C. Barus — Displacemerit Interferometer. 115 



from the front face of mirror and the rear face of grating are 

 in coincidence horizontally and vertically, i^may be revolved 

 for this preliminary adjustment. To find the angle of inci- 

 dence, the graduated plate K is turned from the given position 

 of coincidence until the image of the slit falls upon the slit 

 itself at the end of the collimator A, as explained above. If 

 the fringes are not sharp, they may be made so by further 

 adjusting the set screws of W or the grating, by trial. This 

 usually succeeds easily, remembering that the fringes move 

 about a horizontal axis normal to the mirror when the mirror 

 moves about a horizontal axis parallel to its face. For other 

 details the earlier paper should be consulted. 



15. Other Measureinents : High Temperature, Adiahatic 

 Transformations, etc. — The displacement interferometer con- 

 structed with its arms made of gas pipe is adapted for high 

 temperature investigation, if a current of cold water, at con- 

 stant temperature, be passed through the arms in question ; 

 they will then be kept at invariable length, however much the 

 atmosphere about them may change in temperature. Further- 

 more, since the distance between the central grating and the 

 opaque mirror may easily be increased to a meter or more, 

 tubes of considerable length may be inserted in the interfering 

 beams of light. The displacement interferometer should be 

 used with the angle of incidence nearly zero, in which case 

 this angle vanishes from the micrometer reading and the 

 observing telescope lies in a particularly convenient position 

 side by side with the opaque mirror on the micrometer. This 

 is thus immediately at hand. 



It seemed to me, therefore, that a particularly interesting 

 subject for investigation would be the relation of temperature 

 and pressure of the index of refraction jjl of air. According 

 to Lorentz* the /* — 1 for air follows the equation 



p= 0(^-1)8 



(p pressure, C constant, 8 absolute temperature), coinciding in 

 form to the intrinsic equation of a gas since the temperature 

 coefficient of /n— 1 for air is very nearly equal to its coefficient 

 of expansion. Mascart finds this not quite true. Pressures 

 are to be corrected by (1 + fip) where /3 is equal to -000,007,2 

 relative to cm. of mercury and the temperature coefficient is 

 a — *00382. At all events, the temperature coefficient a is so 

 large that a method of high temperature measurement is not 

 out of the question on the one hand, while on the other the 

 variation of a throughout long ranges of temperature is itself 

 of considerable interest. I have, therefore, made a few tenta- 

 tive measurements at low temperatures to test the apparatus 



* See the admirable summary in Landolt and Boernstein's Tables. 



