SEX AND SEXUAL CHARACTERS 467 



results of every experiment of a more complicated kind. Its 

 experts are continually revising these formulae, and in their 

 most recent improvements they make use largely of purely 

 theoretical and imaginary characters, such as colour factors and 

 inhibiting or epistatic factors which at present have no more 

 perceptual existence than the ether. In fact, many of the Men- 

 delian formulae given in Bateson's recent book are not merely 

 theoretical, but metaphysical — the qualities of a thing are 

 separated from the thing itself. We cannot see the redness of a 

 rose without the rose itself, although in metaphysics we may 

 distinguish an object from its qualities. But the Mendelian 

 finds no difficulty in discussing the inheritance of the bifidity of 

 the comb of the fowl in one gamete and the comb itself in 

 another, or the density of colour in one gamete and the colour 

 itself in another — until we are reminded of Lewis Carrol's 

 description of the smile of the Cheshire cat, which remained 

 behind after the cat had vanished. All this may be perfectly 

 legitimate, if only it is admitted that it is all pure theory — the 

 endeavour which all science makes to arrive at concepts which 

 make the percepts intelligible. But Mendelism makes pre- 

 tensions beyond all this which are not legitimate, and against 

 which it is time to protest. Mendelism deals only with characters 

 which exist. It is true that it shows how new varieties may arise 

 from the omission or addition of existing characters, as the result 

 of crossing ; but it does not maintain that new characters arose 

 in this way. It assumes that all new characters arose as 

 mutations in the gametes ; and there is nothing in the methods 

 or doctrine of Mendelism to afford any evidence of the truth of 

 this assumption. In the passage above, quoted from Mr. Don- 

 caster, we have an example of the unwarrantable pretensions of 

 Mendelians. To assert that the Mendelian theory of sex which 

 he adopts explains the effects of castration is absurd. On the 

 Mendelian view, there is no reason whatever why castration 

 should have any effect at all. This Mendelian formula merely 

 supposes that both the gametes w r hich unite to give rise to a 

 male organism, bear the male sex-character. If we suppose that 

 the latter term includes the male secondary characters, it follows 

 that these develop by heredity in the male organism, and that 

 the female sex-characters are entirely wanting, and cannot 

 develop. But there is absolutely no connection between this 

 conception and the fact that the secondary characters fail to 



