OF GASES AT CONSTANT VOLUJ\IE. 
957 
Looking at the three first experiments upon 18'7647 grammes of gas, it is seen that 
the first two, having nearly the same range, aftord a like precipitation (w) and the 
deduced specific heat is 0'322 ; but the third experiment, which is over a range 
commencing at a temperature (L) only one degree higher than the initial temperature 
of Nos. 1 and 2 , aftbrds a specific heat 0'307. This diiference of 5 per cent, reveals 
the existence of a large thermal effect dependent upon f-j. In short, liquid carbon 
dioxide is present at the initial temperature. Looking down the table it is observable 
that so far as Experiment 10 , similar wide variations of specific heat with small 
change of initial temperature occur. The remainder are sensibly free from this effect 
due to latent heat. Thus, compare 11 with 12 , and 13 with 14. All these last, from 
11 downwards, plot upon the prolongation of previous observation ; the others, as 
might be expected, lie upon a line rapidly bending upwards, away from the axis of 
density, and are not contained m the chart. 
The correction contained in column “ a ” is that due to the different thermal 
capacities of the vessels. It is derived from experiments given in Table XVI. 
Table XVI.—Comparison of the Empty Spheres. 
h- 
^2* 
^2 - h- 
IV. 
15-30 
100-65 
85-35 
- 0-0247 
12-89 
100-42 
87-53 
- 0-0256 
12-79 
99-87 
87-08 
- 0-0258 
86-65 
- 0-0254 
It will be observed that the correction is larger than formerly. This is in part 
accounted for by the increased volume of the active sphere. The change of medium 
during experiment produces an effect = V X 0'000r)2 (the ditierence of density of 
saturated steam and air at 100 °), and in this way the increased volume of nearly 
4 cub. centims. causes an effect upon the balance of 0‘0025 gramme. The outstanding 
increase in co is probably ascribable to lodgement of P 3 O 5 in the vessel, or even 
possibly to some chemical action between the CO 3 and the copper. From the mode 
of estimating W, upon liberation of the gas, this produces no error. 
In column “ b,” Table XV., is contained the result of all the other corrections 
calculated as in the previous experiments. The internal volume of the sphere is now 
89’984 cub. centims. at 13°:2 0. The elastic distension is taken as before. From 
these dtita the numbers in the next table are calculated. 
