Fahrenheit. Centigrade. Reaumur. Fahrenheit. Centigrade. Reaumur. Fahrenheit. Centigrade. Reaumer. Fahrenhelt. Centigrade. Reaumur.

difference is owing to the unequal expansion of the glass of which the thermometers are constructed. These different remarks show that the determination of temperatures presents chances of error, and that the greatest care must be taken in the management of the thermometer.

liquid is introduced into it, in sufficient quantity to fill the horizontal branch, and half of each of the vertical branches of the tube. It is necessary to select a liquid for this purpose which emits no vapour at ordinary temperatures; hence, in general, sulphuric acid coloured red is employed. The apparatus being thus completed, the air is made to pass from one bulb to the other, by heating them unequally, until, after several trials, they are brought to the same temperature, and the level of the liquid becomes the same in both branches. The mark for zero is then put at each extremity of the liquid. It is then graduated by raising one of the bulbs to a temperature 10° higher than that of the other, when the air which it contains expands and forces the liquid column to rise in the other branch; when this column becomes stationary, 10 is marked on each branch at the point which the level of the divided into 10 equal parts, and the divisions are carried above and below zero along each branch of the instrument.

Limits in the use of the Mercurial Thermometer.-Of all the thermometers constructed on the principle of the expansion of liquids, preference ought to be given to the thermometer of mercury, because that this liquid expands in the most regular manner; it has even been observed that its increase in volume, between -36° and 100° centigrade, that is, -328 and 212° Fahrenheit, is proportional to the intensity of the heat. Yet, for temperatures lower than -36° C. or -32-8 F. it is necessary to have recourse to the alcohol or spirit-of-wine thermometer; for mercury freezes at -40° Centigrade, or -40° Fahrenheit; and in approaching this point its expansion is irre-liquid has reached; then the intervals from 0 to 10° are gular, that is, not proportional to the intensity of the heat. In the case of high temperatures, the indications of mercurial thermometers cannot be made to exceed 350° Centigrade or 662° Fahrenheit, because this is reckoned the boiling point of mercury; but, according to Mitscherlich, it is 360° Centigrade or 680° Fahrenheit.

The Alcohol Thermometer.-The alcohol or spirit of wine thermometer differs from the mercurial thermometer only in the liquid employed; the alcohol is coloured red with carmine. But the expansion of liquids is less regular when they approach their boiling point; thus alcohol, which boils at 172 Fahrenheit, expands very irregularly between the two fixed points of the thermometer. Thus, after having determined these points, as for the mercurial thermometer, if the interval between them on the alcohol thermometer be divided into 180°, it will be found that the latter thermometer does not give the same indications as the mercurial one bctween these points: in fact, the indications will be less by several degrees, so that when the mercurial thermometer stands at 122° Fahrenheit, the alcohol thermometer will stand at 111° Fahrenheit. In order, therefore, to construct an accurate alcohol thermometer, it must be done by comparison with a mercurial thermometer as the standard; that is, the two thermometers must be gradually heated together in a bath, and the temperatures indicated by the mercurial thermometer must be marked on the alcohol thermometer at the successive levels shown by the latter liquid. Thus graduated, the alcohol thermometer is equivalent to the mercurial thermometer, that is, it marks the same temperatures when it is placed in the same conditions. The alcohol thermometer is particularly employed to measure low temperatures, because that this liquid is not frozen by the greatest known degrees of cold. Leslie's Differential Thermometer.-Professor Leslie (afterwards Sir John), who died in 1832, a Scotch philosopher, constructed an air thermometer, intended to show the difference of temperature between two material points or media in the close vicinity of each other; hence its name, the Differential Thermometer. This instrument is composed of two glass bulbs filled with air, and connected with one another by a narrow tube twice bent at right angles from the extremities of its horizontal branch, and fixed on a stand, as shown in fig. 161. Before the apparatus is hermetically sealed, a coloured Fig. 161.

The Thermoscope of Rumford.-At the same time that Leslie invented the Differential Thermometer, the American citizen Count Rumford (alias Benjamin Thomson), who died at Auteuil, near Paris, in 1814, invented a thermometer of a similar nature, which has been called Rumford's Thermoscope. The difference between this instrument and the preceding one consists in this, that the two bulbs are larger, the horizontal branch is longer, and it is upon this branch that the graduation is made, see fig. 162. The index 1 is only about thre

Fig. 162.

quarters of an inch in length, and zero is marked at each extremity of it when, the two bulbs being at the same temperature, the index occupies the middle of the horizontal branch. The rest of the graduation is conducted on the same principle as in the preceding instrument. As to the appendage D, it is intended to regulate the apparatus, so that when too much air gets into one of the bulbs, the index is passed into this appendage, and the air is then allowed to pass into the other bulb. In order to make the index come out of this appendage and take the position which it ought to occupy, it is sufficient to incline the thermometer to one side; but this is effected only after several trials. There are other kinds of air. thermometers, to which we shall refer in the sequel.

