1897.] NEW YORK ACADEMY OF SCIENCES. 195 



letter, upon ji white ground when seen with the greatest distinct- 

 ness. In the first part of their research they regarded two 

 quantities of light as equal if, when illuminating successively the 

 same object (black udon a white ground) placed at the same 

 distance from the observer, the details could be distinguished 

 with the same minuteness. The ordinary definition of equality 

 is, that two quantities of light are considered equal if, when 

 illuminating two portions of a given white surface, the two por- 

 tions appear of the same luminosit3^ Their investigations 

 showed that the two definitions are not equivalent, therefore 

 photometers in which the method of distinctness of A'ision is 

 employed do not give correct results if we adopt the latter defini- 

 tion of the equality of two lights. 



Schafhautl * introduces a still different factor in photometric 

 measurements, viz., the time interval which may elapse between 

 two equal light impulses without the eye noticing any interrup- 

 tion. His photometer consists essentially of a pendulum to 

 which is attached a black screen with a single opening in it. 

 The source of light is placed behind the screen and the opening 

 observed through a binocular tube. The light can only reach the 

 eye when the epening in the screen is opposite one of the open- 

 ings in the binocular tube. The interval of time elapsing between 

 the two impressions can then be regulated by changing the 

 effective length of the pendulum. The intensity of the source of 

 light, he assumes, is inversely proportional to the square of the 

 times of swing when there is no flickering, or to the fourth power 

 of the length of the pendulum. No data were attainable from 

 which the accuracy of the method could be determined. 



Charpantierf employs a method for determining the luminos- 

 ity of colored lights which obviates the difficulties introduced 

 "by differences in color. He makes use of a phenomenon previ- 

 ously pointed out b}- Landolt that any source of colored light, if 

 sufficiently weakened, gives rise to a sensation of gray. Char- 

 pantier decreases the illumination until all sensation of color 

 vanishes and then compares the luminosities. It has, however, 

 been pointed out by Purkinje that as the illumination is de- 

 creased the luminosities of the colors in the more refrangible 

 parts of the spectrum are increased relative to those in the less 

 refrangible part, therefore determinations of relative luminosities 

 made under feeble illumination would not hold under normal 

 conditions of illumination. 



In the American Journal of Science for September, 1893, Pro- 

 fessor Rood describes a photometric method which is indepen- 



* Munchener Abhandl. Vol. VII. 1855. 



+ Comptes Rendus, LXXXVIII., page 301. 1879. 



