dependencies among their multifarious dialects.

There are some points of coincidence between the language of Madagascar and those of the Malays, the Philippine islanders, the Boctiama Caffres, and the Coroma Hottentots; there are also a few words, in many of the African dialects, borrowed from the modern Arabic, not, as Court de Gebelin would persuade us, from the Phoenician. Nor can any other of the affinities be very distinctly established.

The Caffres have little of the negro character except the black colour, and less of this as they become more remote from the equator. They are supposed to extend across the whole of Africa, immediately N. of the Hottentots, as far as Benguela and Quillon. The Hottentots, with their neighbours the Bejesemen, speak different dialects of the same singular language, in different parts of their country. Of that of the Damamaras, little or nothing is known. Lichtenstein has classed them as Hottentots; but Baron, who was better acquainted with the considers them as Caffres.

[For a fuller sketch of the African languages in general, see below; here observing that—

a. The Malay of Madagascar is Malay.

b. There are two different populations known as Damamaras, one of which is Caffre, one Hottentot.—n. o. l.]

ADDENDA.

On the Languages of Armenia.

The Eskimo.—The extent to which this language is isolated has been greatly exaggerated. It has decided and direct affinities with the north-eastern Ugrian and the Peninsular languages on one side, and the north-western languages of America on the other.

It is spoken in Greenland, Labrador, the coast of the Arctic Ocean, Russian America, the Aleutian Isles, and the parts about Tshuktschi-Noss in Asia.

It is succeeded by the languages of—

The Athabaskan Group.—The vast size of the area over which the Athabaskan tongues have spread themselves has commanded less attention than it deserves. It should command attention if it were only for the fact of its touching both the oceans—the Atlantic on the one side, the Pacific on the other.

For the northern Athabaskans, the main body of the family, the philological details were, until lately, eminently scanty and insufficient. There was, indeed, an imperfect substitute for them in the statements of several highly trustworthy authors as to certain tribes who spoke a language allied to the Chepewyan, and as to others who did not,—statements which, on the whole, have been shown to be correct; statements, however, which required the confirmation of vocabularies. These have now been procured; if not to the full extent of all the details of the family, to an extent quite sufficient for the purposes of the philologist. They show that the most western branch of the family, the Chepewyan proper, or the language of the Northern Indians, was closely akin to that of the Dogrib, the Hare (or Slave), and the Beaver Indians, and that the Dabodian, called from their reptile habits the Mauvais Monde, were but slightly separated from them. Further west a change took place, but not one of much importance. The interpreters were understood with greater difficulty, but still understood.

The Sikani and Sussee tongues are known by specimens of considerable length and value, and these languages, lying as far S. as the drainage of the Saskatchewan, and as far W. as the Rocky Mountains, are, and have been for some years, also known as Athabaskan.

Then came the Takulli of New Caledonia. This was the Nagaïl, or Chin Indian of Mackenzie, or nearly so. Now, Nagaïl I hold to be the same word as Tackulli, whilst Chin is Tshin = Dinne = Tsoi = Atna = Knoi = Mosa. The Takulli division falls into no less than eleven (7) minor sections, all of which but one end in this root, viz., tin.

1. The Tsu-tin, or Talko-tin. 2. The Talko-tin, or Chilko-tin, perhaps the same word in a different dialect. 3. The Nasco-tin. 4. The Thetlic-tin. 5. The Tatsino-tin. 6. The Nulcan-tin. 7. The Nanaco-tin.

Sir John Richardson, from vocabularies procured by him during his last expedition, the value of which is greatly enhanced by his ethnological chapter on the characteristics of the population which supplied them, has shown, what was before but suspected, that the Loucheux Indians of Mackenzie River and Athabaskan, a most important addition to our knowledge. Now, the Loucheux are a tribe known under many names—besides that of the Quarellers, under that of the Squinters, under that of the Thyroids, and Digots. Sir John Richardson calls them Kutsikam, name which we shall find in several compounds, just as we found the root tin in the several sections of the Takulli, and as we shall find its modified form dinni among the eastern Athabaskans. The particular tribes of the Kutsikam division occupant of either the eastern frontier of Russian America, or the north-western parts of the Hudson's Bay territory, are (according to the same authority) as follows:

1. The Artez-Kutsikam = Hard people. 2. The Tshin-Kutsikam = Water people. 3. The Tatzel-Kutsikam = Rampart people; falling into four bands. 4. The Yoyse-Kutsikam = People of the shelter. 5. The Vanta-Kutsikam = People of the lakes. 6. The Nuyse-Kutsikam = People of the open country. 7. The Tlagga-silla = Little dogs.

This brings us to the Kenay. Word for word, Kenay is Knoi = Tsoi, a modified form of the now familiar root tso = mos, a root which has yet to appear and reappear under various new, and sometimes unfamiliar and unexpected forms. A Kenay vocabulary has long been known. It appears in Lisiansky tabulated with Kadlik, Sitkan, and Unalaskan of the Aleutian Islands. It was supplied by the occupants of Cook's Inlet. Were these Athabaskan? The present writer owes to Mr Jabloster the suggestion that they were Loucheux, and to the same authority he is indebted for the use of a very short Loucheux vocabulary. Having compared this with Lisiansky's, he placed both languages in the same category—rightly in respect to the main point, wrongly in respect to a subordinate. He determined the place of the Loucheux (Kutsikam, as he would now call them) by that of the Kenay, and made both Kolush. He would now reverse the process, and make both Athabaskan, as Sir John Richardson has also suggested.

The Atna, at the mouth of the Copper River, the Kaltahoni, high up the stream, and the Uqatents, about Mount St Elias, are all held by the present writer to be Athabaskan; not, indeed, so decidedly as the Beaver Indians, the Dog-ribs, or the proper Chilcotins, but still Athabaskan.

The most southerly of the Athabaskans are the Sussees, in N. latitude 51°; there are therabouts. But the Sussees, far S. as they lie, are the most northern of the Athabaskans en masse. There are relics of the stock as far S. as the southern parts of Oregon. More than this, there are Athabaskans in California, New Mexico, and Sonora.

Few discoveries respecting the distribution of languages are more interesting than one made by Mr Hale, to the effect that the Umkwa, Kwaliokwa, and Tlatakanai dialects of a district so far S. as the mouth of the River Columbia, and the upper portion of the Umkwa River (further S.), Language, still, were outlying members of the Athabaskan stock—a stock pre-eminently northern, not to say arctic, in its main area.

Yet the dialects just named were shown, by a subsequent discovery of Professor Turner's, to be only penultimate ramifications of their stock, inasmuch as further S., and further S. still, in California, New Mexico, Sonora, and even Chihuahua, as far S. as 30° N. latitude, Athabaskan forms of speech were to be found; the Navaho of Utah and New Mexico, the Jecorilla of New Mexico, and the Apatch of New Mexico, California, and Sonora, being Athabaskan. The Hoopah of California is also Athabaskan.

The Kotum of Sitka, &c., can be made Athabaskan by raising the value of the class, or it can stand as the sample of a separate and allied group.

The first of the populations to the S. of the Athabaskan area, who, lying on or to the W. of the Rocky Mountains, are other than Algonkin, are—

The Kitunama.—The Kitunama, Kutani, Cootanie, or Flatbow area is long rather than broad, and it follows the line of the Rocky Mountains between 52° and 48° N. latitude.

The Kutani differs notably from the tongues with which it is in geographical contact; though, like all the languages of America, it has numerous miscellaneous affinities.

The Atna Group.—W. of the Kutani, and S. of the Takulli Athabaskans, lie the northermost members of a great family, which extends as far S. as the Sahaptin frontier, the Sahaptin being a family of Southern, or American Origin.

