126 Mr. J. Y. Buchanan. On Rapid Variations of [May 31, 



carefully measuring it, and assuming a probable thickness of the 

 glass. In dealing with problems of this sort it is necessary to 

 express the specific heat in terms of the volume, and for this purpose 

 the ordinary numbers which express the capacity for heat of unit 

 weight have to be multiplied by the density, which expresses the 

 weight of 1 c.c. of the substance. The density of mercury is 13'596, 

 and that of ordinary glass is 2 - 45 ; their specific heats per unit weight 

 are - 033 and 0*19 respectively; whence the capacity for heat of 

 1 c.c. of mercury is 0*449, and of glass 0'466. If their specific heats 

 are taken as identical and equal to 0*457, the error made will not be 

 more than 2 per cent., in the extreme case where the bulb is all glass 

 or all mercury. 



Hence it appeared that there was no necessity for knowing the 

 thickness of the glass of the bulb or the weight of mercury in it. 

 For calorimetric purposes, a knowledge of the volume of the bulb 

 suffices, and it is immaterial in what proportion the two substances 

 are present. The figures on which this calculation are based are for 

 ordinary soda or potash glass, which was no doubt nsed in the con- 

 struction of the German thermometers which I was using. 



Using the value 0*457 for the specific heat per unit volume of the 

 bulb, and whirling the thermometer at the uniform rate of 6 m. per 

 second, twelve observations were made of the thickness of the film of 

 air heated to the full amount, corresponding to the fall of tempera- 

 ture of the thermometer. The difference between the initial tem- 

 perature of the thermometer and that of the air varied from 18 C. to 

 2 C., and the resulting computed thicknesses of the film of air heated 

 varied from 0*209 to 0*267 mm. ; the mean value was 0*237 mm. 



The measurement of the volume of the bulb requires some atten- 

 tion. The most convenient form of the bulb is the cylindrical, and 

 it is also the most common. But the bulbs are very rarely truly 

 cylindrical, they are often considerably tapered. It is not sufficient 

 to measure the diameter of the bulb with callipers, it is necessary to 

 measure the circumference at various parts of the bulb. One simple way 

 is to envelop the bulb with a wrapper of tissue paper, like a cigarette, 

 to blacken the edge of the paper which is laid inside. When 

 the paper is neatly and smoothly laid on, pressure with the finger 

 along the line of the inner edge of the paper produces a sharp impres- 

 sion of the edge on the overlapping paper. On unrolling the paper 

 the exact envelope of the bulb lies between the blackened edge of the 

 paper and the impression which it has made on the paper underlying 

 it. The length of the bulb is very easily measured, and when the 

 paper envelope has been, to begin with, given the proper length, it 

 measures the outer surface of the bulb, less the surface of the end. 

 This is assumed to be hemispherical, and is added accordingly. The 

 upper end of the bulb, where the stem joins on, is neglected, as in 



