RATIO OF SPECIFIC HEATS. 



5 



freciiieiitly tlir(iii;:;liout a w liolc ImiM'. Aside from tliis iii;iiii[)ul;itioii, Ihe expefi- 

 ineiits were favoral>lv eii'fimistaiiccd iiiasiiiucli as tliey were made in tlie subterranean 

 (•liam1)ers of the Keichsanstalt. Oii liri lining the temini;iiiire into coincidence with 

 tliat of the room, tlie ohserxcd changes would not 

 exceed the mere fracHoii of a degree foi' hours. 



Accurate measurement of tcnqiei'atui'o is of 

 extreme importance in its Wearing on the present 

 experiments, since an error of liut .(>■_' ('. in the 

 datum foi' 1\ in the case of the pressure values 

 actually employed, will affect the i-esult liy al)out 

 Af. To facilitate tlie reading of the .01 C". re- 

 quired, all tempeialuiT'S were so chosen as to make 

 the mercury meniscus of the thermometer coincide 

 very nearly with one of the division mnrks of the 

 thermometer. After long practice we also learned 

 to keep the tem[ierature constant to .01" C. (l»y 

 adding ice or hot water as occasion re(|uired and 

 continued observation of the thernn^meter), even 

 when the temperature of expei'iment lay consid- 

 erably below that of the room. To be sure that 

 the temperatTire was (piite the same at all parts 

 of the water bath, we made a special test in ^vhich, 

 while the stirring device was in action, a sensitive 

 thei'mometer was carried from place to place within 

 the bath. Simultaneously, of course, temperature 

 was kept constant in the lapse of time, at the 

 single place of obser\ation, by the means above 

 set forth. The thermometer probe used in this 

 test was a branch of one of Lunnner and Kurl- 

 baiun's surface bolometers, carefully jacketed by 

 a tight metallic case and appi-opriately insulated. 

 The change of internal resistance experienced by 

 this instrument gave us the necessary data for 

 changes of temperature encountered during the 



excursions from place to place within the bath. These values, given on pages 18, 24, 

 below, showed the variations of temperatures at any one place in the lapse of time, to 

 exceed the differences of temperature for diffei'ent positions within the bath at any 

 given time, the bath being kept in action in the maiinei' stated. We infer that 



Fig u UK 4. 



