Effects of Castration 135 



either remain undeveloped or if developed are shed prematurely and are not 

 replaced, or replaced ouly by incomplete growths. But structures which are 

 common to both sexes — in species, for instance, in which both sexes possess horns — 

 are not modified by castration. In Arthropoda the correlation between the 

 generative glands and the secondary sexual characters (which in many species are 

 even more marked than in Vertebra ta) does not hold good. Experiments upon 

 caterpillars show that removal of the generative glands has no influence on the 

 development of the male sexual characters of the imago ; nor do the glands, if 

 transplanted into individuals of the other sex, affect the secondary sexual characters 

 or instincts of the host. This need not be taken to mean that the secondary 

 sexual characters in these animals are not the result of an internal secretion, but 

 may be interpreted by supposing that some organ other than the generative glands 

 furnishes the internal secretion which produces those characters. 



In Vertebrata, at any rate, there can be little doubt that the internal secretions 

 of the generative glands are an important if not the chief factor in determining 

 the development of the secondary sexual characters. And that this development 

 is independent of the normal functions of the generative glands is shown by the 

 fact that ligature of the vas deferens has no effect in preventing it. Moreover, 

 transplanted testes and portions of testis (in which the generative cells themselves 

 may completely disappear) have been found capable (in birds) of preventing the 

 results of castration : the comb, wattles, spurs, etc., of the cock being developed 

 in the usual way (Shattock and Seligmaun and others). Xussbaum s experiment 

 on the effect upon the development of the thumb-pad of grafting pieces of testis 

 from another frog into the dorsal lymph sac of a castrated male frog also points 

 to the existence of an internal secretion of the testicles in these animals. 



The Nature and Source of the Testicular Autacoid 



It seems certain, therefore, that the development of the secondary- 

 sexual characters in the male sex is dependent upon an internal secretion 

 of the testicle, and it is highly probable that it is yielded not by the gener- 

 ative cells (gonads) but by the interstitial cells. In eryptorchids, and also 

 after experimental ligature of the vas deferens, in both of which, as we 

 have seen, the seminiferous epithelium is atrophied but the interstitial 

 tissue is well developed, the secondary sexual characters and sexual desire 

 are normal. Successful implantation of the whole or part of a testicle in 

 a young castrated animal is also followed by development of those char- 

 acters, although in most cases the seminiferous epithelium of the graft dis- 

 appears. Loewy found male secondary characters developing in caponised 

 cockerels fed with testicle-substance. Bouin and Ancel state that extract of 

 testicle freed from all morphological elements may, when injected, produce 

 a similar result. In support of the theory that an autacoid which affects 

 the development of the secondary sexual organs and characters is formed 

 by the interstitial cells, they have found that if one testicle is removed 

 from a rabbit and the remaining one has the vas ligatured its interstitial 

 tissue becomes hypertrophied. Further, it is known that if the testicles 

 are exposed to the action of X-rays the seminiferous epithelium undergoes 



