CN IDARIA 



65 



medusa by the larval hydropliillium, whereas the tentacle would have 

 to be explained as the only remaining marginal tentacle of the medusa. 

 This primary individual of the Siphonophore stock, referable, from 

 Haeckel's point of view, to the fundamental form of a Hydromedusa, 

 would in the language of the medusa theory (see pp. 70 and 73) be 

 called a viedusom, and the corresponding larval form a Sip]ionula-stage. 



Only portions of the development of the Vellelidce (TracheophysiB, 

 Chun) are as yet known. A number of young larvje have been de- 

 scribed by A. Agassiz (No. 52), Haeckel (No. 70), Bedot (No. 53), and 

 Chun (No. 57). The youngest larval stage observed by Haeckel, perhaps 

 belonging to the developmental cycle of Porpita, was named Discomila ; 

 it exhibits a distinctly octoradial structure (Fig. 27). From the under- 

 surface of the discoid stem there hangs a central polypite (c), the cavity 

 of which is united by means of eight radial canals to a peripheral ring- 

 canal and eight simple tentacles {t). In the apical part of the gelati- 



FiG. 27. — Two J)is ■odida-stag'es (after Haeckel). ^.younger stage, seen from 

 the upper side; B, somewhat older stage with ramified tentacles, seen from the 

 lower surface ; p, pneumatophore ; q, buds of the blastostyle; c, central polypite 

 with mouth-opening; t, tentacle. 



nous disc is found a central lentiform pneumatocyst (p), surrounded by 

 a circle of eight radial air-chambers, each of which opens to the ex- 

 terior by means of a dorsal i^ore. Haeckel interprets this stage as the 

 ontogenetic rej^roduction of an octoradial ancestral form which would 

 have to be sought for among the Trachomedusffi ; consequently all the 

 Siphonophores assignable to this group must be separated as an inde- 

 pendent sub-class (Disconanthie) from all the remaining ones, which are 

 descended from a bilateral ancestral form, of which the Siphonula larva 

 is the expression. In opposition to this hypothesis of the diphyletic de- 

 rivation of the Sii^honophora, Chun has contended that the octoradial 

 Disconula-stage is probably preceded in the development of the Por- 

 pitidae and Vellelidse by a bilateral Siphonula-stage. Young Rotariic 

 (larvflB of Vellelidfe), still with a simple, unchambered pneumatophore, 

 exhibited four bilaterally arranged tentacles, for a larger tentacle and 

 K. H. E. F 