The Atna, called by Mr Hale, Tahlaik-Selah languages, reach the sea in the parts to the S. of the mouth of Frazer's River, i.e., the parts opposite Vancouver's Island; perhaps they touch it further to the N. also; perhaps, too, some of the Takulli forms of the speech further N. still reach the sea. The current statements, however, are to the effect that to the S. of the parts opposite Sitka, and to the N. of the parts opposite Vancouver's Island, the two families in question are separated from the Pacific by a narrow strip of separate languages—separate and but imperfectly known. These are, beginning from the N.,—

The Haida Group of Languages, spoken by the Skittegats, Massett, Kunashinas, and Kyganis of Queen Charlotte's Islands and the Prince of Wales Archipelago. Its area lies immediately to that of the S. of the so-called Kolush languages.

The Chemesyan, spoken along the sea-coast and islands of N. latitude 55°.

The Billechula, spoken at the mouth of Salmon River; a language to which I have shown elsewhere that a vocabulary, from Mackenzie's Travels, of the dialect spoken at Friendly Village, was referable.

The Hailtsa.—The Hailtsa contains the dialects of the sea-coast between Hawkesbury Island and Broughton's Archipelago; also those of the northern part of Vancouver's Island.

In Gallatin, the Chemesyan, Billechula, and Hailtsa are all thrown in a group, called Naas. The Billechula numerals are certainly the same as the Hailtsa; the remainder of the vocabulary being unlike, though not altogether destitute of coincidence. The Chemesyan is more outlying still. I do not, however, in thus separating these three languages, absolutely deny the validity of the Naas family. I only imagine that if it really contain languages so different as the Chemesyan and Hailtsa, it may also contain the Haida and other groups; e.g., the ones that come next, or

The Wakash of Quadra and Vancouver's Island.

The Tshinak, or Chinal.

The Kalapuya.

The Jaron; all agreeing in the harshness of their phonesis, and (so doing) contrasted with—

The Sahaptin, and

The Shoshoni.

The Sahaptin is separated by Gallatin from the Waialatpu, containing the Cayus or Moleld form of speech. The present writer throws them both into the same group.

The Shoahoni and Sahaptin languages are as remarkable for the apparent ease and simplicity of their phonesis as the Jakon, Kalapuya, and Tshinak are for the opposite qualities. Language. South of the Cavis, Wailatpu, and Whinast, or Western Shoshonis, come—

(a.) The Lutuami; (b.) the Palaii; (c.) the Shasti, thrown by Gallatin into three separate classes. They are, without doubt, mutually unintelligible. Nevertheless, they cannot be very widely separated.

The Lutuami seems somewhat the most Sahaptin of the three, and this is what we expect from its geographical position, it being contiguous with the Molele (or Cayda) and the allied Wailatpu. It is also contiguous with the Whinast, Shoshoni, or Paduca, as is the Palaii. Both Palaii and Lutuami (along with the Shasti) have Shoshoni affinities.

Of three languages spoken in the N. of California, and mentioned in Schoolcraft by name, though not given in specimen—(1), the Watshehwa; (2), the Howetech; and (3), the Nabilite, the first of which is said to be that of the Shasti bands.

Of the Howetech I can say nothing.

The Nabilite is probably the language of the Tototume; at least, Rogers's River is its locality, and the Rascal Indians is an English name for the Tototume.

South of the Shasti and Lutuami areas we find—

1. The Ennik. 2. The Tahlewal.

The junction of the Rivers Klamath and Trinity gives us the locality for—

The Languages akin to the Weitspeck.—The Weitspeck itself is spoken at the junction, but its dialects of the Weyot and Wishoek extend far into Humboldt county, where they are probably the prevailing forms of speech, being used on the Mad River, and the parts about Cape Mendocino.

The Weyot and Wishoek are more dialects of the same language. From the Weitspeck they differ much more than they do from each other. It is in the names of the parts of the body where the chief resemblances lie.

The Mendocino (?) Group.—This is the name suggested for the Choweshak, Batemadaii, Kulamapo, Yukai, and Khaw-kalamayo forms of speech collectively.

1, 2. The Choweshak and Batemadaii are spoken on Eel River, and in the direction of the southern branches of the Weitspeck group, with which they have affinities.

3, 4, 5. The Kulamapo is spoken about Clear Lake, the Yakes on Russian River. These forms of speech, closely allied to each other, are also allied to the Khwaklamayo.

We may now turn to the drainage of the Sacramento and the parts S. of the Shasti area. Here we shall find three vocabularies, of which the chief is called—

The Coren.—How far this will eventually turn out to be a convenient name for the group (or how far the group itself will be real) is uncertain. A vocabulary in Gallatin from the Upper Sacramento, and one from Mag Readings (in the S. of Shasti county), belong to the group.

The Pujuni.—Concerning this we have a notice in Hale to the effect that, about 80 or 100 miles from its mouth, the River Sacramento formed a division between two languages, one using mozi, the other kik = water.

The Pujuni, &c., say mozi, as did the speakers of the Chepo.

For this group we have—(a.) Pujuni, (b.) Sesumne, and (c.) Tumumne. There is also the Cushna vocabulary, from the country Yuba, the Cushna numerals, as well as other words, being nearly the same as the Sesumne.

The Moquahumne Group.—The name Moquahumne is proposed, inasmuch as we find a Moquahumne Hill (in Calaveras county), and a Moquahumne River within the area over which the languages belonging to it are spoken. Again, the names of the tribes that speak them end largely in mane, Chupumane, &c. As far S. as Tuolumne county the language belongs to this division.

These make a provisional division for a vocabulary called—

The Costano.—The tribes under the supervision of the Mission of Dolores were five in number—the Ahwastes, the Ohone, or Costanos of the coast, the Altahmos, the Talomes, and the Romonosha.

The Mariposa Languages.—In the N. of Mariposa county, and not far S. of the Tuolumne area, the language seems changed, and the Cocoomons is spoken by some bands on the Merced River, under a chief named Nuella. They are said to be the remnants of three distinct bands, each with its own distinct language.

The Salinas Group.—This is a name which I propose for a group of considerable compass, and one which contains more than one mutually unintelligible form of speech. It is taken from the River Salinas, the drainage of which lies in the counties of Monterey and San Luis Obispo.

The Ruslen, Ealen, San Antonio, and San Miguel of this group are probably four mutually unintelligible languages.

The Salinas languages are succeeded to the S. by the forms of speech of—

The Santa Barbara Group, containing the Santa Barbara, Santa Izex, and San Luis Obispo languages.

The Capistrano Group.—Capistrano is a name suggested by that of the Mission of San Juan Capistrano. The group, I think, falls into two divisions—

1. The Proper Capistrano, or Nctela, of San Luis Rey and San Juan Capistrano.

2. The San Gabriel, or Kij, of San Gabriel and San Fernando.

The Yuma Languages.—At the junction of the Gila and Colorado stands Fort Yuma, in the district of the Yuma Indians. They occupy each side of the Colorado, both above and below its junction with the Gila. How far they extend northward is unknown, probably more than 100 miles. They are also called Cuchans, and are a fierce predatory nation, encroaching equally on tribes of their own language and on aliens.

From these Yuma Indians I take the name for the group, now under notice. It contains, besides the Yuma Proper, the Diegrano of San Diego, the Coco-maricopas, and the Cochimi (succeeded by the Walker and Porcupine of Old California).

The Coco-maricopa Indians are joint occupants of certain villages on the Gila, the population with which they are associated being Pima. Alike in other respects, the Pimas and Coco-maricopas Indians differ in language.

That the Pima group contains the Pima Proper, the Opata, and the Budwe, may be seen from the Mihidristes. That the language of the Papagos, or Papago-cadam, is also Pima, rests upon good external evidence. Whether the speech of the Ciris, and population of the island of Tiburon and the parts opposite, be also Pima is at present uncertain.

