52 SCIENCE PROGRESS 



change with hydrogen would be the same as with other gases. 

 If these conclusions are correct the results should be intensified 

 in the case of helium, which has a much lower critical tempera- 

 ture and pressure. 



It appears to be probable also that at temperatures below the 

 critical Cp — Cv may be negative, in which case K = Cp/Cv would 

 further be less than unity. If this should be substantiated by 

 investigation it will throw some doubt upon the deductions 

 which are customarily made about the connection between k 

 and the total and external energy of the molecule. Certainly 

 the main conclusions are justified, and the deduction that k would 

 have its maximum value with monatomic molecules has been 

 abundantly demonstrated, first with mercury vapour and subse- 

 quently with the gases of the argon group, where the experi- 

 mental results all show values differing very little from 5/3. In 

 the liquid state the molecular heat of mercury is about 67 

 and in the solid 64, which would appear to indicate that even 

 in the solid state it is monatomic, as this value coincides with 

 the general value of the atomic heat of solid elements. 



However, the elements with simple molecules in the gaseous 

 state are still very little studied in the liquid and solid states, 

 partly owing to the low temperatures at which they would have 

 to be observed and partly because the importance of these 

 measurements is not very generally recognised except among 

 those who are fully occupied with these and similar questions. 

 There are three separate lines of experimental research which 

 are all very fruitful and which are at present only connected 

 together in a very imperfect way theoretically owing to the 

 want of sufficient data. The accurate study of isothermals, 

 which is the absolutely necessary foundation for an advance 

 in the theory of coincident condition, and the possibility of 

 arriving at a generally applicable equation of state can receive 

 most important assistance from the study of the Joule-Kelvin 

 effect and the specific heats. It must, however, be emphasised 

 that the preliminary and pioneer stages are past, and that unless 

 measurements are exact they have really very little value or are 

 actually harmful because they form the basis of false con- 

 clusions. 



In isothermal work it is possible at about the ordinary tem- 

 perature to arrive at an accuracy of about 0*02 per cent, mean 

 error in the determinations. As lower temperatures are used, 



