Another case is that of tortoise-shell, _i.e._ black and yellow cats. The tortoise-shell with very rare exceptions is female, the corresponding male being yellow, without any black colour. Doncaster found that a yellow male mated to a black female produced black male offspring and tortoise-shell females. When a black male is mated to a yellow female, the female kittens are tortoise-shell as before, but the males yellow. The Mendelian hypothesis which explains these results is that the male is always heterozygous, or has only one colour factor whether yellow or black, and transmits these colours only to his daughters, while the female has two colour factors, either _BB_, _YY_, or _BY_. Thus the crosses are:-- YELLOW male x BLACK female YO male BB female | \/ | | /\ | YB female BO male Tortoise-shell female BLACK male BLACK male x YELLOW female BO male YY female | \/ | | /\ | BY female YO male Tortoise-shell female YELLOW male The sex must be determined therefore by the spermatozoa, as in the case of colour-blindness, etc., in man, and the colour factor must always be in the female-producing sperm. SEXUAL DIMORPHISM It is obvious from the above facts that however interesting and important sex-linked heredity may be, it is not the same thing as the heredity of secondary sexual characters, and does not in the least explain sexual dimorphism. In the first place, the term sex-linked does not mean occurring always exclusively in one sex, but the direct contrary-- transmitted by one sex to the opposite sex--and in the second place there is no suggestion that the development of the character is dependent in any way on the presence or function of the gonad. The problem I am proposing to consider is what light the facts throw on the origin of the secondary sexual characters in evolution. In endeavouring to answer this question there are only two alternatives: either the characters are blastogenic-- that is, they arise from some change in the gametocytes occurring somewhere in the succession of cell-divisions of these cells--or they arise in the soma and are impressed on the gametocytes by the influence of the soma within which these gametocytes are contained--that is to say, they are somatogenic. That characters do originate by the first of these processes may be considered to be proved by recent researches, and such characters are called mutations. There can be little doubt that the so- called sex-linked characters, of which examples have been given above, have originated in this way, and that their relation to sex is part of the mutation. According to T. H. Morgan, it is simply due to the fact that the determinants for such characters are situated in the sex-chromosome. Morgan, however, also states that a case of true sexual dimorphism arose