THE AID OF THE ACHROMATIC FRINGES. 87 



a mixture of these) , would have to be used for the examination of glasses in 

 general. The difference could then be expressed by the Billet or the air-mi- 

 crometer. The former has a large range, controls the fringes with remarkable 

 facility, but has to be standardized in terms of them, for each color. The 

 vacuum air-micrometer has a small range and is cumbersome. Finally, 

 there are two annoyances incident to the method as a whole, viz., the rela- 

 tively large dispersion coefficient of liquids and their relatively high tempera- 

 ture coefficient. The experiments therefore require extreme care as to details 

 throughout. Otherwise the present method is quite self-contained for BB f , 

 the difference of dispersion coefficients, is found on passing the ellipses 

 through the spectrum. 



