142 



The Journal of Heredity 



each other. With the same pattern of 

 stripes, the banding may vary from 

 very wide, giving the cat a nearly yellow 

 appearance, to very narrow. Indeed in 

 solid black cats, in which the banding is 

 wholly eliminated, the pattern of stripes 

 can often be made out, especially in 

 kittens, in a so-called ghost pattern. 

 Darwin'^ mentions this point. Accord- 

 ing to Pocock,^^ the ghost pattern is due 

 to a difference in glossiness in stripes 

 representing those of tabbies. 



In cats, with given widths of banding 

 the character of the stripes may vary 

 very greatly. There are three distinct 

 types according to Whiting. In the 

 so-called blotched tabby the stripes are 

 very wide, few in nimiber, and much 

 broken up. The common tiger pattern 

 has been described above. There is 

 also an almost uniform tabby in which 

 the stripes have become so narrow that 

 they lose their distinctness for the most 

 part. Ghost patterns of all of these 

 types may be seen in different black 

 kittens. 



CLASSIFICATION OF VARIATIONS 



The possibility of combining any 

 grade of banding with any kind of strip- 

 ing indicates that independent kinds of 

 factors are involved. Before consider- 

 ing the mode of inheritance, we may 

 attemjit to classify the factors involved. 

 Fr)r this purpose a comparison of an 

 ordinary- black-and-yellow tortoise shell 

 with a tabby-and-yellow tortoiseshcll is 

 very instructive. The black and the 

 tabby spots differ just as do solid black 

 and solid tabby coats. The yellow 

 areas in the two coats, on the other 

 hand, will be found indistinguishable. 

 This, however, is not because the 

 pattern is lost. The yellow of the black- 

 and-\-cllow tortoiseshcll, as well as of the 

 taljby-and-yellow kind, shows distinct 

 alternate stripes of orange and cream. 

 In the tabby tortoise, the dark stripes of 

 the tabby areas continue as orange 

 strii);_'s in the yellow part of the coat 

 while the light, banded stripes f)f the 

 tabby continue as cream in the yellow 

 part. 



The fact that there is no \isible dif- 

 ference between the black and the tabby 

 tortoiseshells in the yellow ])arts of th? 



fur shows that the banding factor, by 

 which alone these varieties differ, does 

 not act on color in general, but merely on 

 black. We may describe its action bv 

 saying that it has no effect on enzs-me I, 

 but inhibits enzvone II during part of the 

 development of the hair. Doncaster'^ 

 showed that tabby is dominant over 

 black and differs by only one factor. 

 Thus banding in cats appears to be in 

 every way comparable with the banding 

 of agouti-colored rodents and the domi- 

 nant factor may be put at once in 

 class 2ai. The variations in the width 

 of the bands may be due either to mul- 

 tiple allelomorphs of this factor as in 

 cases discussed in the papers on mice, 

 rabbits, and guinea-pigs, or they may 

 be due to subsidiary factors, which vary 

 the density of the black, which the 

 banding factor must inhibit; and_ similar 

 to those noted in guinea-pigs. 



The case is very different with the 

 stripes. The contrast in intensity of 

 the orange and cream stripes shows that 

 these differ in the activity of the fvmda- 

 mental color-producing enzyme, enzyme 



I. As there is little or no effect on the 

 intensity of black, the difference in 

 intensity is more like that due to factors 

 of class lb than of class las. Greater 

 activity of enzyme I, in the intense 

 stripes, does not, however, account 

 satisfactorily for the reduction or elimi- 

 nation of the bands in the latter, in 

 tabby cats. To take a ])arallel case, 

 reduction of the black-and-red agouti 

 variety of guinea-pig to the sejDia-and- 

 cream agouti variety by a factor of 

 class lb, has no appreciable effect on the 

 width of the red or cream l^and. It is 

 necessary to suppose that the dark 

 stri])cs have a greater activity of enzyme 



II, as well as of enzyme I, and that this 

 tends to neutralize the inhibiting action 

 of the banding factor. It will be seen 

 that this interpretation is identical with 

 that reached in the case of the agouti 

 patterns of rodents, if the whole l)ack 

 of the mouse, rat, rabbit, or guinea-i)ig 

 is com]iare(l to a dark stri]);; of the cat, 

 while the belly of the rodent is com- 

 pared to a light stri]K'. It should be 

 added, however, that there is also a 

 difference between the back and brlly 

 of cats, |)-obabl\' c)m])aral)le to the 



