ZOOLOGY AND BOTANY, MICROSCOPY, ETC. 517 



side of which a gastric tube with filaments and a group of genunas is 

 formed. After some notes of and criticisms on later studies and 

 especially on those of Chun, the author comes to the conclusion that the 

 Siphonophora arise from a stage in the development of the Hydroidea 

 not unlike that represented by the Hydractinidse ; this form did 

 not, however, become fixed, but continued to lead a free-swimming 

 pelagic life, and so became capable of further development. It is 

 probable that the cell-material collected at the aboral pole of the 

 growing polypoid body formed a zone of gemmation from which, 

 without stem or stolons being formed, polyps and medusoid buds could 

 be given off. If the oral circlet of tentacles remained arrested in 

 development one or another polyp-bud might, if it elongated and did 

 not become provided with a mouth, become converted into a tentacular 

 appendage which would take on the function of a " fishing-line " 

 (Fangfaden) ; a medusoid bud would become converted into and be 

 gradually set free as a sexual form, while its place as the nectocalyx 

 would be taken by a fresh medusa formed by gemmation. It is almost 

 certain that this sexual animal had at first the marginal filaments and 

 eye-spots, which were only completely lost during the later processes 

 of adaptation, in just the same way as the Mnestra which is attached to 

 PhjUirhoe has lost the marginal organs of the medusoid body. In the 

 course of further changes we must suppose that the first developed 

 medusoid bodies would lose the power of developing generative 

 elements and would persistently retain the function of the nectocalyces 

 or of the hydrophyllia. The sterile medusa would either become 

 bell-shaped or have a deeply excavated swimming-sac, or cartilaginous 

 covering pieces with atrophied subumbrella, and subumbrellar vascular 

 apparatus. 



The oldest Siphonophora must have gone through a number of 

 changes before they became converted into the present Calycophora, 

 while the development of a pneumatophore was necessary for forms 

 like the Physophoridae ; it is almost certain that this divergence did 

 not commence at a simple, but at a very advanced, stage. The appear- 

 ance of the first-mentioned hydrostatic apparatus must be taken as a 

 character of great importance, and we may group together those forms 

 — Physophoridae, Physalid^e, and Discoidea — which possess it under 

 the name of the Pneumatofhora. The organ would seem to have 

 been developed, phylogenetically, at a stage much later than ontological 

 evidence alone would lead us to expect ; there are many points of 

 affinity between Hippopodius, which is a Calycophore with a number 

 of cells, and the physophorid Apolemia. The problem of whether 

 the pneumatophora was a " neomorph " or a metamorphosed bud 

 cannot yet be exactly determined, though the latter view is the more 

 probable. The fact that this pneumatophore has an opening to the 

 exterior explains its great importance as a hydrostatic organ, and the 

 correlated suppression of the nectocalyces. We may pass through 

 the simpler stage of BhizopTiysa with an elongated stolon, to the more 

 metamorphosed Physalidae, which are bladder-shaped ; and from them 

 to the Velellidse, the form of which is discoid ; these are the most 

 aberrant of the whole group. 