W. of the Pima lies the Tarahumara, and S. of it the Hiaqui, succeeded by the Tubar and Cora of Sinaloa.

The Paternosters of these four languages, compared with the Opata dialect of the Pima, suggest the likelihood of a closer relationship than is usually admitted.

With these end our data, but not our lists of dialects. The names Maya, Guazavo, Heria, Sicuraba, Xixime, Topia, Tepeguana, and Acaxee, all being either in Herivas, or elsewhere, as applied to the different forms of speech of Sonora and Sinaloa, to which may be added the Tahu, the Pacasa, and the Acaxee, which is probably the same word as Acaxee, as Huima is the same as Yuma, and Zaque as Hiaqui. Of the Guazavo a particular dialect is named as the Abome. Add also the Zoe and Huitcole, probably the same as the Huite.

That some of these unrepresented forms of speech belong to the same class with the Pima, Hiaqui, &c., is nearly certain. How many, however, do so is another question; it may be that all are in the same predicament; it may be only a few.

The languages of Mechoacan are—

1. The Pirinda. 2. The Tarasca. 3. The Oromi.

More has been written on the Otomi than any other language of these parts, the proper Mexican not excepted. It was observed by Naxera that it was monosyllabic rather than polysyllabic, as many of the American languages are, with somewhat doubtful propriety, demonstrated. A Mexican language, with a Chinese characteristic, could scarcely fail to suggest comparisons. Hence the first operation on the Otomi was to disconnect it from the languages of the new, and to connect it with those of the old world. With his accustomed caution, Gallatin satisfies himself with stating that others have said, his own opinion evidently being that the relation to the Chinese was one of analogy rather than affinity.

Doubtless this is the sounder view, and one confirmed by three series of comparisons made by the present writer.

The first shows that the Otomi, as compared with the mo- Language, monosyllabic languages of Asia or Mexico, has several words in common. But the second qualifies our inferences by showing that the Maya, a language more distant from China than the Otomi, and by no means insensibly monosyllabic in its structure, has, there or thereabouts, as many. The third forbids any separation of the Otomi from the other languages of America, by showing that it has the ordinary amount of miscellaneous affinities.

In respect to the Chinese, &c., the real question is, not whether it has so many affinities with the Otomi, but whether it has more affinities with the Otomi than with the Maya or any other American language,—a matter which we must not investigate without remembering that some difference in favour of the Otomi is to be expected, inasmuch as two languages with short or monosyllabic words will, from the very fact of the shortness and simplicity of their constituent elements, have more words alike than two polysyllabic forms of speech.

The fact, however, which most affects the place of the Otomi language is the monosyllabic character of other American languages; e.g., the Athabaskan and the Atakapa.

Utah and New Mexico.—The comparative civilization of the Pueblo Indians of New Mexico has always attracted the attention of the ethnologist. Until lately, however, he had but a minimum amount of trustworthy information concerning either their habits or their language. He has now a fair amount of data for both. For philological purposes he has vocabularies for six (probably for all) of them.

Of the Pueblo languages two belong to the drainage of the Rio Colorado, and four to that of the Rio Grande. Of these two divisions the former lies the farthest W., and of the two Colorado Pueblos, the most western is that of—

The Moqui.—The Moqui vocabulary was procured by Lieut. Simpson from a Moqui Indian who happened to be at Chelly.

The Zuni country lies in 35° N. latitude, to the S. and E. of the Moqui, and is probably divided by the Sierra de Zuni—from—

The Acoma, or Laguna, the most southern of the Pueblos of the Rio Grande. N. of the Acoma area lies that of—

The Jemez, on the San Josef.

The two that still stand over lie on the main stream of the Rio Grande itself. They are—

The Tesuque; and

The Taos or Picuri.—The northern boundaries of the Tesuque seem to be the southern ones of Taos. Connect these Pueblos with the town of Taos, and the Tesuque with Santa Fe, and the ordinary maps give us the geography.

The philological affinities of the Pueblo languages scarcely coincide with the geographical relations. The Moqui lies far W. Laying the emphasis on the question, the three that most strike the eye in tables, as agreeing with each other, are the Laguna, the Jemez, and the Tesuque. The other two that thus outwardly agree are the Taos and the Zuni,—two that are not in the most immediate geographical juxtaposition.

The Moqui, which is not to be separated from the other Pueblo languages, has, out of twenty-one words compared, eight coinciding with the Utah.

Neither are there wanting words common to the Pueblo languages and those of the Athabaskan Navahos, Jecorillas and Apaches.

It is convenient, in a notice of the languages of the state of Texas, to bear in mind its early, as well as its present, relations to the United States. In a country where the spread of the population from the other portions of the Union has been so rapid, and where the occupancy is so complete, we are prepared to expect but a small proportion of aborigines. And such, upon the whole, is the case. The displacement of the Indian tribes of Texas has been great. Even, however, when Mexican, Texas was not in the category of the older and more original portions of Mexico. It was not brought under the regime of the missionaries, as we may see by turning to that portion of the Mithridates which treats of the parts W. of the Mississippi. The references here are to Dapratz, to Lewis and Clarke, to Charlevoix, to French and English writers, rather than to the great authority for the other parts of Spanish America—Hervas. And the information is less precise and complete. All this is because Texas, in the earlier part of its history, was, in respect to its exploration and description, a part of Louisiana (and, as such, French) rather than a part of Mexico (and as such, Spanish).

We notice in Texas, in the Mithridates, taken along with our subsequent data, are to the effect that (a.) the Caddo, (b.) the Adai or Adahi, (c.) the Attacapa, and (d.) the Choctaw, are the prevailing languages of Texas; to which may be added a few others of minor importance.

The details as to the distribution of the subordinate forms of speech over these four leading languages are as follows:—

(a.) The Nandakoes, Nabadaches, Alisch (or Eyish), and Ini, or Tachi, are expressly stated to be Caddo; and, as it is from the name of the last of these that the word Texas is derived, we have satisfactory evidence that some members, at least, of the Caddo family are truly and originally Texian.

(b.) The Yatasi, Natchitoches, Adai or Adahi, Nacogdoches, and Keyes, belong to the Caddo confederacy, but without speaking the Caddo language.

(c.) The Caranconas, the Attacapas, the Apelunas, and the Mayes, speak dialects of the same language.

(d.) The Tunics speak the same language as the Choctaws.

Concerning the philology of the Washas, the Bedies, the Accosetaws, and the Canees no statements are made.

It is obvious that the information supplied by the Mithridates is measured by the extent of our knowledge of the four languages to which it refers.

Of these, the Choctaw, which Adelung calls the Mobilian, is the only one for which the Mithridates itself supplies, or could supply, specimens; the other three being unrepresented by any sample whatever. Hence, to say that the Tachi was Caddo, that the Yatasi was Adahi, or that the Caranconas was Attacapa, was to give an instance, in the way of explanation, of the obscurem per obscurem. Since the publication of the Mithridates, however, we have got samples of all three—Caddo, Adahi, and Attacapa.

The tendency of the Mithridates is to give prominence to the Caddo, Attakapa, and Adahi tongues, and to incline the investigator, when dealing with the other forms of speech, to ask how far they are connected with one of these three. The tendency of the later writers is to give prominence to the Cumanche, and to suggest the question: How far is this (or that) form of speech Cumanche, or other than Cumanche?

A Washita vocabulary, which will be referred to in the sequel, concludes the list of Texian languages known by specimens.

