RELATIVE PHYSIOLOGICAL VALUE OF SPECTRAL LIGHTS • 509 



mm. X 1.5 mm. and a resistance of 2.8 ohms. The galvanometer has 

 a resistance of 10.17 ohms, period 10 seconds, deflection in millimeters 

 at 1 meter for 1 micro-amp 415, and deflection in millimeters at 1 meter 

 for 1 micro-volt 40.81. The resistance was reduced by connecting the 

 coils in parallel. 



In order to control the sensitivity of this apparatus, readings of the 

 radiation from a 16 c.p. Standard carbon lamp, burning always on 

 100 V, were taken. By means of the control magnet, the sensitivity 

 of the galvanometer was varied imtil deflections of 400 mm., scale at 1 

 meter, were obtained. 



In addition to this apparatus we have a Franz Schmidt and Haensch 

 radiomicrometer (Schmidt (30) and Coblentz (29) ) lent to us by 

 Professor Bumstead. 



Before proceeding to the description of our final apparatus for ob- 

 taining colored lights a brief reference may be made to one which was 

 finished and then discarded for what will be obvious reasons. This 

 apparatus was based on the plan of the light generators used by Lau- 

 rens, Day and Gross (20), (21), (22). The light from a Nernst glower 

 (220 V.) was passed through a lens, emerging as a parallel beam. To 

 prevent over-heating a circulating water-cell was placed between the 

 glower and the lens. Later this was given up because it lengthened 

 the focal distance, in favor of a blast of compressed air emerging from 

 a wing-tip burner, directed against the front face of the lens. Aside 

 from the objection of cooling the glower and thus reducing the inten- 

 sity of light, the glower was made to buckle backward, due to the 

 imequal heating on the front and back surfaces. This happened al- 

 though the lower connecting wire of the glower was suspended in a 

 cup of mercury. The parallel rays of light were directed upon one of 

 the faces of a large 60 degree angle prism-bottle (10.5 cm. by 10.5 cm.) 

 filled with carbon bisulphide and placed at the angle of minimum devia- 

 tion. The spectrum thus obtained was focussed by a converging lens 

 and bands of light of the desired wave-lengths were obtained by plac- 

 ing a micrometer slit, opened to the appropriate width, at the focal 

 plane of the converging lens. 



This apparatus had the advantage of yielding a spectrum of large 

 size and of great brilliancy. But it was found impossible to obtain 

 with it pure and saturated light when the bands were examined with a 

 spectrometer. 



Since this large spectrometer was not successful we decided to ob- 

 tain our spectral lights by passing white light through a Hilger constant 



