CTENOPHORA 155 



important evidence which militates against the derivation 

 of the Ctenophoi^a fi'om the comparatively highly organized 

 and specialized Cladonemidse. 



It appears to us, however, as if it were not these difficulties 

 alone, but rather grounds of a more general nature, that 

 have been influential recently in causing several writers (R. 

 Hertwig, Lang, and Hatschek) to concede a more independ- 

 ent position to the Ctenophora. We have learned to consider 

 the attached polyp, a Hydra-like creature, as the ancestral 

 form and archetype of the Cnidaria, and believe it probable 

 that in this instance the radial structure has been developed 

 in connection with the attached mode of life, as is so 

 frequently the case. Wherever among the Cnidaria pelagic 

 species occur, we can easily refer them to attached forms, 

 from which they have descended. The medasa must there- 

 fore be looked upon as a modified polyp that has aKained 

 the power of free locomotion. All these pelagic Cnidaria 

 have, however, as evidence that they ai-e organisms 

 secondarily derived from an attached form, the following 

 characteristics: () the loss of the general coat of cilia and 

 the development of new locomotor organs depending upon 

 muscular action ; (2) little tendency on the part of the ex- 

 umbrellar portion of the bell to produce any organs what- 

 ever. This latter feature of the Cnidarian medusa is con- 

 nected with the original function of its apical pole as a point 

 of attachment and the former comparatively unexposed 

 and unimportant position of the ex-umbrellar side, which 

 corresponds to the lower surface of the cup of the polyp. 



The Ctenophora do not exhibit any polypoid stage in their 

 ontogeny. We would not ascribe too great an importance 

 to the absence of this, for the ontogeny of Geryonia and 

 Pelagia furnishes us with an example of how quickly in an 

 abbreviated development just this stage is blotted out past 

 identification ; it is therefore not the circumstance that 

 the ontogeny of the Ctenophora contains no indication of an 

 attached stage, but rather the existence of certain prominent 

 features of organization in the Ctenophora, which makes it 

 seem to us probable that an attached stage has never been 

 interpolated in their series of ancestors. A system depend- 