At present, the chief question respecting the philology of Texas is one of distribution. Given, as centres to certain groups—

1. The Choctaw, 2. The Caddo, 3. The Adahi, 4. The Attakapa, 5. The Cumanche, and 6. The Washita languages—

how do we arrange the tribes just enumerated? Two works help us here—1. A Letter from the Ex-President Burnett to Schoolcraft on the Indians of Texas; date 1847. 2. A Statistical Notice of the same by Jesse Stem; date 1831.

Stem's statistics run thus:—

| TRIBES | NUMBERS | |--------|---------| | Tomacarros | 141 | | Wacos | 114 | | Ketchies | 38 | | Towacans | 1,152 | | Wichitas | 100 |

| TRIBES | NUMBERS | |--------|---------| | Caddos | 161 | | Andarcos | 292 | | Ioni | 113 | | Lipans | 560 | | Cumanche | 20,000 |

giving us several of the names that have already appeared; giving also great prominence to the Cumanche—numerically at least.

In Mr. Barnett's Letter the term Caddo is prominent; but whether it denote the Caddo language, or merely the Caddo confederation, is uncertain. Neither can I find from the context whether the statements respecting the Indians of the Caddo connection (for this is what we must call it at present), are made on the personal authority of the writer, or whether they are taken, either directly or indirectly, from the Mithridates. The term that Burnett uses is stock, his statement being that the Waco, the Towacan, the Towish, the Aynic, the San Pedro Indians, the Nabaducho, and the Nacodochets, are all both Texian in origin and Caddo in stock.

His other tribes are— Language.

1. The Ketchi, a small tribe on Trinity River, hated by the Cumanchees as sorcerers, and, perhaps, the same as—

2. The Hitchi, once a distinct tribe, now assimilated with their neighbours.

3. The Tonkawas, a separate tribe, of which, however, the distinctive characteristics are not stated.

Whatever may be the exact details of the languages, dialects, and subdialects of Texas, the general outline is simple. The Choctaw forms of speech are anything but native. They are of foreign origin and recent introduction. So are certain Sioux and other dialects spoken within the Texian area.

The Cumanche is in the same predicament; though not, perhaps, so decidedly. It belongs to the Paleo class, and its affinities are with the Shoshoni and Wintuan of Oregon.

The Caddo Proper is said to be intrusive, having been introduced so late as 1519 from the parts between the great Raft and the Natchitoches, or Red River. I hold, however, that some Caddo forms of speech must be indigenous.

The Wichita is probably one of these.

The Adahi has already been noticed as being a comparatively isolated language, but, nevertheless, a language with numerous miscellaneous affinities.

The Attacapa is one of the pauro-syllabic languages of America, by which I mean languages that, if not monosyllabic after the fashion of the languages of south-eastern Asia, have the appearance of being so. They form a remarkable class, but it is doubtful whether they form a natural one; i.e., whether they are more closely connected with each other in the other elements of philological affinity than they are with the tongues not so characterized. They deserve, however, what cannot be given in the present paper, a special consideration.

For the north-eastern district of Mexico, New Leon, Tamaulipas, &c., i.e., for the parts between the Rio Grande and Tampico, no language is known to us by specimens. It is only known that the Cumanche dips deeply into Mexico. So does the Apach.

A tribe, lately mentioned, that of the Lipans, is, perhaps, Apach. Barnett states that they agree with the Mescalero and Seraties of the parts about the Paso del Norte. For these, however, we still want vocabularies its nominibus.

For the parts about Tampico the language belongs to the Huasteca branch of—

The Maya.—The Maya succeeds the languages just enumerated on the east. On the west, the Otomi, Pirinda, and Tarascos succeeded by—

The Mexican Proper.

The languages that, from their comparative obscurity, claim the attention of the investigator, are those which are other than Maya, and other than Mexican Proper.

Of these, the first succeeds the Huastecos of Huastecanapán, or the parts about Tampico; which it separates, or helps to separate, from the northern branches of the Maya Proper, being—

The Totonaca of Vera Cruz.

Cross the watershed from Vera Par to Oaxaca, and you come to the area of—

The Mixteca.—In the ordinary maps, Tepezcolula, on the boundaries of Oaxaca and Puebla, is the locality for its chief dialect, of which there are several.

The Mixteca is the language of Northern Oaxaca.

The Zapoteca that of Southern Oaxaca.

South of the areas of the three languages just enumerated comes the main division of the Maya—the Maya of Guatemala and Yucatan, as opposed to the Huasteca of the parts about Tampico. This, however, we pass over siaco pede, for—

The Lencas of Honduras, San Salvador, &c.; these being from the four Pueblos of Guajiquiro, Opatoro, Intibeken, and Sirsimton, those of the last being shorter and less complete than the others. They are quite recent, and are to be found only in the Spanish edition of Mr Squier's Notes on Central America. The English is without them.

As Mr Squier is the sole authority for the Lencas of San Salvador and Honduras, so he is for Nicaragua, for which we have specimens of—

1. The Chorotega; 2. The Nagrande; and 3. The Wulwa, of the Chontal district.

The language of the Mosquito country gives us a fourth form of speech; at least (I think) as different from the Chorotega, Nagrande, Wulwa, and Lencas, as they are from each other.

This is—

The Waikna of the Indians of the coast, and, probably, of several allied tribes inland.

In Costa Rica, a Talamanca; in Veragua, a Bayano, a Savanese, and a Chalo vocabulary carry us into South America, which must not, however, be entered before the languages of North America, to the E. of the Rocky Mountains, have been noticed.

S. of the Eskimo, and S. and E. of the Athabaskan group, come the great class of the languages called—

Algonkin, falling into the following (more or less) provisional divisions:—

1. The Bethuck of Newfoundland; 2. The Central Group, containing the Cree, Skodji, Micmac, Esquimaux, Abnaki, Algonkin Proper, Ojibwa, Delaware, Pequot, Mohican, Narraganset, Natick, Massachusetts, Mississaugi, Pamiocoghi, Menomini, Sank, Potowatomie, &c.; 3. The Skegwan and Arapaho; 4. The Blackfoot.

The vast area of this class extends from Labrador to North Carolina. It surrounds that of the—

Iroquois—the Mohawk, Oneida, Onondago, Seneca, Cayuga, Huron, Nottoway, and Tuscarora, &c.

The Sioux.—Winnebago, Dakota (with divisions and subdivisions), Assiniboin, Upsaroka (Crow), Mandan, Minetari, Osage. The Sioux area lies chiefly between the Mississippi and the Rocky Mountains.

The Woccon (with which associate the Catawba) of Carolina.

The Cherokee of Tennessee, &c.

The Choctaw, Mescogee, and Creek of Florida, &c.

The Caddo (see notice of Texas) have all Sioux and Iroquois affinities; as have also—

The Riccarie and Pawnee of Kansas and Nebraska.

In Alabama, Mississippi, and Louisiana.

The Natchez, Uche, and Chelomacha, have miscellaneous affinities, but still have to be definitely fixed.

They are succeeded by the Adahi and Attacapa, already noticed.

In South America we begin with a very obscure area. In the Mithridates there is no specimen of any sufficient length of any of the languages of New Grenada, of which the chief was called the Muyuka. Neither, until lately, did the present writer know of any vocabularies. He wrongly imagined that all the Indian forms of speech were extinct, and has erroneously stated as much in his work on the Varieties of Man. This opinion he now corrects. The Bibliotheca Glottica of Ludwig gives the following data. For (see sub. sec.)—

The Andagui of the territory of Mocoa, a vocabulary taken in 1854 by the Presbyter Manuel Marea Albis.

Notices of some words of the Cocomeos, Polindaras, and Guambas, in the province of Cundinamarca.

Vocabulary of the Corregganjo, spoken in the territory of Mocoa.

Do. of the Guanque, ibid.

Do. of Inganase, ibid.

