an instrument for measuring the degree of heat or cold in any body.

The thermometer was invented about the beginning of the 17th century; but, like many other useful inventions, it has been found impossible to ascertain to whom the honour of it belongs. Boerhaave* attributes it to Cornelius Drebbel of Alcmar, his own countryman. Fulgenzius† attributes it to his master Paul Sarpi, the great oracle of the Venetian republic; and Viviani gives the honour of it to Galileo‡. But all these are posthumous claims. Sanctorio|| claims this honour to himself; and his assertion is corroborated by Borelli§ and Malpighi ‡ of the Florentine academy, whose partiality is not to be suspected in favour of a member of the Patavian school.

Perhaps the best way to reconcile these different claims would be, to suppose that the thermometer was really invented by different persons about the same time. We know that there are certain periods in the progress of the arts when the stream of human genius runs in the same direction, and moves towards the same object. That part of the current which reaches the object first may possess the title; but the other parts follow so rapidly and arrive too soon after, that it is impossible for a spectator to decide which is first in point of time.

The first form of this instrument for measuring the degrees of heat and cold, was the air-thermometer. It is a well known fact that air expands with heat so as to occupy more space than it does when cold, and that it is condensed by cold so as to occupy less space than when warmed, and that this expansion and condensation is greater or less according to the degree of heat or cold applied. The principle then on which the air-thermometer was constructed is very simple. The air was confined in a tube by means of some coloured liquor; the liquor rose or fell according as the air became expanded or condensed. What the first form of the tube was, cannot now perhaps be well known; but the following description of the air-thermometer will fully explain its nature.

The air-thermometer consists of a glass tube BE, connected at one end with a large glass ball A, and at the other end immersed in an open vessel, or terminating in a ball DE, with a narrow orifice at D; which vessel, or ball, contains any coloured liquor that will not easily freeze. Aquafortis tinged of a fine blue colour with a solution of vitriol or copper, or spirit of wine tinged with cochineal, will answer this purpose. But the ball A must be first moderately warmed so that a part of the air contained in it may be expelled through the orifice D; and then the liquor pressed by the weight of the atmosphere will enter the ball DE, and rise, for example, to the middle of the tube at C, at a mean temperature of the weather; and in this state the liquor by its weight, and the air included in the ball A, &c., by its elasticity, will counterbalance the weight of the atmosphere. As the surrounding air becomes warmer, the air in the ball and upper part of the tube, expanding by heat, will drive the liquor into the lower ball, and consequently its surface will descend; on the contrary, as the ambient air becomes colder, that in the ball is condensed, and the liquor pressed by the weight of the atmosphere will ascend: so that the liquor in the tube will ascend or descend more or less according to the state of the air contiguous to the instrument. To the tube is affixed a scale of the same length, divided upwards and downwards from the middle C into 100 equal parts, by means of which the ascent and descent of the liquor in the tube, and consequently the variations in the cold or heat of the atmosphere, may be observed.

This instrument was extremely defective; for the air in the tube was not only affected by the heat and cold of the atmosphere, but also by its weight.

The air being found improper for measuring with accuracy the variations of heat and cold according to the form of the thermometer which was first adopted, another fluid was proposed about the middle of the 17th century by the Florentine academy. This fluid was spirit of wine, or alcohol, as it is now generally named. The alcohol being coloured, was inclosed in a very fine cylindrical glass tube previously exhausted of its air, having a hollow ball at one end A, and hermetically sealed at the other end D. The ball and tube are filled with rectified spirit of wine to a convenient height, as to C, when the weather is of a mean temperature, which may be done by inverting the tube into a vessel of flagrant coloured spirit, under a receiver of the air-pump, or in any other way. When the thermometer is properly filled, the end D is heated red hot by a lamp, and then hermetically sealed, leaving the included air of about one-third of its natural density, to prevent the air which is in the spirit from dividing it in its expansion. To the tube is applied a scale, divided from the middle, into 100 equal parts, upwards and downwards.

