4 dorsey: luminescence measurements 



square of the distance of the lamp from the milk-glass screen, 

 the constant of proportionality being determined from a care- 

 ful photometric standardization of the lamp and of the screen 

 and color filter as used. For example, the photometer screen 

 and a certain specimen were found to be of equal brightness 

 when the lamp was 70.5 cm. from the screen. The lamp was a 

 tungsten lamp burning at the color of a 4 wpc carbon lamp and 

 had a candlepower of 1.38 in the direction of the length of the 

 photometer. For the light from such a source the surface 

 brightness of the screen used had been found to be 3.96 micro- 

 lamberts per meter-candle illumination from the rear. Conse- 

 quently, this specimen had a surface brightness of 3.96 X 1.38 

 -f- (0.705) 2 = 11.0 microlamberts. A micro lambert is the 

 brightness of a perfectly diffusing and completely reflecting 

 white surface when illuminated by a source of unit candlepower 

 placed at a distance of 10 meters. 



When greater precision is desired, the top of the milk-glass 

 window should lie in the central section of the end of the photom- 

 eter and be so placed that the vessel containing the material 

 under study can be placed immediately above it and so thatthe 

 luminescent surface lies in the plane of the window. For the 

 highest precision it is desirable that the material be contained 

 in a thin vertical vessel with sides of thin plane glass and bot- 

 tom of thin material so that the two luminous fields can be 

 brought as nearly as possible into contact, so as to secure a mini- 

 mum visibility for the line of separation. 



In addition to the well known errors of ordinary photometry, 

 others of prime importance have to be guarded against. For 

 low luminosities it is necessary that the eye of the observer be 

 in a highly sensitive state; this condition is not attained until 

 after he has been in total darkness for at least 15 or 20 min- 

 utes. In order to retain this sensitive state the illumination 

 used for reading the scale, recording the observations, or other 

 purposes must be as faint as is consistent with the work, and 

 must be continued only as long as is absolutely necessary. 

 Furthermore, this light must be of such a nature that it will 

 not affect the luminosity of the material under study. These 



