158 PROCEEDINGS OF THE AMERICAN ACADEMY 



of torsion rend off on a scale on the edge of k I gives the correction to 

 be added to or subtracted from the weights op. 



One observer constantly reads the circle k I, and the other con- 

 stantly records the transits of the mercury over the divisions of the 

 thermometer. 



A series extending over from one half to a whole hour, and record- 

 ing a rise of 15° C. to perhaps 25° C, and in which a record was 

 made for perhaps each tenth of a degree, would thus contain several 

 hundred observations, from any two of which the equivalent of heat 

 could be determined, though they would not all be independent. 

 Such a series would evidently have immense weight; and, in fact, I 

 estimate that, neglecting constant errors, a single series has more 

 weight than all of Joule's experiments of 1849, on water, put 

 together.* 



The correction for radiation is inversely proportional to the ratio of 

 the rate of work generated to the rate at which the heat is lost; and 

 this for equal ranges of temperature is only ^ as great in my 

 measures as in Joule's ; for Joule's rate of increase was about 0°.G2 C. 

 per hour, while mine is about 35° C. in the same time, and can be 

 increased to over 45° C. per hour. 



3. Details, 

 The Calorimeter. 



Joule's calorimeter was made in a very simple manner, with few 

 paddles, and without reference to the production of currents to mix 

 up the water. Hence the paddles were made without solder, and 

 were screwed together. Indeed, there was no solder about the 

 apparatus. 



But, for my purpose, the number of paddles must be multiplied, so 

 that tbere shall be no jerk in the motion, and that the resistance 

 may be great: they must be stronger, to resist the force from the 

 engine, and they must be light, so as not to add an uncertain quantity 

 to the calorific capacity. Besides this, the shape must be such as to 

 cause the whole of the water to run in a constant stream past the 

 thermometer, and to cause constant exchange between the water at 

 the top and at the bottom. 



* Forty experiments, with an average rise of temperature of 0°.5G F., equal 

 to 0°31 C ., gives a total rise of 12°.4 C, which is only about two thirds the 

 average of one of my experiments. As my work is measured with equal accu- 

 racy, and my radiation with greater, the statement seems to be correct. 