As spirit of wine is capable of a very considerable degree of rarefaction and condensation by heat and cold, when the heat of the atmosphere increases the spirit dilutes, and consequently rises in the tube; and when the heat decreases, the spirit descends, and the degree or quantity of the motion is shown by a scale.

The spirit of wine thermometer was not subject to its defects, some of the inconveniences which attended the air thermometer. In particular, it was not affected by variations in the weight of the atmosphere: accordingly it soon came into general use among philosophers. It was, Martinez at an early period, introduced into Britain by Mr Boyle. Ephys.

To this instrument, as then used, there are, however, many objections. The liquor was of different degrees of strength, and therefore different tubes filled with it, when exposed to the same degree of heat, would not correspond. There was also another defect: the scale which was adjusted to the thermometer did not commence at any fixed point. The highest term was adjusted to the great sunshine heats of Florence, which are too variable and undetermined; and frequently the workman formed the scale according to his own fancy. While the thermometer laboured under such disadvantages it could not be of general use.

To obtain some fixed unalterable point by which a determined scale might be discovered, to which all thermometers might be accurately adjusted, was the subject propounded by which next drew the attention of philosophers. Mr Boyle, who seems at an early period to have studied this subject with much anxiety, propounded the freezing of the essential oil of anniseeds as a convenient point for graduating thermometers; but this opinion he soon laid aside. Dr Halley next proposed that thermometers should should be graduated in a deep pit under ground, where the temperature both in winter and summer is pretty uniform; and that the point to which the spirit of wine should rise in such a subterraneous place should be the point from which the scale should commence. But this proposal was evidently attended with such inconveniences that it was soon abandoned. He made experiments on the boiling point of water, of mercury, and of spirit of wine; and he seems rather to give a preference to the spirit of wine*. He objected to the freezing of water as a fixed point, because he thought that it admitted considerable latitude.

It seems to have been referred to the all-conquering genius of Sir Isaac Newton to determine this important point, on which the accuracy and value of the thermometer depends. He chose, as fixed, those points at which water freezes and boils; the very points which the experiments of succeeding philosophers have determined to be the most fixed and convenient. Sensible of the disadvantages of spirit of wine, he tried another liquor which was homogeneous enough, capable of a considerable rarefaction, about 15 times greater than spirit of wine. This was linseed oil. It has not been observed to freeze even in very great colds, and it bears a heat about four times that of water before it boils. With these advantages it was made use of by Sir Isaac Newton, who discovered by it the comparative degree of heat for boiling water, melting wax, boiling spirit of wine, and melting tin; beyond which it does not appear that this thermometer was applied. The method he used for adjusting the scale of this oil thermometer was as follows: Supposing the bulb, when immersed in thawing snow, to contain 10,000 parts, he found the oil expand by the heat of the human body so as to take up \( \frac{3}{4} \)th more space, or 10,256 such parts; and by the heat of water boiling strongly 10,725; and by the heat of melting tin 11,516. So that reckoning the freezing point as a common limit between heat and cold, he began his scale there, marking it 0, and the heat of the human body he made 12°; and consequently, the degrees of heat being proportional to the degrees of rarefaction, or \( 256 : 725 :: 12 : 34 \), this number 34 will express the heat of boiling water; and by the same rule, 72 that of melting tin†. This thermometer was constructed in 1701.

To the application of oil as a measure of heat and cold, there are insuperable objections. It is so viscous, that it adheres too strongly to the sides of the tube. On this account it ascends and descends too slowly in case of a sudden heat or cold. In a sudden cold, so great a portion remains adhering to the sides of the tube after the rest has subsided, that the surface appears lower than the corresponding temperature of the air requires. An oil thermometer is therefore not a proper measure of heat and cold.