The Muyuka Proper, or Chilcha, nearly extinct, is spoken in the neighbourhood of Santa Fé de Bogota. The chief dictionary of it is in MSS. (See Ludwig, ad voc. Muyuka).

To the south the languages of New Grenada are succeeded by—

The Quichua of Bolivia and Peru, the language of the Incas, spoken from the equator to S. Lat. 28°. The Quichua area almost surrounds that of—

The Aymara of the parts about the Lake Titicaca, to which it is allied.

The Quichua is spoken as far inland as Tocuman; but over a great portion of its area it is an intrusive and foreign language, a fact which lends us to the consideration of the minor, and less known language, within its area, or on its precincts. These, though probably numerous, are little known by specimen. There is one of the Zapaha language of the province of Mainas, which looks like an independent form of speech.

There is also the—

Yurakares of the eastern slope of the Andes, in 17° S. Lat.

In Venezuela and Demerara, the great— Language.

Canin family takes the same prominence as the Quichua in Peru. To it belong the Chayma, Tamanak, Arawak, Accawa, Macusi, Areccuna, Malongkong, Guiana, Soerikong, Wayawara, Maopetyan, Tiverighatto, Pianogkotto, and most of the numerous languages whose names end in ghto.

Akin to the Carib are several of the languages of the Orinoco, Rio Negro, and northern bank of the Amazon, of which we know by specimens the—

Saliva, Maypure, Yarura, and Ottomaca groups, of undetermined value.

With Carib and Maypure affinities, in proportions hitherto unascertained, we find on the northern affluents of the Amazon, the Usinamben, Tariana, Isanna, Barre, and Tomo-Maroa forms of speech; and somewhat more isolated (though with miscellaneous affinities), the Juri, the Javita, the Coreto, Cobou, and Tucano, known from vocabularies in Mr Wallace's work on the Amazon and Rio Negro.

In a similar state of isolation, as compared with the definite and undoubted members of the Carib family (though still with miscellaneous affinities), are the Warow, Taruma, and Wapi-siana of Demerara.

In Brazil there is the great family of the Guarani tongues, and, by the side of it, the analogues of the minor languages of the Inca and Carib districts. Thus, of the Brazilian languages

Other than Guarani we know, by specimens, the Botocudo, Corope, Coroado, Puri, Machacalli, Malali, Camacum, Meneng, Cachriaba, Kiriri, Sabujah, Timbira, and others. Of—

The Guarani, the Guarani Proper (the basis of the Lingua Geral, Lingua Franca), the Chiriguano, the Omagua, the Mundruce. The Omagua is spoken on the Napo, the Chiriguano on the La Plata; so wide is the vast Guarani area. (See Ethnology.)

Passing over large tracts of terra incognita, in the Missions of Mooxos and Chiquitos, the Moxo, Movima, Cayuvava, Saraboca, Corovexa, Pacoconca, Curumimaca, and Palocomeca—forms of speech, known by specimens, with miscellaneous affinities, and leading to the language of the Chaco, which form a class of considerable complexity called—

Abiponian—Abiponan, Papier, Mbocobi, Tobu, Payaguá (?) Lenguas, and Zamuco (?). The exact relations of these to the Lule is uncertain. We approach, however, another group.

The Chileno-Patagonian—The Araucanian (Chileno Proper), the Pampa (?), the Patagonian, and the Fuegian.

A recent and valuable contribution to the philological ethnography of America, by Mr Daa of Christiania, besides giving a comparison of the Samoyed of Asia with the Dakota of America (along with much other subsidiary matter), contains more than one suggestion respecting the forces by which changes in language are effected. Great stress is laid upon the size of the community that uses it. "When language is confined to the daily use of a family, or a small knot of acquaintance, it stands in a quite contrary relation to the use of men, than when it is the common medium that combines millions of human beings. In the last case, the individual license in changing the adopted sounds and significations of words, whereby are introduced novelties of speech, is continually checked by the impossibility of making all such unnecessary changes comprehensible to the mass of those who speak. Thus we see that in the present English and French languages, this license of adding to what is the common property of millions in both hemispheres, is a privilege for only a few distinguished inventors of new things, or authors of widely-read books. The power of changing language is so much repressed, that it can only be observed by comparing two remote periods of the history of the language, as you observe the geological changes by considering generations as merely a single day. The habit of speaking distinctly is then kept up and cultivated as a necessary means to be comprehended by the many unknown persons you continually meet with.

"In a small island in the South Sea, or an insignificant tribe in the wildernesses of America or Siberia, the facility of changing language may easily be conceived to be next to unlimited. Everybody who speaks must become understood, because his hearers almost know beforehand what he is to say. The most arbitrary changes of languages are thus introduced continually, as may be proved historically.

"Almost all those languages that are spoken by nations living either in a natural (geographical) isolation, or in an arbitrary and artificial one, want a great number of letters. For one letter in one dialect is substituted another letter in the next tribe; because every word is as well understood whether you pronounce it with the letter r, or l, or v. Accidental and individual defects of uttering are thus changed into national peculiarities, and a general indistinctness of pronunciation is introduced. The sounds that are hardly perceptible to a stranger will, among close relatives, appear sufficiently intelligible.

"The strange practices of mutilating the nose and the lips must have contributed a great deal to disfigure the enunciation of language itself. The insertion of one or more large pieces of wood into incisions in the lips or the nose, still practised on the Pacific coast, and from which custom a tribe is called Nas Perce, has no doubt been more frequent formerly. These mutilations would evidently make it next to impossible to pronounce any labial consonant, and they would in return introduce a nasal articulation. Now a paucity of labial, and a superfluity of nasal sounds, is just what we observe in many American languages.

"Similar permutations of letters of course happen among all languages of the world, and in fact form the basis and the principal means by which the differences in language are introduced and produced. Only among those nations who lead an isolated life are these changes more violent, and appear to separate tribes that evidently, from their general habits and manners, must be very closely related. Thus the Dakotas, forming only a nation of 25,000 individuals, are split into tribes divided by such considerable differences of dialect as these—one tribe changes d into t, and t into r; another changes k into k', a third changes h into g; d is altogether rejected, and t substituted in its place; another band of Dakotas g at the end of syllables, and l does not occur; thus, the word holok, 'to go home,' becomes holok and glol. Different dialects."

In the syllabic alphabet invented for the Cherokee tongue by a native, the whole number of possible syllables is merely seventy, including the vowels. In the excellent Dakota Dictionary of Mr Riggs we find abundant proof how a scarcity of radical words and simple ideas is made to expand into a language of endless compositions. But, from the variety of objects to be expressed, these composed words in a great measure must contain the most arbitrary descriptions of things: a continual make-shift of substitutes for the thing that is thus obscurely brought before the mind. For instance, the word nezgo means 'metal of any kind,' gadii 'merchandise;' hence are derived, by addition of other substantives, or adjectives and particles, compositions expressing an anchor, iron pot, bracelet, bell, trap, chair, gun and all its parts, pistol, cannon, lock, ransomed, &c., nail, steelyard, blacksmith, spade, finger-ring, store, skates, sword, iron, silver, money, dollar, shilling, bank-note, medal, gold, lead, bullet, moulds, copper, peester, button, spoon, pan, brass, file, hammer, pincers, tongs. Like manner the syllable to comprehends all ruminating animals and their parts.

