NOVEMBEB 15, 1907] 



SCIENCE 



653 



sisting essentially not in the presence of 

 separate factprs for the dominant and for 

 the recessive characters, but in the pres- 

 ence of something constituting the domi- 

 nant character which is absent from the 

 recessive gametes. So satisfactory, indeed, 

 are the results of this mode of representa- 

 tion that the probabilities are greatly in 

 favor of its truth. Indeed, when the inter- 

 relations of a complicated series of varietal 

 types have to be dealt with, the presence- 

 and-absence system, as we may call it, 

 applies so readily that its correctness is 

 scarcely doubtful. 



In simple cases, for instance, in that of 

 the rat, we may regard the color gray and 

 black as due to the operations of gray and 

 black determiners acting upon a distinct 

 factor for color. According to the scheme 

 promulgated by Cuenot, the two determin- 

 ers, gray and black, are regarded as allelo- 

 morphic to each other. 



Such a system, however, fails when, as 

 in the case of mice, a third color-type (in 

 addition to the albino) viz., chocolate, has 

 to be expressed. If, on the contrary, each 

 determiner is regarded as allelomorphic to 

 its own absence, a workable system is pro- 

 vided, which can deal with almost all the 

 observed facts. The gray— or technically, 

 agouti — mouse, then, contains all the fac- 

 tors. The black is black because it is minus 

 the determiner for agouti, and the choco- 

 late is wanting in the determiners both for 

 agouti and for black. The relations of all 

 the color types to each other are thus clear 

 except in so far as the relation of yellow 

 to the other colors is not quite satisfactorily 

 accoimted for on either system. 



It is at present beyond my purpose to 

 examine the suggestions made to deal with 

 that particular difficulty, but leaving this 

 special question on one side, we can draw 

 the clear deduction that each of these 

 varieties owes its existence to the absence 



or removal of some factor, from the gamete 

 of the type. 



Conversely in other cases we perceive 

 with equal certainty that the variety is due 

 to the addition of such a factor. 



To deal with this series of interactions, 

 the simple conception of dominant and re- 

 cessive is inadequate. We now need a 

 term to denote the relation between dom- 

 inant factors belonging to distinct pairs of 

 allelomorphs. 



Till lately we spoke of the relations be- 

 tween the gray color of the mouse to the 

 black color in terms of dominance. Those 

 terms, strictly speaking, should only be 

 applied to members of the same allelo- 

 morphic pair. We can perhaps best 

 express the relation between the gray 

 and the black by the use of the metaphor 

 "higher and lower," and I therefore sug- 

 gest the term epistaiic as applicable to 

 characters which have to be, as it were, 

 lifted off in order to permit the lower or 

 hypostatic character to appear. The same 

 method of representation is, of course, ap- 

 plicable to the series of factors for pattern 

 and for intensity of color. 



The case of patterns is in a special way 

 instructive. Symbolically we can repre- 

 sent pattern as due to determining factors, 

 like those which cause the tint or the in- 

 tensity of color. 



Though justifiable as a symbolic repre- 

 sentation, it is evident that the "factor" 

 for pattern may really be a quantitative 

 difference in the amount of one of the ele- 

 ments, presumably the chromogen. We 

 may imagine that the color appears on spe- 

 cial parts, just as color takes on the pre- 

 pared surface of a lithographer's stone, 

 always remembering that though the dis- 

 tinction between, for example, self -pattern, 

 the Dutch-pattern and the English-pattern 

 rabbit may thus be quantitative, the quan- 

 titative stages are fairly well defined. 



The point is of interest inasmuch as 



