340 Dr. E. A. Cockayne on the relation between 



ovaries and female accessory glands are present, but no 

 male internal organs are found. Other primary hybrids 

 occasionally show a similar condition. On the other hand, 

 a primary hybrid Ennomos bred by Harrison showed fully- 

 developed male organs including two testes, and yet parts 

 of the external genitalia and some external somatic 

 characters were purely female in appearance. 



Many secondary hybrid Saturnias (emperor moths) and 

 Bistoninae show a very coarse mosaic of male and female 

 somatic characters, including the external genitalia, though 

 their internal organs are entirely female. 



The secondary hybrid Amorpha daubi, Standfuss, 

 A. ocellatus $ X A. langi $, also has females replaced by 

 gynandromorphs or intersexes. This is also the case with 

 many mongrel Lymantrias, though in these the external 

 genitalia are intermediate between those normally found 

 in the two sexes. Here, again, we have several examples 

 of male secondary sexual characters where all the male 

 internal organs are absent and all the female present, 

 and one example of the opposite condition. 



The chief argument against the castration experiments 

 is that they can only be performed comparatively late in 

 the larval stage, and it has been suggested that at this 

 stage an internal secretion has been produced and already 

 acted on the tissues of the body in such a way that they 

 are able to attain to their full sexual differentiation, even 

 after the gonads have been removed. Such an argument 

 cannot be advanced in the case of halved gynandromorphs 

 in which the failure of development of the sexual organs 

 Avhen it occurs must take place at a much earlier period, 

 and still less can it be advanced in the case of the hybrid 

 gynandromorphs or " intersexes." 



The whole of the evidence derived from experimental 

 castration, and from the study of halved gynandromorphs 

 and of intersexes, seems to prove definitely that the 

 secondary sexual characters are in no way dependent on 

 the gonads or any other portion of the internal sexual 

 apparatus. There is, however, no proof to be derived 

 from them that some gland quite independent of the 

 sexual organs does not produce an internal secretion com- 

 parable to that of the interstitial cells of the vertebrate 

 testis and ovary. 



But if we examine the condition of affairs in mixed 

 and transverse gynandromorphs even this possibility seems 



