472 SCIENCE PROGRESS 



assumption ; namely, that the character in question is coupled 

 or associated with the female sex-character ; but this coupling 

 is not complete or absolute — the character can be present in 

 dominant spermatozoa, and so appear in males. In Lepidoptera 

 such as Ocneria dispar, on the other hand, we may assume that 

 the coupling is complete, and therefore when the spermatozoon 

 is dominant the male character only develops in the males, and 

 conversely the female character in the females. It is not, 

 however, proved at present that all secondary characters in 

 Lepidoptera are determined at fertilisation, and are unaffected by 

 castration. It must be remembered that the characters and 

 organs which are dependent on the sexual hormones in Verte- 

 brates and Crustacea have some function in the relations of the 

 sexes, and in Lepidoptera such functions as these — connected 

 with fighting, courtship, and care of the young — seem to be 

 generally absent. It would be interesting to ascertain whether 

 the highly developed special organs in male beetles, which are 

 certainly sexual in function as well as in limitation — such as the 

 mandibles of the male stag-beetle — were also independent of the 

 function of the gonads. Their varying development in different 

 individuals indicates very forcibly that they are not. Even in 

 Lepidoptera, the absence of wings in the female only in some 

 species is probably due to their disuse in this sex only ; and in 

 this case it would be expected that the wings would be more 

 developed if the ovaries were removed in the caterpillar. There 

 is much room for further investigation. 



Thus it seems to me we arrive in considering sex-limited 

 characters at the same distinction which I have insisted upon in 

 other characters, a distinction which is persistently ignored by 

 the Mendelians. This is in general the distinction between 

 adaptive and non-adaptive characters — which is, I believe, the 

 same thing as the distinction between continuous and dis- 

 continuous variations. Mendelians assume that all characters 

 arose as mutations — that all hereditary variations are discon- 

 tinuous. To me it seems impossible to maintain that the organs 

 which distinguish the frog from the tadpole — the amphibian from 

 the fish — arose as mutations, since they develop continuously in 

 the individual. The lungs develop gradually pari passu with the 

 change from aquatic to aerial respiration. The same gradual 

 development (corresponding with a change of external condi- 

 tions) is seen in the metamorphosis of the flat-fish and of the 



