2 n.VVlO OK SPECIFIC UKAl'S. 



\\;is cMiiru-oiisly i>lafed at our disjuisal l)\ llit- 8iiiitljsuiiiaii liiNliiiitinii in \\ asliiii<(- 

 tini. It mve.s us iiiiu-li pleasuie ai^aiii to tender to the Iiistitiition our irrateful 

 acknowledgments for tbe favois received. 



Tlie results of the pieseiit investigatii)n were cominuincated 1>\ ptTniission to 

 the British Assoi-iation at the O.xfonl meeting.' We purpo-sely withheld the pnbli- 

 i-ition of the body of oui' wmk, howevci', inasmuch as we hoped to cair\ the 

 experiments forward into ivgions uf much higher tem[)eratuie ami to include these 

 results, tlic pidiminary preparations lor w liich had long been completed, in a single 

 memoir. In view of the technical ditticulties encountered in the further develop- 

 ment of our work, this plan was abandoned. It will take some time before the 

 necessary e.\])erience for the coustnu-tion of high-temperature baths of sufficient 

 constancy is at hand. 



I. ThK MkTIIoD in PlMNCII'l.K. 



The method used for the measurement of the specilichcat ratio i.f gases 

 depends, in the present as in ineceding investigations in the same direction, un the 

 law of inliulxitir expansion of tlie gas under treatment. Foi; the ca.se uf i\. j>trfe<-'t 

 gas expanding adialtatically from an initial pressure yy, to a linal pressure y^,, the 

 ratio of absolute temperatures, 7J and T^, corresponding to y>, andy^., ma\ be writ- 

 ten (J, >Z') 



J{ = (;;K ■<■> 



where « = ^ is the ratio of thesi)ecitic heats at constant piessure and at constant 

 volume. I'fom e(juation (I) « is found as 



1 ^' 



"= .^^---r ^^) 



lOjr /l _ log I 



Hence H may be computed if a jierfect gas is permitted to expand adiabatically in 

 such .1 way that 'Pi,p-i, 'l\, and T.. arc .ill measurable. " Foi- the case of a gas com- 

 pressed at the temperature 7J as far as the pressure /y, , suddcuK expanding to 

 atmos[>heric pressure y>;^,, the three <|uaiitities ^7, , ^'a. ^i '"'t' f<"ind with relative 

 ease. The chief diflicidty lies in finding the temperatuie T.,, or the final tempera- 

 ture of the gas which has been cooleil by sudden expansion from />, toy<o; for the 

 change of temperature from 7[ to T., nnist invariably occur in a very short time if 

 the change of pressure is to be rapid enough to lie compatible with the conditions 

 of adiabatic expansion. On the other hand the cooled gas will not remain at the 



' Rt|n)rt, Urilish Association, Oxford, 1H94. 



