370 



THE AMERICAN NATURAEIST 



[Vol. XLI 



ceive colors." Two colors of equal l)rightness are distinguished 

 better than two grays of equal brightness; and though the bright- 

 nesses are the same, colors may be distinguished from grays. ' 



In the dancing mouse, however, the cones of wliich are at least 

 very rod-like, Yerkes has recently found that color vision is ex- 

 tremely poor. There is some evidence of (hscrimination of red 

 and green, and of red and blue, but none whatever of l)lue and 

 green. A]){)arcntly such visual guidance as is received results 

 from dif^'ercnccs in brightness. The mouse discriminates blacks 

 grays, and whites.^ 



Because of the inherent difhculties in the investigation of color 

 vision in the lower animals, comprehensive results liave not 

 yet been obtained, but the newer methods |)romise notable 

 discoveries. 



Development. Since color vision is a complex differentiation, 

 it might be expected that in the course of development, an individ- 

 ual should successively pass through the simpler stages by which 

 it was acquired. Anatomically it has been shown that the retinal 

 layers first become distinct at the center of the retinal cup, and 

 that the differentiation of tlie retinal cells decreases from the 

 center toward the periphery. In the chick it is said that the cone 

 nuclei may be identified at an earlier stage than the rod nuclei,^ 

 but it is not generallv recognized that one t'onn of visnal cell ])re- 

 cedes the other. 



The development of color vision has hvvu theon^tieallv consid- 

 ered l,y Mr^. l.iidd Franklin.' Il.-r theory assnnies that tli,> color- 



phoio-ehemieai substance called the gray substance, which is com- 



These gray molecules, which persist in their {)rimitive state only 

 in the rods, upon disas.sociation furnish us with the gray sensa- 



