HYDROIDA 



structure of the hydropolyps is throughout quite houiogeneous and according!)' affords no holds of 

 use to the systeniatist. A careful study, however, of the slight information occurring in the literature as 

 to the structure of the polyps, will show that the old view is wrong. In this connection it will be 

 sufficient to refer to the excellent passage written by Kvihn (1913), Die x\usbildung des Polypenkorpers, 

 in which are mentioned several examples of the heterogeneous structure of the polyps in the various 

 groups of hydroids. 



Nevertheless Kiihn himself (1913), in constructing his system of the hydroids, has, according 

 to the old practice, left this fact almost wholly out of account. I will afterwards come back to the 

 subject and point out how unreasonable it is to set aside this part of systematics even though a 

 thorough investigation of the anatomy of the polyps may seem almost impracticable because of the 

 state of preservation in which the hydroids generally occur in the materials of the great expeditions 

 and collections. In fact, it is confirmed that Levinsen (1893) is right, emphasizing that the systeniatist 

 in treating of hydroids should rather lay stress on all about the nourishing individuals than on the 

 varying development and organisation of the generative individuals. 



a. The Hydroid Gonophores bearing on classification. 



Our knowledge of the hydroids has advanced a vast stride ahead owing to the thorough 

 researches of Kiihn into the development and the organisation of the gonophores (1910). As to the 

 importance of the gonophores to systematics, it has lately (1913) been asserted by the same author 

 that no weight whatever can be given to the various gonophoral development, but that particular 

 stress should be laid on the structure of the full-grown gonophores. On account of difference of structure 

 he distinguishes between fourt)pes of sessile gonophores: eumedusoids, cryptomedusoids, heteromedusoids, 

 and, finally, the simple gonophores, called styloids (Bonne vie 1898). The eumedusoid gonophores, indeed, 

 retain the structure of the medusa throughout, and occasionally breaking awaj' [Campanularia iufcgra 

 Mc. Gillivr. — Agastra mira Hartlaub) occur like defective medusae. The structure of the crypto- 

 medusoid gonophores is more reduced. Certainly they keep their endocodon and their umbrellary cavity, 

 but have a single-layered umbrellary endoderm; however, also of cryptomedusoids occur exceptional 

 instances breaking away, as in Pachycordylc Weisiiiai/iii Hargitt. The heteromedusoids are entirely 

 wanting the umbrellary endoderm, and the inner umbrellary ectoderm is formed by delamination, not 

 by invagination of the outer ectoderm (endocodon). The simplest gonophores, finally, are without any 

 trace of medusoid organisation. These four types of sessil gonophores Kiihn (1913) regards as char- 

 acters important to the systeniatist for the division of h\droids into genera. 



Stechow (1913) in his important work on Japanese hydroids occupies another standpoint, 

 attaching more importance to the conditions of the gonophores. Thus he draws nearer to earlier 

 principles of classification, but at the same time he tries to make his systematic division more manage- 

 able by drawing out characters from the outward morphology of the colonies, thus effecting the 

 transition to the American school. The Americans, headed by Nutting, look on classification only as 

 "a matter of convenience", and accordingly in their grouping.s, as a matter of fact, relinquish the idea 

 of constructing a natural system aiming at summing up and drawing by critical sifting of the char- 

 acters a skeletonlike picture of the phylogenetic affinities of the group of animals in question. Con- 



