18 



THE INTERFEROMETRY OF 



9. Continuation. Air column. An air compensator was now installed 

 consisting of a tube 0=15 cm. long and about 2 cm. in diameter, closed with 

 glass plates. The fringes were easily found and sharp. Unfortunately the 

 pump was not quite tight, so that, on breaking the count of fringes at low pres- 

 sures, it was difficult to state when the conditions had become isothermal. 

 Hence the results in table 5 are rough. Temp. 19.7. 



TABLE 5. 



The mean value thus appears as X = icr 5 X5.88 for sodium light. The equa- 

 tions used are 



(1) n\=e(n i) 



where n is the number of fringes counted, e the tube-length, and /i the index 

 of refraction of air. Again, 



(2) p 



where p is the pressure, tf the absolute temperature, and the constant C 

 computed from normal conditions (76 cm. and o C.) is (Mascart's values) 

 C = 952.6. Hence 



e p e dp e dp 



" " C d(n&) " Cddn 



(3) 



when # is constant. It is this assumption which is not quite guaranteed above. 

 To obviate this in the following experiments, the total number of fringes 

 were counted (table 6) from exhaustion to plenum. Their number was definite 



to the fraction of a fringe. 



TABLE 6. 