All the thermometers hitherto proposed were liable to many inconveniences, and could not be considered as exact standards for pointing out the various degrees of temperature. This led Reaumur to attempt a new one, an account of which was published in the year 1730 in the Memoirs of the Academy of Sciences. This thermometer was made with spirit of wine. He took a large ball and tube, the dimensions and capacities of which were known; he then graduated the tube, so that the space from one division to another might contain 1000th part of the liquor; the liquor containing 1000 parts when it stood at the freezing point. He adjusted the thermometer to the freezing point by an artificial congelation of water: then putting the ball of spirit of wine into boiling water, he observed whether it rose 80 divisions; if it exceeded these, he changed his liquor, and by adding water lowered it, till upon trial it should just rise 80 divisions; or if the liquor, being too low, fell short of 80 divisions, he raised it by adding rectified spirit to it. The liquor thus prepared suited his purpose, and served for making a thermometer of any size, whose scale would agree with his standard.

This thermometer was far from being perfect. As its defects, the bulbs were three or four inches in diameter, the surrounding ice would be melted before its temperature could be propagated to the whole spirits in the bulb, and consequently the freezing point would be marked higher than it should be. Dr Martine accordingly found, that instead of coinciding with the 32nd degree of Fahrenheit, it corresponded with the 34th, or a point a little above it. Reaumur committed a mistake also respecting the boiling point; for he thought that the spirit of wine, whether weak or strong, when immersed in boiling water, received the same degree of heat with the boiling water. But it is well known that highly rectified spirit of wine cannot be heated much beyond the 173rd degree of Fahrenheit, while boiling water raises the quicksilver 37 degrees higher. There is another thermometer that goes by the name of Reaumur's, which shall be afterwards described.

At length a different fluid was proposed, by which Mercurial thermometers could be made free from most of the defects hitherto mentioned. This fluid was mercury, and seems first to have occurred to Dr Halley in the last century; but was not adopted by him on account of its having a smaller degree of expansibility than the other fluids used at that time*. Boerhaave says that the mercurial thermometer was first constructed by Olaus Roemer; but the honour of this invention is generally given to Fahrenheit of Amsterdam, who presented an account of it to the Royal Society of London in 1724.

That we may judge the more accurately of the propriety of employing mercury, we will compare its qualities with those of the fluids already mentioned, air, alcohol, and oil.

Air is the most expandable fluid, but it does not receive nor part with its heat so quickly as mercury. Alcohol does not expand much by heat. In its ordinary state it does not bear a much greater heat than 173° of Fahrenheit; but when highly rectified it can bear a greater degree of cold than any other liquor hitherto employed as a measure of temperature. At Hudson's Bay, Mr Macnab, by a mixture of vitriolic acid and snow, made it to descend to 69 below 0 of Fahrenheit. This is an inconvenience, however, attending the use of this liquor; it is not possible to get it always of the same degree of strength. As to oil, its expansion is about 15 times greater than that of alcohol; it sustains a heat of 600°, and its freezing point is so low that it has not been determined; but its viscosity renders it useless.

Mercury is superior to alcohol and oil, and is much more manageable than air. 1. As far as the experiments of mercury, ments already made can determine, it is of all the fluids hitherto employed in the construction of thermometers, that which measures most exactly equal differences of heat by equal differences of its bulk: its dilatations are in fact very nearly proportional to the augmentations of heat applied to it (A). 2. Of all liquids it is the most easily freed from air. 3. It is fitted to measure high degrees of heat and cold. It sustains a heat of 60° of Fahrenheit's scale, and does not congeal till it fall 30 or 40 degrees below 0°. 4. It is the most sensible of any fluid to heat and cold, even air not excepted. Count Rumford found that mercury was heated from the freezing to the boiling point in 8 seconds, while water took two minutes 13 seconds, and common air 10 minutes and 17 seconds. 5. Mercury is a homogeneous fluid, and every portion of it is equally dilated or contracted by equal variations of heat. Any one thermometer made of pure mercury is, cæteris paribus, possessed of the same properties with every other thermometer made of pure mercury. Its power of expansion is indeed about six times less than that of spirit of wine, but it is great enough to answer most of the purposes for which a thermometer is wanted.

