84 DISPLACEMENT INTERFEROMETRY BY 



for these intervals, and dx may be taken from the table. The results now 

 come out more smoothly. In this way we find 



Lines ............ C D b F 



io 4 (X/Aa;) .......... 2.84 2.55 2.24 2.10 



8x ................ 5.5 8.8 9.9 



io*(i/X 2 -i/X' 2 ).... 5-57 8.58 4.95 



io n (B-B') ........ 4-4 4-1 6.77 



The datum computed above for (B B') io u is 4.2 to 4.4 as far as the b line 

 and does not differ much from the tabulated values between C and b. 



Later I made an actual count of wedge displacement for 100 fringes at three 



spectrum-lines with results as follows (8x and A ( ^ J being as above) : 



Lines .......... C D b 



Ax ............ 24.9 22.9 20.1 (observed) 



io 4 (\/Ax) ...... 2.64 2.57 2.57 



io"(B-B').... 4-25 4-35 



io ll (B B'). ... 4.2 4.4 (as originally taken) 



These values are so nearly in accord with the data used (particularly as the 

 A and b lines are dark) that a discrepancy here can hardly be looked for. The 

 results of the rigorous equation would be io- 11 X4.5 and io~ n X4-4, respectively. 



It is rather unfortunate that the achromatic fringes can not be used for the 

 present purposes, as they are very clear; but in the first place the wave-length 

 to which /JL belongs is not implied. In the second place, with an increase of 

 glass-path, they soon multiply beyond recognition. Using the nearly colorless 

 central fringes of the group (flanked on either side by reddish fringes), the 

 data, also given in table 6, were obtained. The value of /.i computed therefrom 

 lies about midway between the aniline and benzol results. This is much 

 better than was anticipated. 



Experiments were now made with the air compensator for comparison. In 

 this case, if AN is the equivalent air-path removed by the exhaustion of the 

 compensator of length E, we may write 



if no is the index of refraction of air under normal conditions and B"= i.6sX 

 io~ 14 the dispersion constant of air. Thus the equation becomes in the above 

 notation 



B~B' 



If for MO i we write 292. 7 X io~ 6 and neglect B" (E= 13 cm. being the length 

 of the compensator used) we get 



AA T = 292. 7 Xio~ 6 X 13X273^/767= i. 367 X 10' -p/r 



