66 INVEETEBRATA chap. 



converted into the body of the first hydroid polyp, at the apex of 

 which the mouth is formed ; but the broad end develops an ectodermic 

 invagination, the rudiment of the float, at the spot where the 

 attached base would naturally be looked for if we were dealing with 

 planulae of Hydromedusae. A single tentacle sprouts from the base 

 of the polyp, and above this, i.e. nearer the float, is a growing zone, 

 from which other polyps arise. 



In many Siphonophora certain of the medusoid buds lose their 

 genital cells, and even the manubrium, and become organs of 

 locomotion merely. These organs arise in the part of the growing 

 zone nearest the float, morphologically the most basal part. In one 

 group the adult relies on these modified medusae (nectocalyces) 

 alone for swimming, the float having disappeared ; and in such cases, 

 which we regard as the most modified of all, the endoderm cells in 

 the base of the planula secrete oil drops, and the first definite organ 

 to be formed in this region is a huge nectocalyx. 



To this theory of Korschelt and Heider there are opposed two 

 other theories, viz. the medusome theory of Haeckel (1888) and the 

 theory of Chun (1887). 



According to Haeckel, the whole Siphonophore colony is merely 

 a medusa which budded, as some few medusae are known to do. 

 Every person is supposed to be a modified medusa ; the bells of the 

 medusae are supposed to be represented by the translucent leaf-like 

 bracts, termed hydrophyllia, which many species possess ; and the 

 hydroid-like persons are their "manubria," which are supposed to 

 have migrated out from them through a slit in the bell. The violent 

 dislocations required by this theory belong to the period of imagina- 

 tive morphology. 



Chun agrees with Korschelt and Heider in regarding the 

 Siphonophore as a Hydromedusan colony, containing both hydroid 

 and medusoid persons ; but he regards the float as a modified medusa, 

 in which air has replaced water. It is, however, very difficult, if 

 not impossible, to picture a series of ancestors in which one of the 

 medusa bells gradually replaced its contained water by air. In other 

 words, Chun's hypothesis transgresses the law of functional con- 

 tinuity, which should be exemplified in any supposed phylogenetic 

 change. 



NARCOMEDUSAB AND TEACHYMBDUSAE 



The Narcomedusae aud Trachymedusae are usually stated to be 

 Hydromedusae, in which the egg develops directly into a medusa 

 without an 'intervening hydroid stage. A more correct statement 

 of the case would seem to be that the egg develops into a modified 

 hydroid person, which does not bud, but which, by the forma- 

 tion of a web, becomes directly transformed into a medusa. 

 The fact that both these groups are pelagic in their habit has 

 rendered the formation of a fixed budding colony of hydroids 

 impossible. Therefore the development is hurried on, and the first 



