398 Dr. Andrews on the Heat developed during the Formation of the 



glasa balls within, and immediately afterwards secured in the interior of the 

 larger cylinder, fig. 6, where the whole was rapidly rotated, for the space of five 

 and a half minutes, from the time of observing the temperature. It was then 

 removed, and the temperature of the water again observed. In the case of bro- 

 mine and iodine, all that now remained to complete the experiment was to weigh 

 the water in the copper vessel, but, in the case of chlorine, the original volume of 

 the gas had to be determined. For this purpose, the glass vessel was placed in a 

 water-trough, and the cork withdrawn. From the quantity of water which rushed 

 in, the bulk of the chlorine was easily estimated. It is almost unnecessary to add, 

 that, in every instance, the whole of the chlorine had entered into combination ; 

 the small residue being atmospheric air, unavoidably introduced when the bulbs 

 were inserted. 



1 1 . The accuracy of experiments of this kind greatly depends upon the heat 

 which is gained or lost by the apparatus during the course of the experiment. 

 In a vessel placed apart from other sources of heat, the losses and gains of heat 

 will evidently be equal to one another for equal diflPerences of temperature above 

 and below that of the surrounding air. But in the apparatus I have just de- 

 scribed, from the proximity of the person of the observer, and the necessity of 

 grasping the tin cylinder while placing it in, and removing it from, the rotating 

 machine, this middle point is no longer the temperature of the air, but 1°.4 above 

 that point. Direct experiments also showed that the water had nearly attained 

 its maximum point in 45", from the time when the glass balls were ruptured, and 

 15" usually elapsed from the observation of the first temperature to the latter 

 moment. We may, therefore, assume that the water is at the maximum tempe- 

 rature during 4-|-', and at the minimum during 15". If we put e for the excess 

 of the final temperature above the air, e' for the difference between the initial 

 temperature and the same, and r and r' for the corrections to be applied for the 

 cooling and heating of the apparatus, during periods of 4^' and 15" respectively, ' 

 we shall have 



R = + (e-r.4)X 0.049, 



r' zz — (e + r.4) X 0.003 -fO°. 03. 



12. The constant quantity 0°.03 is added to the correction for simple heat- 

 ing, as an allowance for the heat, transmitted by the hand through the apparatus, 

 while rupturing the balls. The temperature of the water being generally so ad- 