Breguet's Metallic Thermometer.-A thermometer constructed on the principle of the unequal expansibility of metals, and remarkable for its extreme sensibility, was invented by A. Breguet, a watchmaker of Paris, who died in 1823. This instrument is composed of three equal slips of metal placed one above another, namely, platinum, gold, and silver, and soldered together throughout their whole length; they are then passed through a flatting machine, so as to form a very thin metallic ribbon. This ribbon is wound in a spiral form round a support, as shown in fig. 163, and the upper extremity being fixed on it, there is attached to the other extremity, which is free, a slight needle or index, which is also free to move over a horizontal dial carrying a centigrade scale. The silver, which is the most expansible of the three metals, forms the interior face of the spiral ribbon; the platinum, which is the least expansible, forms the outward face; and the gold is between them, When tho temperaturo is valued, the silver expanding more than the platinum or the gold, the

spiral unwinds itself from left to right, as shown in the figure; | the contrary effect takes place when the temperature is lowered. The gold is placed between the two other metals, because it has an expansion intermediate between that of Fig. 163.

lowest temperature to which the instrument has been subjected.

The Thermometrograph.-The spiral thermometer of M. Breguet has been modified by his nephew, of the same name, so as to indicate the hourly state of the temperature. In order to effect this, the index or needle carries a small style or pen full of ink, and below it is a moveable plate, on which are drawn twenty-four equal and equidistant arcs, having the centigrade graduation. At every hour, a clock movement causes the plate to advance by a quantity equal to the interval between two arcs, and at the same time to strike a small blow on the style of the index which then marks a black point on the arc. The number of the arc indicates the hour, and the position of the black point gives the corresponding temperature.

Wedgwood's Pyrometer.-The instruments employed to measure the high temperatures for which the mereurial thermometer could not be employed because of the vaporisation of the liquid and the melting of the glass, are called pyrometers, (from the Greek, fire-measure). There are, however, no good pyrometers. All those which have been hitherto constructed are far from giving the exact measure of temperatures. Wedgwood, whose name has long figured on our best pottery ware, invented a pyrometer founded on the principle that clay contracts by the action of heat. This instrument is composed of a plate of brass, on which are fixed three bars of the same metal, fig. 165. The length of each is about six inches. The first two, which are at first apart from each other by about half an inch, converge by about a sixth part of that distance between their extremities. The second and third, which follow the former, also converge by the same quantity; so that the whole length of the gauge is about a foot, and the convergence from one end to the other, about one part in seventy-two of the length. Every inch of the length is divided into 20°, which gives 240° on the whole length. In applying Fig. 164.

silver and that of platinum. By only employing the two latter metals, their difference of expansion would occasion a rupture of the apparatus. This thermometer is graduated by comparing it with the mercurial thermometer as the standard. Rutherford's Thermometer.-In the case of meteorological observations, it is important to know the highest temperature of the day, and the lowest temperature of the night. The common thermometers can only give us this information by continuous observation at every minute of the 24 hours, a thing which is quite impracticable. In order to obviate this difficulty a great number of instruments has been invented; of which the best is that of Rutherford. On a rectangular glass plate, fig. 164, are fixed two thermometers, with their stems

bent horizontally instead of being placed vertically. The first A is a mercurial thermometer; the second в is an alcohol thermometer. In the stem of the former there is a small iron cylinder at A, which slides freely in the tube, and serves as an index; for when it is put in contact with the extremity of the column of mercury, the instrument being arranged horizontally, if the mercury expands, it pushes the index before it; this index, however, is stopt in its motion as soon as the mercury ceases to expand; but it remains at the same point of the stem when the mercury contracts, for there is no adhesion between this liquid and the iron. The point where the index stops, therefore, marks the highest temperature which has been produced in a given period; in the preceding figure, the index shows about thirty-one degrees.

In the stem of the alcohol thermometer there is placed a small enamel cylinder B which is used as an index. If the temperature is lowered while the cylinder is at the extremity of the liquid column, the latter, when it contracts, draws the cylinder with it by the effect of adhesion, and the index moves onward to the point where the maximum contraction of the liquid takes place. When the temperature is raised, the alcohol expands and passes between the sides of the tube and the index, without displacing the letter. Consequently, the extremity of the index opposite to the bulb indicates the