108 



S. B. Williams — An Achromatoscojpe. 





tance from the center of the rings equal to the 

 distance this long slit is from the central band. 

 Equation (12) is applicable here also. 



There remains one other simple device for 

 studying the dispersion of a medium where 

 the interference bands from a thin film of the 

 substance is at hand as in fig. 1. In equation 

 (2) we see by differentiating N with respect to 

 A, that for non-dispersive substances, and air 

 may for all practical purposes be taken as such, 

 that 



d^/-dX = 2D/A y , 



while for a dispersive substance, 

 2D 



rfN/- dX 



X' 2 



( Xn 



Xdp 

 dX 



(2a) 



(2b) 



The quantity in parenthesis is the factor which 

 will, in general, cause the interference bands 

 from the dispersive film to crowd together 

 more rapidly than those from an air or vacuum 

 film as one goes toward shorter wave lengths 

 in the spectrum. Hence, superimposing a set 

 of air bands upon a set of bands from a dis- 

 persive film, there will be found interference 

 effects between the two sets of bands as they 

 get out of step with each other. This is shown 

 in fig. 6 where a set of air bands has been 

 superimposed upon a set of bands from a thin 

 film of alcohol. The photographic plates 

 were made to overlap part way and then 

 printing on " Yelox " paper occurred through 

 the overlapped plates. This gives simultane- 

 ously the air bands, the alcohol bands and the 

 bands due to superposition. If the air bands 

 are shifted so that the distances between the 

 bands become more nearly like those between 

 the bands from the dispersive film, then the 

 interference bands due to superposition become 

 less and less and any irregularities in dispersion 

 are very easily picked out. A set of air bands 

 may be photographed once for all and then 

 used as comparison bands for any other set of 

 bands taken from a dispersive film. 



