A Uetermination ui- riii': Raiio («) of iiu-: Stkcimc IIfats at 



Constant Prmssl're and at Consiant VolumI'; i-ok Air, 



0\\c;i{N, C/VRiioN-DioxiDi:, and IIndrocen. 



By O. Lumjikk and E. Pkinosiikim. 



Introductory. 



The experiments detailed in the following pages are based on a method of 

 research/ which in all its essential features we had carried through fur air as far 

 back as the year 1887. The data obtained in this original test were mutually ac- 

 cordant to within as little as ^fn, but their absolute value was decidedly too small, 

 throuo'hout. 



The researches, however, were (piite sufficient tt) convince us that the bolo- 

 metric method of temperature measurement is well adapted for the direct de- 

 termination of such increments of temperature as are met with in the adia])atic 

 expansion of gases; but that the method, nevertheless, can be expected to lead to 

 trustworthy results only when an extreme of sensitiveness and efficiency has been 

 imparted to the bolometric thermometer. This must therefore be practically 

 instantaneous in its registry (/. e., free from inertia), and the effect of heat con- 

 ducted either way through the terminals must be eliminated. It is to the latter 

 source of error that we chiefly attribute the discrepancy in our earlier results. 



In the meantime bolometric resistances^ meeting every recpiirement had been 

 successfully constructed. Thus it seemed reasonable to assume that in a repetition 

 of the above experiments with the new bolometric material now available, the error 

 due to terminal heat conduction might be reduced in any desirable degree. We 

 were therefore fortunate in being able to repeat our work at the Pliysikalisch- 

 Technische Reichsanstalt in consequence of a grant froTu the Hodgkins fund Avhich 



' Verhatidl. d. Physik. Ges. zu Berlin, 1887, pp. 136-140. 



' O. Lummer and F. Kurlbaum : IVieJ. Ann., vol. xlvi., pp. 204-224, 1892 ; Zeiischr. f. Imlru- 

 mcntenkunde, vol. xii., pp. 81-89, 1892. 



1 