"The state of small isolated tribes or clans in which the half-savage nations live will as easily introduce an endless change of significations. In a family, or amongst the inmates of the same house, it is quite as easy to make arbitrary expressions or slang words understood and ultimately accepted of as an indistinct utterance of the common words. Instead of father, you may say master, governor, husband, the old one, and the original word father you may restrict to God only; instead of child you may use any word signifying little or dear, &c. We have special accounts of two remarkable instances of the action of this principle among the rude tribes. One is the superstitions custom of the South Sea islanders, on the death of a king whose name is composed of a couple of common words, to abstain altogether from the use of those words that form his name, and to substitute others. The practice is either ascribed to a reverence for him, or to some religious sentiment connected with omens. Such a custom will, of course, in many instances, lead to a permanent instead of a temporary change of language. The other fact upon this head is the sacred languages employed by the conjurors or priests. As far as this has been accurately found out, for instance, in the Greenlandic, it seems to be chiefly an arbitrary perver- Language, sign of the significations of old and known words. It is then the same principle as in Europe has formed any slang, for instance, among vagrants and thieves. Yet these words of the conjurers have been so far altered that any double meaning is sufficiently avoided.

"From the effect of these causes it appears probable that as one savage tribe may, from trifling occasions, suddenly split into two, that separate widely from each other, thus also their languages, in comparatively short time, deviate into two very different dialects. If there were means of investigating the state of a given language at stages in different periods, it would perhaps be demonstrable, that its formation as a peculiar dialect, or a variety of speech, does not require thousands of years, as one might suppose, who starts from the fact that a great many Greek and Hebrew words have been preserved uncorrupted for thousands of years, through the influence of literature and civilization."

ON THE LANGUAGES OF AFRICA.

Our data for the investigation of the African languages are greater than we expect d priori. Many unexplored districts are known in respect to their philology. This is because missionaries, and other investigators, have taken down the accounts of slaves transported from the more inaccessible portions of the continent to either localities on the coast, or to America.

The general distribution of the African languages is that of the languages of the rest of the world, viz., irregular in respect to the different areas of the different tongues. This means that sometimes a single form of speech, with either a paucity of dialects and sub-dialects, or else with a slight amount of difference between its extreme forms, is spread over a large extent of country; whereas, on the other hand, a small tract of land shall contain several mutually unintelligible tongues. The Berber, the Fulah, and the Kaffir tongues are remarkable for the ground they cover; all the parts S. of the equator, with the exception of the Hottentot, being Kaffir, and all the parts N. of Negroland, with the exception of the Arabic and Tibboos districts, being Berber.

The small areas with the greatest number of mutually unintelligible tongues, are the alluvial tracts between the Gambia and Cape Mount, i.e., the deltas and valleys of the Casamance, Rio Grande, Nunez, &c., then certain districts in Abyssinia. The areas of moderate or intermediate magnitude are those of the Hansa, Woloff, Mandingo, &c.

The evidence that all the languages of Africa are related to each other is in the mind of the present writer, though not, perhaps, to the majority of investigators conclusive. This rests upon two series of facts—the accumulation of new materials and the breaking down of distinctions which at first sight appear more definite and trenchant than they really are.

This latter process has taken place most especially with the tongues of the extreme N. and the extreme S.

To begin with the former. The languages akin to the Hebrew and Arabic—Semitic as they are called—were long either isolated, or, if connected with those of any other class, connected with the so-called Indo-European forms of speech. This was on the strength of the higher civilization, greater historical importance, and superior physical organization of the nations which spoke them. Writers, however, were not slow to observe that the populations of northern Africa in general were, to a great extent, possessed of the same characteristics. Such were the Egyptians, whose language was the Coptic, a language which was one of the first to be recognized as one exhibiting Semitic characteristics. This was not doubted. It was only doubted whether the Coptic was, in the ordinary sense of the word, African.

Then came a language to which the French conquest of Algeria gave prominence, whilst it also made it accessible; the language of the Kabyles, Tuaregs, Siwans, and Canary Islanders. This was recognized, if not as actually Semitic, as what was designated by the new term, sub-Semitic.

That other tongues, especially those in geographical contact with the Kabyle (Berber or Amazigh), were, more or less, what the Berber was, was shown by even the Berber scholars; the foremost of whom recognized, in the Hansa of Sudania, Berber elements. (See a paper by Francis Newman in the Appendix to Prichard's Physical History of Man.)

Then came the turn for the tongues to the south of the Language. Coptic area to be considered as, more or less, Coptic; e.g., the Bisharye, the Nubian, and the Galla; and, finally, that for the languages of Abyssinia in contact with the recognized Semitic tongues (such as the Tigre, Amharic, &c.), but not themselves Semitic. Of these the Agow and Falasha forms of speech are the chief.

With this relation between the Semitic and sub-Semitic classes—a relation made patent by the name itself—the question as to the relations of the African languages at large must either remain stationary, or one of two alternatives be resorted to.

Either languages like the Hansa, Nubian, Agow, &c., must lead to the true Negro tongues, or they must be wholly separated from them. It is not too much to say that, on the part of the proper Semitic philologues, the tendency was towards separation. This, however, was impossible. Whoever knew anything of the other African languages knew that for every step from such languages as the Coptic and Berber, towards the Hebrew and Arabic, a similar advance could be made in the opposite direction, i.e., towards the Fellatah, Mandingo, and Woloff, and through these to the most Negro languages of the whole continent.

On the South there are the same apparent differences, and the same abrogation of them; two languages coming more especially under notice—

1. The Hottentot.—It is doubtful whether the presumed isolation of the Hottentot forms of speech rests on purely philologic grounds; indeed, it is pretty certain that it does not. The same is the case with the Semitic and sub-Semitic tongues, wherein physical difference has as much to do as glossarial. The contrast between the Bushman, and not only the so-called higher races in general, but his immediate neighbours, is great. At the same time there are many more transitional and intermediate forms than is generally believed, especially in the northern parts of the Hottentot area. The articulate sound described as a click, characterizes the Hottentot languages. Like most points of phonesis, its import has been overvalued. It is not limited to the Hottentot, and would not prove isolation if it were. More than one Kaffir dialect has it. Then there is a kind of gender in the Hottentot grammar; a fact which has led some able philologists to the preposterous doctrine that it has an affinity with the more cultivated tongues of Europe. It has some Kaffir affinities, too, and others (perhaps more numerous) with the languages N. of the Kaffir area—a matter that should by no means surprise us, inasmuch as the amount of displacement along the whole frontier has been great.

The comparison of the scanty specimens of a scanty language gives us the following miscellaneous Hottentot (H) affinities with the other languages of Africa (A):

| English | sun. | English | mountain. | |---------|-----|---------|----------| | H.S | t'kora. | H. Corana | t'kusa. | | Hottentot | torre. | A. Falasha | daba. | | Corana | soreb. | English | ear. | | A. Agow | guorah. | H. Corana | t'fusna. | | Shilluck | glurrek. | A. Bullom | mofusa. | | Kra | gweva. | English | star. | | Kanga | jirov. | H. Corana | kambroka. | | Wawn | jiri. | A. Kossa | rumbereki. |

| English | tongue. | English | bird. | |---------|--------|---------|------| | H. Corana | t'omma. | H. Bushman | t'kumani. | | A. Fertit | t'isna. | A. Mandingo | t'umani. |

| English | neck. | English | asleep. | |---------|------|---------|--------| | H. Bushman | t'kusa. | H. Corana | t'kusa. | | A. Darfur | kira. | Bushman | t'kusa. |

| English | hand. | English | fire. | |---------|------|---------|------| | H. Corana | t'koom. | H. Corana | t'kusa. | | A. Shilluck | kiwa. | Congo | taba. | | English | tree. | Somalil | dab. | | H. Corana | pibona. | Bushman | t'kusa. |

VOL. XIII. In respect to the so-called genders, the following table shows how little they are those of the European languages. The forms are simply compounds. They are taken from the *Abriss der Formenlehre der Namauqua-Sprache* (Barmen, 1854); the German *ich*, *du*, &c., being translated into Latin:

| Masc. | Fem. | Com. | |-------|------|-----| | Ta | Ego | Ta | | Ko | So | Da | | Khom | Tma | Uma | | Khoma | Tma | Uma |

II.

| Masc. | Fem. | Com. | |-------|------|-----| | Z | S | S | | Ko | So | Do | | Kho | Ro | Ro |

III.

| Masc. | Fem. | Com. | |-------|------|-----| | B | S | T | | Ku, ka | Ti, te | Na, na | | Kho | Ra | Ro |

The Kaffir tongues have two notable characteristics, — (1.) Alliteration; (2.) The system of prefixes.

