THE ULTIMATE SEX-ELEMENTS. 9 1 



In coelenterates the phenomena are of much interest ; the origin of the 

 sex-cells is very diverse. Some time ago considerable emphasis was laid, 

 by E. van Beneden and others, on the fact that, in certain Hydrozoa, "the 

 ova are derived from the endoderm, and the sperms from the ectoderm." 

 Thus Gegenbaur, accepting this, remarks that in such cases "the 

 endoderm is the female, and the ectoderm the male germinal layer." 

 Such a generalisation, if established, would be plausible enough, seeing 

 that the inner or endoderm layer is the more nutritive or anabolic of the 

 two. A controversy however soon arose, the result of which was to over- 

 throw the generalisation. In hydra, we have already noticed that both 

 products arise from the ectoderm ; the same was shown by Ciamician to be 

 true of Ttibularia mesembryanthenium ; while in the Etidendrmm ramosum 

 the ova appeared to arise from the ectoderm^ and the male elements from 

 the endoderm^ the very reverse of Van Beneden's conclusion. The matter 

 was settled, so far as the general facts are concerned, by Weismann, who 

 established the fact of active migration of the elements from one layer to 

 another. He has since been followed by other investigators, {a) The sex- 

 elements, both male and female, may appear first in the endoderm, whether 

 they originate there or not, and from this inner layer they migrate to the 

 ectoderm, where they ripen. {h) In rare cases they even ripen in the 

 endoderm. (<:) Very commonly the sex-cells originate in the ectoderm and 

 ripen there, or they may pass thence into the endoderm and back again to 

 the ectoderm, {d) In the medusa of Obelia, the ova appear to ripen partly 

 in° both layers. These facts, a convenient summary of which will be 

 found in Hatchett Jackson's erudite edition of Rolleston's " Forms of 

 Animal Life," show plainly enough how varied are the origin and history 

 of the sex-cells in these forms. 



The colonial hydroids typically produce well-marked reproductive 

 individuals or sexual zooids, set free as " swimming-bells " or medusoids 

 (in a process to be afterwards described under " Alternation of Genera- 

 tions "). In these the reproductive elements are typically developed. But 

 in varying degrees these medusoids have degenerated, and are frequently 

 not only not liberated, but lose their characteristic features, and become 

 mere reproductive buds. In these buds the sex-cells are normally 

 developed. But it very frequently happens that they arise more or less in 

 the body of the asexual vegetative hydroid. They ripen early, and sub- 

 sequently migrate to their proper place ; the asexual stage incorporating 

 more and more of the originally separate sexual generation. Weismann has 

 emphasised the value of this early ripening as an advantage to the race, 

 lessening the danger of its extinction; and this has doubtless to be con- 

 sidered, though it can hardly be regarded as a physiology of the facts. 



§ 7. Early Separatio7i of Sex- Cells. — Having noted the 

 general fact of mesodermic origin, and some of the interesting 

 phenomena observed in coelenterates, we shall not further pursue 

 the subject except as regards one question, the period at which 

 the reproductive cells make their appearance. This is sometimes 

 early, sometimes late ; and it is not yet decisively known how 

 widely early separation occurs, nor how far the fact is of much sig- 

 nificance. The question will have to be discussed in the volume 

 treating of heredity ; only a brief reference is here possible. 



