50 SCIENCE PROGRESS 



Cv can be measured independently, or their difference and their 

 ratio can be determined experimentally and their values be 

 thus obtained. 



The earlier experimenters, whose values were not very accur- 

 ate and who mostly used the permanent gases for their measure- 

 ments, concluded that Cp and Cv varied little with temperature, 

 and that Cv at any rate did not vary with the volume. Such 

 conclusions are quite in accord with deductions to be drawn 

 from a consideration of ideal gases obeying equation (2), with 

 which the difference of the specific heats will be a constant 

 from equation (13). 



However, experiment shows that both specific heats not only 

 vary considerably with change of temperature, but with change 

 of density also. There are not many substances on which 

 experiments have been made in several states, but the general 

 trend of change is indicated by what is known. 



In the solid state the majority of the elements have atomic 

 heats approximating to 6*5, even hydrogen being 5*88 as deduced 

 from the results of the change in the specific heat of palladium 

 by occluded hydrogen ; in the gaseous state it is 3*4 at o° C. If 

 we assume that the molecular heat is strictly additive, as it 

 appears to be in a large number of cases, we can compare the 

 heats of simple compounds such as water, which is particularly 

 interesting because it is the standard calorimetric substance. 

 It will be convenient to give molecular heats to avoid any 

 question about the atomic heats in the molecule, and to assume 

 the simple molecule in all states for this purpose, although it 

 is certainly more complex in many liquids and solids. We have 

 not always both specific heats at the different temperatures of the 

 vapour and gas, so must assume that the difference is equal to 

 2 gramme calories, which is very nearly the value for an ideal 

 gas. Taking then the constant pressure value throughout, we 

 find for ice at o'o° C. 9/36 and decreasing with the temperature, 

 for water at about 16 C, 18 ; for steam at 100, 6*35 + 2 = 8*35 ; for 

 water vapour at 1,000, u'52 + 2 = 13*52, assuming that the same 

 law holds as at lower temperatures. 



The changes here shown appear to be general. Starting 

 from the minimum at absolute zero, the value grows until it 

 reaches a maximum at some temperature coinciding with the 

 liquid state at moderate pressures, then again decreases to a 

 second minimum at a temperature corresponding with the 