According to the former, when two words stand in certain grammatical relations to one another, the initial letter of the language subordinate is changed to that of the governing term, just as if we said, in English, *bus-beam* instead of *sun-beam*.

According to the latter, every noun has, as its concomitant, some non-radical prefix, so necessary, that when the missionaries would introduce such English words as *priest* or *priories*, the form they gave them in Kaffir was *un-priest*, *un-priories*.

These two characteristics give very remarkable physiognomy to the language. Doing this, they have a tendency to create broad and definite lines of demarcation. Hence the separation between the Semitic tongues on the N. and the inland and western dialects, is repeated in the S. between the Kaffir and the non-Kaffir languages.

The ethnological import of the two characteristics in question has never been very closely considered. They may mean much; they may mean little.

Their absence may separate languages, otherwise like, from the Kaffir; their presence may connect languages otherwise unlike. So, at least, was the opinion of many investigators; for the fact suggested is a real one.

Both have been discovered elsewhere. The Timman is a language of the parts about Sierra Leone. It is a language of Northern Africa. That it exhibits the so-called characteristics of the Kaffir tongues of South Africa, is notified by Bishop Vidal in the preface to his edition of Crother's *Tororo Grammar*. Mr Norris had, by independent researches, arrived at the same conclusion.

The import of this coincidence is another question. That the most northern of all the Negro languages, the Wolof, exhibited a remarkable series of euphonic initial changes, had been known since the time of Dard. The Wolof changes, however, were not so Kaffir as the Timman.

But there is another series of facts. The language of all the languages of Africa, wherein the Kaffir characters are at a minimum, is the Mandingo—at least in the proper Mandingo dialect. Writers who have made everything that was Kaffir-like Kaffir, have still left the Mandingo undisturbed. The most that has been done towards even approximating it to Kaffir-like languages has been done by Mr Norris and Dr Bleek. The former has suggested the likelihood of the signs having been lost—obscure traces of them still being discoverable. The latter has, at one and the same time, pronounced the Grebo dialects to be Mandingo, in respect to the class to which they belong, and South African (Kaffir) in their structure.

But the Timman is connected by its vocabulary with the Mandingo,—not directly, but still as a member of the same class; that class being one of no inordinate magnitude, and one which does not contain the Kaffir; one, too, which the geographical conditions indicate as likely. How does this affect our classification? Much more than it ought to do. The present tendencies are to make everything Kaffir. True criticism only tells us to look twice before we allow a single characteristic to become the basis of a classification. Naturalists know this. Philologists have yet to learn it. In the mind of the present writer, the Timman, and other languages in the same category, are Kaffir, in the way that an oak is a smoke, a whale a fish, or a bat a bird.

There is another preliminary to the classification of the languages of Africa. We cannot treat them as we treat those of Europe. Everything in Europe is definite; i.e., it is Latin or Greek, Roman or Celtic, Sarmatian or Turkish,—there being no such thing as an intermediate, transitional, or equivocal form. Hence, we can classify by *definition*, separating tongue from tongue. In Africa we must classify by *type*: i.e., group certain forms of speech round certain others, without being able to separate the outlying members of one class from the outlying members of another. This makes the place of many forms of speech equivocal,—equivocal rather than actually doubtful or obscure. They may belong to two classes at once.

Thus guarded against the expectation of finding greater definitude than the nature of the inquiry will allow, we begin with the parts where Asia and Africa join, and where the (so-called) Semitic touch the (so-called)— II. The Nubian group, containing: (a) The Kensi, Noub, and Baraba of the specimens, the Dongoloway of Dongola. (b) The Koldaij of Kordovan. (c) The Shelluk, Dinka, Takei, Tumali, Shabun, and Ferit of the parts to the S. of Kordovan. (d) The Furian of Darfur, allied to—

III. The Bisharye, between the Nile and Red Sea; and,

IV. The Fazoglo and Qdnamyf of Bertat.

To the west of Darfur follow— V. The Borgo (of Wadai or Darseleh). VI. The Beqhami. VII. The Mandara. VIII. The Berari or Canowry group; which will, probably, when better known, be more closely connected with each other, and also with—

IX. The Tibou of the eastern part of the desert. To the W. of Berari lie the languages of— X. The Haussa; and— XI. The native language of the parts about Timbuctu.

XII. The Fulah groups, leading to the complex philology of the languages known to us on the coast of the Atlantic. Between the southern boundary of the desert and the equator we find, passing from N. to S.—

XIII. The Wolof of the Gambia. XIV. The Mandingo of the Senegal. XV. The Felup, &c., for the mouths of the Nunez, Rio Grande, &c. XVI. The Grebo or Kru, of the parts about Cape Palmas; passing through— XVII. Arokoem, of the Ivory Coast; to— XVIII. The Gold Coast group, containing— (a) The Fanti (Ashanti); (b) Aca (Cape Castle); (c) Whidah (of Dahomey).

Akin to which are— XIX. The Formos. XX. The Ibo (of the Lower Niger). XXI. The Nufi (of the Tshadda) leading through— XXII.-XXIII. The Efik of Old Calabar, and the Fernando Po languages, to—

XXIV. The Empeonges of the Gaboon. XXV. The Bath of Adamawa has miscellaneous affinities with the language here mentioned, and with— XXVI. The great Kaffir class, which reaches to— XXVII. The Hottentot area. On the E., Kaffir languages are spoken as far N. as (there or thereabouts) the equator, where they are contemporaneous with a transitional form of speech. XXVIII. The Ulwof; and also with— XXIX. The Galla, which leads us up to Nubia and Abyssinia. In the interior, a class of languages spoken by the negroes of Abyssinia is represented by— XXX. The Darweichegan; and— XXXI. The Tocarce vocabularies of the specimens; and further inland still, to the S.W. of Abyssinia, it find constituting a class conveniently called— XXXII. The Tonga, the Kaffa, Woratta, Wolaitta, and Fangaro forms of speech.

It is not pretended that the value of all these groups is the same. It is only submitted that there is no African language, hitherto known (and, with few exceptions, all are known), which cannot be referred to some of the previous groups. It is also added, that they have all miscellaneous affinities; in other words, there is no such a thing as isolation.

The preceding arrangement is a modification of one laid before the British Association for the Advancement of Science in 1844, and expanded (with slight modifications) in 1847, the original form being as follows. It is given for the sake of foreshadowing the more complex classification of divisions and sub-divisions of higher and lower groups, &c., which a more advanced state of our investigations will require.

1. The Coptic class, containing the extinct dialects of Egypt. 2. The Berber class, containing the non-Arabic languages of northern Africa. 3. The Hottentot class. 4. The Kaffir class, extending from the limits of the Hottentot country as far northward as Loango, and the River Juba W. and E.

5. The fifth and last class, which was left unnamed, fell into eleven subordinate groups— 1. The Nubian group. 2. The Galla group. 3. The Borgio group. 4. The Beqhami group. 5. The Berari group. 6. The Mandara group. 7. The Hawasa group. 8. The Mandingo group. 9. The Wolof group. 10. The Fulah group. 11. The Ibo-Ashanti group.

Of these groups the ten first were undivided; whilst the eleventh fell into the following sub-divisions— a. The Fante division. b. The Aca division. c. The Dahomey division. d. The Yoruba division.

