510 



HENRY LAURENS AND HENRY D. HOOKER, JR. 



deviation type spectrometer. At first we used the Nernst glower as a 

 source. It was soon found, when we came to m.easure the radiant 

 energy of the shorter wave-lengths, that more than one such glower 

 was going to be necessary to give a light containing the unit energy 

 content which, from preliminary experiments we had selected as being 

 most satisfactory. By m.eans of this apparatus we were able to obtain 

 eight consecutive colored lights, each 30 wave-lengths broad, extending 

 from 430 ju/x to 670 yu/x and containing the same amount of energy 

 (see table 1). The glowers were clustered together and so connected 

 by switches, with their ballasts, that any number, from one to five, 

 could be burned at the same time. For the lights of shorter wave- 

 length a mirror was placed behind the glowers so that their reflected 

 image was focussed on the slit of the collimator. 



TABLE 1 



Shotoing the conditions binder ivhich the eight lights, each 30 wave-lengths wide and 

 equal in radiant eJiergy content icere obtained 



These eight lights were perfectly satisfactory but for the incon- 

 venience occasioned by burning five glowers so closely together and 

 the consequent bending and displacement. We therefore decided to 

 use a 1000 watt gas-filled Mazda lamp, burning on a 110 V city cir- 

 cuit, the variations occasioned by the fluctuations in the intensity of 

 this current being nullified by the large number of determinations of 

 the radiant energy content which were always made. Furthermore 

 we soon came to the conclusion that eight such lights were not all 

 that we desired, for since we wished to study the relative stimulating 

 value of the various wave-lengths in the spectrum, as well as to find 

 the wave-lengths of maximum stimulating efficiency, it was deemed 

 necessary, in order to be able to say, with an approach to accuracy, 



