444 



PROCEEDINGS OP THE AMERICAN ACADEMY. 



LITHIUM. 



sharp enough to show a doubling, if tlie components were even closer 

 than a tenth-meter, and no sucli doubling has been observed. The 

 curious reversals obtained by Hagenbach certainly call for an explana- 

 tion, but it is difficult to adopt the one given by him in view of the 

 absence of doubling in those lines where -it would be most easily detected. 

 The writer took a set of six photographs of the 4602 group on a film, 

 using the second order spectrum (5.5 t.-m. to the mm.) and varying the 

 exposures and amounts of vapor in the arc so as to furnish wide differ- 

 ences among the set. In these photographs the points of maximum 

 density in the image on either side of the "reversal" remain in constant 



