No. 6.] THE MORPHOLOGY OF THE ACTINOZOA. 269 



ance; (3) that in the same embryos the streaks make their 

 appearance on mesenteries that are not connected in any way 

 apparently with the ectoderm, and (4) that in bud embryos 

 the ciliated bands appear before the glandular streaks. 



It seems to me from these facts that the ciliated bands must 

 be regarded as being ontogenetically distinct from the glandu- 

 lar st'reaks. The two have been very generally regarded as 

 different parts of the same structure, but this idea is, I think, 



untenable. 



If they be recognized as distinct structures, there are no 

 a priori reasons for regarding both as products of the same 

 germ layer. The question of the origin of the filaments, 

 whether from the ectoderm or from the endoderm, is one that 

 has been frequently discussed, and with very varying answers. 

 The majority of authors have regarded both parts as ectodermal 

 or as endodermal, E. B. Wilson having been the first, from his 

 studies on the Alcyonaria, to point out the probability of the 

 development of the ciliated bands in these forms from the ecto- 

 derm and that of the glandular streaks from the endoderm. In 

 my studies on the development of the hexactinians ('91), I 

 reached the same conclusion, and the evidence presented above 

 seems to point to a similar story in the zoanthids. 



However, there is a more fundamental consideration at the 

 base of all questions as to ectodermal and endodermal origin 

 in the Coelentera. Is there sufficient fixity of the germ layers 

 in these forms, whether the layers be regarded from the mor- 

 phological or the physiological standpoint, to warrant the impor- 

 tance which has generally been attached to them >. The germ 

 layers have evolved ; like other structures, they have had a 

 phylogeny, and it may be remarked that just as in other struc- 

 tures we find discrepancies between the phylogenetic and onto- 

 genetic development, so too we may expect and undoubtedly 

 do find discrepancies between the ontogeny and the phylogeny 

 of the o-erm layers. It has generally been accepted that the 

 Coelentera represent a stage in the phylogeny of the germ lay- 

 ers two of them being fully differentiated; indeed, Huxley s 

 homoloo-y of the coelenterate ectoderm and endoderm with the 

 epiblast%nd hypoblast of the embryologists may be regarded 



