LATENT ANCESTRAL CHARACTERS 163 



allelomorph in which the grey was permanently latent. It 

 would appear, therefore, that the essential features of Mendelian 

 inheritance are, not segregration and gametic purity, but the perma- 

 nent dominance of one alternative character and the permanent latency 

 of the other. 



278. Any number of similar instances might be quoted. Thus 

 when Mr A. D. Darbishire crossed fawn-and-white Japanese 

 waltzing mice v/ith albinos, the descendants exhibited a number 

 of colours which he classified as yellow, fawn colour, grey (pale 

 and dark, the colours of the wild mouse), black, lilac (pale and blue- 

 grey), and chocolate. 1 In this as in many other instances (e.g. 

 Stanley sweet-pea, quoted above) the allelomorphs from both the 

 crossed varieties were evidently highly compound, several latent 

 characters being revealed. The appearance of the black coloration 

 is particularly interesting. It has appeared also when pigmented 

 guinea-pigs and rabbits were crossed with an albino variety. 

 Thus Mr C. C. Hurst crossed Belgian rabbits which have yellow- 

 grey fur with angoras which are albinos. 2 Both varieties had 

 been pure-bred for at least eight and probably many more genera- 

 tions. The offspring all displayed " grey coats like that of the 

 common wild rabbit." Bred together the mongrels produced 

 wild greys, albinos, Dutch-marked, and black offspring. The 

 three latter types bred true when mated with their own kind. 

 They were pure recessives. The greys were pure and impure 

 dominants, with albino, black, or Dutch markings as the latent 

 element. There was never a return to the yellow-grey Belgian 

 coloration. A further series of experiments proved that the 

 black coloration as well as the grey and Dutch markings had 

 been introduced by the various albinos of the original cross. Now 

 albinism is a typical recessive character which, according to the 

 doctrine of unit segregration, should always be pure. Yet we see 

 that in mice, rabbits, and guinea-pigs it may be associated with 

 latent grey, black, or Dutch coloration. Plants furnish evidence 

 similar in kind, and even greater in quantity. Thus if two white- 

 flowered varieties of garden plants be crossed the offspring may 

 have coloured flowers. 



279. The reproduction of ancestral traits seems to afford 

 evidence that is decisive in favour of the hypothesis of alternate 



1 " On the Result of Crossing Japanese Waltzing with Albino Mice" Biometrika, 

 Jan. 1904. 



*" Experimental Studies on Heredity" Linnean Society's Journal, Zoology, 

 vol. xxix. 



