132 
DRS. L. MORD, W. RAMSAY, AND J. SHIELDS ON THE 
separated out,* The ice sheath was then made to grow to the required size by pour¬ 
ing ether into the inner tube and blowing a current of air through it. Lastly, suffi¬ 
cient salt solution, S, cooled down to zero, was introduced into the inner tube, so that 
when the experimental tube was inserted, the whole of the bulb and a considerable 
portion of the stem were covered with the solution ; and care was also taken that 
the surface of the salt solution was 3 or 4 centims. below the top of the ice-sheath, 
so that most of the heat radiating from it would be caught in the sheath. The 
amount of mercury sucked into or expelled from the apparatus was preferably deter¬ 
mined by the deflection of the meniscus in the capillary tube, rather than by 
weighing, since the progress of the reaction could easily be seen at any instant. No 
ej'i’ors due to the “sticktion” of the mercury need be feared if the tube is tapped 
Fig. 1. 
occasionally. The capillary tube was one especially selected from a lai’ge number. 
For over a metre of its length its mean capacity was 0‘0001196 cub. centim. per 
millim., and the deviations from the mean value at ten points along its length were 
respectively -|- 9, -|- 17, fl- G, — 1, +4, — 11, — 16, — 1, + L, — 7, units in 
the last significant figure. In translating the deflection of the mercury meniscus in 
millims. into heat units, we have accepted DiETERici’st value, that 1 gram calorie 
corresponds to the displacement of 0’01544 gram of mercury ; and hence a deflection 
of 1 millim. in the tube we employed represents 0T053 of a gram calorie, or O'OOlOoS 
of a hundred-gram calorie, denoted by K in tins paper. 
Before proceeding to use the calorimeter for our experiments, the specific heats ol 
'* If this precaution is not taken, the water may be cooied several degrees belo\Y 0“ C., and, when once 
ice begins to form, it does so so suddenly that the apparatus is liable to burst, 
t DiETEiaci, ‘ Wied. Ann.,’ 1889, vol. 37, p. 499. 
