REVERSED AND NON-REVERSED SPECTRA. 



77 



In conclusion, I may recall that if we regard 100 fringes between the D 

 and b lines as still available for counting under proper facilities, the succes- 

 sive ranges of measurements will be roughly as follows : 



The transition from fine fringes to medium is a little abrupt. Otherwise, in 

 cases where manual interference is not permissible, all thickness of air-films, 

 from a fraction of a wave-length of light to nearly 2 cm., may be adequately 

 measured in this way to advantage. It is probable, moreover, that it would 

 be advisable to observe the fine fringes by transmitted light, but to leave the 

 grating (which may be a film grating) clear, and to observe the medium and 

 coarse fringes by reflected light. A concave mirror and lens (reflecting tele- 

 scope) should be used for this purpose, as this will put the observer behind 

 the plates in all cases. 



