This is an equation to find k. Also when we know k, we have

from this we can find w for the different temperatures.

A few of the readings' taken are:

Thermometer contained in the test tube of the calorimeter.

Corresponding reading of the scale of the calorimeter.

In ice.

At temp. of the room.

15.2

15.8

15.9

82

100

103

Taking the expansion of glass '000025

mercury 000179

per degree of temperature, and substituting in the above equation, we obtain values for w; but the calculation is rather long, and does not give very good results. It is thought that the temperature of the mercury is not uniform throughout and the same as the water, as the mercury is heated much sooner than the water. In the instrument for this purpose it would be better to be without the tube C.

J. T. KNIGHT.

1 These readings are interesting as giving evidence that the expansion of water is continued when the temperature is reduced below the freezing point. But as the rise of temperature was probably due in great part to conduction along the mercury, the numbers cannot be expected to give measurements of the coefficient of expansion.

No. 12, FAVRE AND SILBERMANN'S CALORIMETER.

THIS instrument, of which an account may be found in Jamin's Cours de Physique, Vol. II. or Witz' Cours de Manipulations de Physique, depends on the assumption that when a quantity of heat enters a mass of mercury the expansion is proportional to that quantity of heat. It is desirable to use as large a mass of mercury as possible, since by that means the increase in the temperature of the instrument in any one experiment may be rendered insensible. Favre and Silbermann used a calorimeter containing about 250 kilogrammes; but for laboratory models one-tenth of that quantity will suffice; the instrument here described would only hold about 16 kilogrammes.

The calorimeter used here consists of a spherical iron shell, cast in the mechanical workshop, 5 inches in diameter and inch thick. This shell is provided with three holes; into one is screwed a receiving tube of iron (like a test tube), which penetrates beyond the centre of the shell and which is only about 0.25 mm. thick, so that the hot bodies introduced may readily give up their heat to the mercury; the second hole, which is small, is the one to which the capillary reading tube is attached; this tube is sealed into a steel collar and screwed down to the shell, and it is also bent at right angles so that the part in which the expansion is measured can be placed horizontally against a millimetre scale; in the third hole a screw works, by means of which the position of the end of the mercury column can be adjusted in the reading tube1.

1 With this instrument it was found necessary to use a tube of small bore (about mm.). This makes it necessary to pay very great attention to the cleaning and drying of the tube, to avoid as far as possible the sticking of the mercury.

The calorimeter was filled as far as possible through the opening in which the adjusting screw works; this screw was then replaced and the whole heated to nearly 300° C., so as to drive out the air that might have been clinging to the inside of the shell; while this was going on, the filling up was completed by means of a funnel fitted into a cork and screwed down on the hole for the reading tube, which was at the highest point of the shell; when the filling was nearly completed the shell was moved about so as to dislodge any air still clinging to the adjusting screw or round the receiving tube.

The instrument was then placed in a wooden case and packed round as completely as possible with fossil meal to preserve it from change of temperature; the reading-tube was attached and adjusted to a scale. During the experiments all except the mouth of the receiving tube was covered up with tow.

In making the experiments the hot body was shot straight into the receiving tube, in which a quantity of mercury was placed, so as to conduct the heat rapidly to the sides of the receiving tube, and so into the calorimeter; but if the hot substance is one that acts on either iron or mercury, a glass test tube is inserted in the receiving tube and kept there against the buoyancy of the mercury by means of a cork collar in the mouth of the receiving tube. It was found necessary to keep a cork collar in the mouth of the receiving tube in all cases, so as to minimise the loss of heat caused by the hot body coming into contact with that part of the iron tube which is outside the body of the calorimeter.

The method of experimenting was as follows:

(1) The position of the end of the mercury thread in the reading tube was noted every 5 minutes.

(2) When sufficient readings had been taken to render it possible to deduce the law of variation due to outside influences the hot body was introduced.

(3) The reading tube was again observed as soon as possible and readings were taken, usually at intervals of 1 or 2 minutes, until the law of variation due to external causes could be estimated.

Thus by combining the results of (1) and (3) the necessary correction could be applied, as in the following example:

Between 10 and 20′ mercury moves 299-92 mm. = 207; deduct 0.84 x 10 for external heating, and this leaves 207-8.4 = 198.6 mm. as effect due to hot body.

=

Thus if the weight, and the original and final temperatures of the hot body are known, its specific heat can be calculated when the heat value of the scale divisions is known, and this was obtained by using a known weight of mercury heated to about 100° C.

In the above table it will be seen that when the hot body was introduced the mercury rushed forward 362 mm. in the tube, and then almost immediately withdrew 1633 mm.; this outrush and subsequent contraction is due to the sudden heating and expansion of the receiving tube, which speedily parts with its heat to the surrounding mercury, hence the contraction.

Some trouble was experienced in graduating the instrument owing to the difficulty of conveying the mercury from the hypsometer in which it was heated to the calorimeter; this was at last managed by enclosing the mercury in three test tubes, this giving three thicknesses of glass, and two air spaces for the heat to pass through. The inner test tubes were kept steady by means of cork collars at their ends, and the three were fastened together and fixed in a cork, to which was attached

a handle of copper wire. By means of this arrangement the temperature of the mercury remained constant near the boiling point when withdrawn from the hypsometer for 15", and then began to fall, when cooling effect had penetrated, at the rate of 0.2° C. per 15". The entire series of operations of removing the cork from the receiving tube, reading the temperature of the hot mercury, removing the thermometer, carrying the mercury to the receiving tube and pouring it in, and replacing the cork, was always completed in less than 1 minute, so that the reading tube could be observed; thus the hot mercury was not exposed much, if any, beyond the 15" during which its temperature had been proved to remain steady.

It will be seen later that these precautions were probably not sufficient to guarantee the mercury arriving in the calorimeter at the temperature of the steam. It is not unlikely that it was cooled considerably in passing over the cooler part of the test tube.

The specific heat of Platinum was taken as 0.0324, that of Mercury 0·0333.