INTERNAL SECRETION OF OVARY 265 



there is no definite histological or functional distinction between 

 the different formations which are derived from the follicle. 

 I think that at all the stages of follicular development, beginning 

 with a very early one up to a ripe Graafian follicle and to a corpus 

 luteuni, the follicle can produce an internal secretion, but its 

 activity varies according to the respective stages. A similar 

 view is held by Bucura (1913), by Sand (1918, 1921), and by 

 Athias (1923); but I cannot agree with Bucura that the inter- 

 stitial cells of the ovary act only as a store for the hormones 

 produced in the cells of the follicle. There is no evidence for 

 such an assumption. This conception of Bucura follows from 

 his belief that there is a sharp histological distinction between 

 the "follicle " and the " interstitial tissue," the first consisting of 

 epithelial cells of granulosa, the second of connective cells in 

 the theca interna. But we have seen that such a distinction 

 has no real existence. 



When we have realised that, in regard to function, only 

 quantitative differences exist between the different stages of 

 follicular development, we can understand why cyclical uterine 

 changes, as in the premenstrual phase, can begin before a ripe 

 Graafian follicle or a corpus luteum is present. The ruptured 

 and vascularized follicle on the one hand, and the premenstrual 

 changes in the mucosa on the other, represent only stages in 

 the chain of causation. Evidently the follicles at different 

 stages of development differ from one another as to this endo- 

 crine activity only quantitatively; this explains why after 

 irradiation by X-rays, when an increased follicular atresia is 

 caused, and when an increased number of interstitial cells is 

 activated, the same hormonic effect can be obtained as by means 

 of a corpus luteum. By assuming that only quantitative 

 differences exist in endocrine activity at different stages of 

 foUicular development, we can also understand why menstrua- 

 tion may occur in some animals (as with monkeys) in the 

 absence of either a ripe Graafian follicle or a corpus luteum. 



The same is no doubt true also for the heat periods of the 

 lower mammals. It does not seem to me justifiable to centre 

 the whole problem around the question as to whether a ripe 

 Graafian follicle or a corpus luteum is present at the time of the 

 prooestrum or not. It is necessary to determine rather whether 

 at the prooestrum those changes take place in the ovary, which 

 could be considered as signs of increased activity on the part 