The fixed points which are now universally chosen for adjusting thermometers to a scale, and to one another, are the boiling and freezing water points. The boiling water point, it is well known, is not an invariable point, but varies some degrees according to the weight and temperature of the atmosphere. In an exhausted receiver, water will boil with a heat of 98° or 100°; whereas in Papin's digester it will require a heat of 412°. Hence it appears that water will boil at a lower point, according to its height in the atmosphere, or to the weight of the column of air which presses upon it. In order to ensure uniformity therefore in the construction of thermometers, it is now agreed that the bulb of the tube be plunged in the water when it boils violently, the barometer standing at 30 English inches (which is its mean height round London), and the temperature of the atmosphere 55°. A thermometer made in this way, with its boiling point at 212°, is called by Dr Horsley Bird's Fahrenheit, because Mr Bird was the first person who attended to the state of the barometer in constructing thermometers.

As artists may be often obliged to adjust thermometers' rules for under very different pressures of the atmosphere, philo-adapting sophers have been at pains to discover a general rule which might be applied on all occasions. M. de Luc, in his "Recherches sur les Mod. de l'Atmosphère" from a series of experiments, has given an equation for the allowance on account of this difference, in Paris measure, which has been verified by Sir George Shuckburgh *; also * Phil. Trans. for Dr Horsley, Dr Mackelyne, and Sir George Shuckburgh, have adapted the equation and rules to English and measures, and have reduced the allowances into tables, for the use of the artist. Dr Horsley's rule, deduced from De Luc's, is this:

\[ \frac{99}{8990000} \log_{10} x - 92.804 = h. \]

where \( h \) denotes the height of a thermometer plunged in boiling water, above the point of melting ice, in degrees of Bird's Fahrenheit, and \( x \) the height of the barometer in 10ths of an inch. From this rule he has computed the following table, for finding the heights, to which a good Bird's Fahrenheit will rise when plunged in boiling water, in all states of the barometer, from 27 to 31 English inches; which will serve, among other uses, to direct instrument-makers in making a true allowance for the effect of the variation of the barometer, if they should be obliged to finish a thermometer at a time when the barometer is above or below 30 inches; though it is best to fix the boiling point when the barometer is at that height.

---

(A) We have affirmed that the expansions of the bulk of quicksilver by heat are nearly (for they are not strictly so) in a regular arithmetical progression, according to the quantity of heat it is exposed to; and such seems to be the case according to the Table published by Mr de Luc, at page 309 of his first volume on the Modifications of the Atmosphere. The following extract of this table shows these variations: and the first and second differences are added, in order to render these irregularities more sensible. They are such as can hardly be conceived from the nature of any substance, without the influence of extraneous and accidental causes, which may have escaped the eye, vol. ii. attention of the observer; neither have they been found exactly true by Dr Crawford. Mr de Luc supposes the whole heat from melting ice to that of boiling water to be divided into 80 parts; by the fractional subdivisions of which he expresses the absolute quantities of heat, answering to each 5 or 10 degrees of Reaumur's thermometer (22.5 of Fahrenheit's scale); so that the whole sum of these fractions amounts exactly to the assumed number 80. They are as follow:

| Degrees | Reaumur's Thermometer | Fahrenheit's Thermometer | Quantities of heat | First differences | Second differences | |---------|-----------------------|--------------------------|-------------------|------------------|------------------| | 80 | | | | | | | 70 | | | | | | | 60 | | | | | | | 50 | | | | | | | 40 | | | | | | | 30 | | | | | | | 20 | | | | | | | 10 | | | | | | | 0 | | | | | |

---

*Phil. Trans. for 1736.