C. Barus — Displacement Interferometer. 119 



Inasmuch as the water circulation was omitted during the ring 

 measurements and the exact value of X was not specially found, 

 the agreement is as close as may be expected. 



The two data selected from many similar results show how 

 easily both methods may be used for mutual corroboration. 

 Clearly the ring method, since it involves the evanescence of 

 236 rings, is more sensitive, but also less expeditious. It is not, 

 however, necessary to observe all the rings ; the disappearance 

 of a reasonable number, say 25 or 50, establishes the rate of 

 evanescence per cm. of mercury, or more conveniently the 

 number of cm. of mercury per vanishing ring. If this is 

 determined at the beginning and end of exhaustion the mean 

 result is adequate. 



If the air tube of the apparatus is so modified that the air 

 may be heated electrically the ring method should be equally 

 available for temperature measurements. The results could be 

 compared with a thermo-couple having its fine junction inserted 

 in the tube and read simultaneously. 



As a method of pressure measurement, since for a tube 23'8 cm 

 long, 3*12 rings vanish per cm. of mercury, i. e., about *1313 

 rings per cm. of tube length per cm. of mercury pressure, the 

 method is not very sensitive unless a long tube be used. A tube 

 1 meter long would, for instance, give 13 rings per cm. of mer- 

 cury, admitting of the measurement of pressure to *08 cm per 

 ring. One might estimate to *5 cm , while a more highly refract- 

 ing gas would secure greater precision. 



The method of pressure measurement has the rare advan- 

 tage, however, of being absolutely instantaneous, as reproducing 

 immediately the state of the gas. It is therefore remarkably well 

 adapted for the study of adiabatic phenomena. Many equa- 

 tions relatively to such transformation of gases are thus open 

 to investigation. Since (/*— 1) = A N/e = n\/2e and p = 

 CQ*-1)8, 



-£-= C. AiT/e = C\n/2e, 



the variations of p/8 are directly given by the number, n, of 

 rings vanishing. But the relation of p and 8 is also given 

 either by the intrinsic equation of the gas or by its adiabatic 

 equation, according to the transformation which has been 

 imposed on the gas, so that p and 8 are each determinable. 



Brown University, Providence, E. I. 



