Sbptembise 9. 1898.] 



SCIENCE. 



313 



rise to the sensation of yellow, in the sec- 

 ond stage to that of bine. The visual 

 purple pertains to that element of the reti- 

 nal complex known as rods. These are not 

 present in the central portion of the retina, 

 and the visual purple is apparently absent 

 there also. But the fovea is sensitive to 

 blue and yellow, as also to green and red 

 Ebbinghaus supposes that a red-green sub- 

 stance exists, even in the fovea, green in its 

 first stage, red in its second; that the yel- 

 low-blue substance exists also in the fovea, 

 but that the two, present there in about 

 equal quantity, and nearly complementary 

 in color, neutralize each other, leaving the 

 fovea colorless. A white-sensitive substance 

 is also supposed to exist, more sensitive to 

 light than any of the colored substances, 

 and thus we arrive at three sets of color- 

 processes, similar to those of Hering. The 

 two types of color-blindness are explained 

 by reference to the fact that there are two 

 kinds of visual purple, found in the eyes of 

 different animals, one more relatively red 

 in tone, the other inclining more to violet. 

 The red-blind are supposed to possess one 

 of these, the green-blind the other. Cer- 

 tain anomalous and pathological color- sen- 

 sations are supposed to be due to disturb- 

 ances in the conducting nerves or the central 

 organs, and hence need not be fitted into 

 the scheme thus outlined. 



The physiological character of this theory, 

 as Mrs. Franklin has shown, can probably 

 not be sustained. It is difficult to believe 

 that such a balance between the visual 

 purple-yellow and the supposed visual red- 

 green could exist, in all stages of both, that 

 they would remain always complementary, 

 and so the latter always invisible. Yellow 

 light, according to this theory, should be 

 most active upon the visual purple, but, as 

 a matter of fact, this material is bleached 

 very slowly by sodium light, and, indeed, 

 Konig has shown that its maximum absorp- 

 tion ia not in the yellow, but in the green. 



The visual purple exhibits other striking 

 properties, of which the theory takes no ac- 

 count. It seems probable, on the whole, 

 that the office of this substance is really a 

 very different one, and that if it is con- 

 cerned at all with vision it is with the 

 sensation of white light, not colored. 



The theory of Ebbinghaus, then, if we 

 deny its connection with the visual purple, 

 rests upon the same basis as that of Hering, 

 — a visual substance not identified, perhaps 

 not discoverable, but recognizable only 

 through the precision with which it ex- 

 plains phenomena, and the hypothesis itself 

 becomes in the main a modification of Her- 

 ing's with the well-known pairs of photo- 

 chemical substances, modified in their char- 

 acter so as to meet the facts more perfectly, 

 removing some difliculties, but introduc- 

 ing others. 



The chief advantage of the hypothesis 

 for explanatory or speculative purposes lies 

 in its greater freedom. The theory of Her- 

 ing demands six color-processes. These 

 are so connected together that they make 

 not six, but three, independent variables. 

 Ebbinghaus so constructs his substances as 

 to leave them nearly independent, the blue, 

 for instance, no longer serving as the an- 

 tagonistic substance to the yellow, but re- 

 garded as developed out of it, and possess- 

 ing specific properties of its own. Under 

 certain conditions the color-substances are 

 supposed to neutralize each other, as with 

 Hering, a supposition which adds greatly 

 to the difiiculty of the hypothesis ; but, in 

 general, five independent variables are at 

 the command of the theorist, which, it is 

 evident, may be endowed with such various 

 properties as to explain almost any con- 

 ceivable difficulty of color-vision. 



It may also be said that, with such an as- 

 sortment of visual substances at command, 

 the properties of which have at present no 

 known chemical, physical or physiological 

 relations, but are deduced entirely from the 