Notwithstanding these sub-divisions, the following languages remained unplaced— 1. The Agow dialects. 2. The Bisharye dialects. 3. The Serawooli. 4. The Serere. 5. The Akvambas. 6. The Kra.

These last are now distributed as follows:—The first is sub-Semitic; the second allied to the Nubian and Coptic; the third and fourth Mandingo, with Fulah elements; the fifth Fanti; the sixth intermediate to the Mandingo and Fanti.

The Mandingo, Kru, Ashanti, Woloff, and Fulah will, probably, be brought by future researches more closely together.

That this classification is right in its main elements is rendered probable by the confirmation it has received from subsequent investigations. The Polyglotta Africana of Mr Kolle (published in 1854), as far as it goes—which is to the extent that specimens collected at Sierra Leone carried the writer—gives the same groups, though under different names, the investigation being (I believe) independent. Thus Kolle's— I. North-western Atlantic languages fall into four groups, represented by the (a.) Felup, (b.) Popel, (c.) Biafra, and (d.) Timmani languages, respectively, each falling into dialects and subdialects.

II. The North-western High Sudan, or Mandingo languages, constitute the second group.

III. The Upper Guinea, or Middle-coast languages, the third. This means, the forms of speech akin to (a.) the Kru, (b.) the Dahomey, (c.) the Yoruba.

IV. The North-eastern High Sudan languages are spoken inland, at the back of the Ashanti country, and along the eastern range of the Kong Mountains. They are akin to the (a.) Mossi, (b.) Kouri, (c.) Koamn, and (d.) Yula forms of speech.

V. The Niger-delta group falls into the (a.) Iosaama, (b.) Sobo, and (c.) Okoloma divisions.

VI. The Niger-Tshadda languages as those akin to the Nufi.

VII. The Central African division contains the languages allied to (a.) the Bornui, and (b.) the Pika.

In Part II. we have the South African languages distinguished by an initial inflection, as has been already stated. It excludes the Hottentot, and includes the Old Calabar, Cameroons, and Gaboro languages. Doing this, it coincides with the so-called Kaffir-class of tongues, in its widest form, i.e., in the form it has taken since it was shown that the Pongwee, the Tembe, the Efik, and others, all exhibited a similar series of initial changes to those of the Kaffir and Bichuan.

There are of course some differences of detail; the main groups, however, are the same. The first is one for which details were especially wanted. The fourth also is important. It should probably form a separate substantive division of the group called Ibo-Ashanti.

If the foregoing is what the writer believes it to be, viz., the sketch of the classification into which the languages of Africa ought to fall, it is far from being that which the very newest investigations have suggested. The very newest investigations are those of which Mr Norris, Dr Bleek, and Mr Kolle are the chief representatives; and, in a less degree, perhaps, Professor Lepsius. It is the overvaluation of the Kaffir characteristics (along with certain other points of less importance) that makes these arrangements exceptionable; and it is this against which the preliminary criticism of the previous pages is directed.

Lest it should be thought unnecessary, I subjoin the following table from Dr Bleek's paper, De Nomina Generebus:— One of the best measures of the truth of a principle, is to be found in the result to which it leads. Here we have one, which, separating the Hottentots from the Kaffirs, connects them with the Egyptians, Algerians (Berbers), Gallus, Jews, and Arabs (Semites), and, finally (though more indirectly), with the Slavonians, Germans, Greeks, and Latins (Indo-Germans). This is what it comes to when we classify on the strength of a single characteristic. The actual details of the ethnographical philology of Africa are important. The principle, however, on which investigators arrive at them is much more so.

To conclude: The preface of a work, not two months old—The Languages of the Mozambique, or Vocabularies of the Dialects of Lorenzo Marques Inhambane, Sofala, &c., from the MSS. of Dr Peters, by Dr Bleek—contains the following passage: "The languages of these vocabularies all belong to that great family which, with the exception of the Hottentot dialects, includes the whole of South Africa, and most of the tongues of Western Africa, &c. If so, and if the table be equally accurate, the affinities of the Bushman are with the Parisian, the Londoner, and the Athenian, rather than with the Negro of the Gold Coast, or Sierra Leone.

It is the old error—a bat has wings, therefore it is a bird; an oyster lives in the sea, therefore it is a fish; and so on through the whole range of possible misnomers.

On the so-called Indo-European Class.

As this stands, it contains, according to all authorities, the Celtic, the German, the Sarmatian, the Classical (the Latin and Greek), and Sanscrit groups. To these many (perhaps most) add the Armenian and Ossete; some the Albanian. In respect to the import of these terms, the ordinary views make Gothic and German nearly synonymous, and Sanscrit the name of a language of Asiatic origin, to which the Zend was allied, and out of which and its congeners some Indian and Persian forms of speech have been developed, and to which they are allied, as the Italian, &c., is to the Latin. The greater part of this is believed by the present writer to rest on an illegitimate amount of assumption.

Firstly, that the Keltic was not isolated, but allied to the German, Slavonic, &c., was shown by Dr Prichard, and excellent service he did in showing it. But that it is allied to these tongues, as they are to each other, has never been shown. Yet this is what is wanted, in order to make an addition to a class without raising its value. This, however, should never be done without consideration. The principle upon which the value of a class is raised, in order to admit one addition, may be extended so far that, by successive increases of value, it may be no class at all; as the so-called Indo-European class is in a fair way of becoming. Good reasons can be given for limiting rather than widening it. It can, of course, when thus limited, be subordinated to some other.

Secondly, in respect to the German division, a vast superstructure totters as soon as we inquire for evidence to the assumed fact of any German population whatever having been called Goth anterior to its settlement in the country of the Getae. No such evidence exists. On the contrary, the Germans were Goths as they were Britons, i.e., occupants of British soil, but no Britons at all—not yet Goths.

Thirdly, the philological evidence of the Sanscrit being of Asiatic origin is that of the Latin of Spain having been Hispanian; certainly no better. It consists in the fact of its appearing on the soil at an early (admittedly a very early) period; a literature indigenous to the Indian soil having existed in it, along with inscriptions, and a certain amount of influence upon the later language. In India, the appearance of the Sanscrit on the soil is earlier than that of the Latin in Spain. On the other hand, no one has doubted the Latin origin of the Spanish language, whilst more than one competent judge has doubted the Sanscrit origin of the Hindui, Bengalee, &c. Yet we know from history that the Latin of Spain was not only not indigenous to the Peninsula, but that it was not even derived from the next country. All that rests upon legitimate inference, in respect to the Sanscrit in India, are the facts of its early introduction and the greatness of its subsequent influence. Its Indian origin, its Persian origin, and the like, are all based upon covert, but undue, assumptions.

Akin to these are the explanations of the affinity between the Sanscrit on the one side, and the Lithuanic, Slavonic, Latin, Greek, and German on the other. The latter are deduced from Asia, i.e., five species of a genus are deduced from the scite of one, instead of the one from the scite of the other five. This is all improbable, except so far as it can be shown that the Dnieper and Danube are nearer to the Indus and Ganges than the Ganges and Indus are to the Danube and Dnieper.

The Asiatic origin of the Greek, Latin, Lithuanic, Slavonic, German, and Keltic, has been assumed at once, without any consideration whatever of the other alternative, viz., the European origin of the Sanscrit. This is over-hasty.

The recent deductions from the Ossete, and its Indo-European affinities, verify what has been said concerning the Keltic. Whatever the Ossete is, the Georgian is also; and the Georgian has been made Indo-European. But what the Georgian is, the Circassian is; which is more Chinese than Greek, more Tibetan than Latin, more Burmese than German. Classifications of this kind prove too much.

It is held that the class in question is most conveniently limited to (1.) the German; (2.) the Classical (Greek and