1884.] Ill [Hyatt. 



to Prof. L. Agassiz's homology of the actinostome with the pro- 

 boscis or hypostome of the Hydrozoa. It seems to us that the 

 composition of its walls and the mesenterial filaments cannot be 

 accounted for unless we admit their identity in both of these 

 branches. The ectodermic origin of the numerous folds in the 

 actinostome and the upper parts of the mesenteries, and their cir- 

 culatory functions, described by Wilson, are apparently explicable 

 on the supposition that they are similar to the fringes and tentacles 

 of the actinostome (hypostome) in Hydrozoa. The comparison 

 between the septal chambers and the coelomic sacs of the embryos 

 of Amphioxus and Peripatus is rendered very complete, and even 

 the mode of formation of the middle gut or mesenteron of the 

 Articulata and Vertebrata is suggested by this remarkable inves- 

 tigation. By combining all the views which have been cited we 

 can then understand Van Wijhe's two neuropores as corre- 

 sponding anterior and posterior tubes belonging to the primitive 

 neural canal, which would thus become homologous with the 

 modified actinostome of some ancestor of the Vertebrata, the 

 hypophysis and blastoneuropore as remnants of tubes which 

 belonged to the period when the actinostome began to be divided 

 from the region of the mesenteron below, but was still connected 

 therewith. The stomodeum and proctodeum, however, would 

 have to be considered as independent invaginations comparable 

 with the similar ones in the Invertebrata, the former becoming 

 the mouth and the latter fusing with the blastoneuropore and 

 becoming the anus. When facts favoring the independent appear- 

 ance of these invaginations in the Invertebrata are so strong, and 

 the simultaneous presence of the three anterior invaginations in 

 the embryos of Petromyzon, as given by Dohrn, is admitted, it 

 certainly appears to be incorrect to allow oneself to imagine 

 that any of them are necessarily derivatives of the differentiated 

 tubes of the actinostome and necessarily parts of the blastopore. 

 Hubrecht's homology of the hypophysis and the proboscis of 

 nemerteans (Quart. Journ. Micr. Sci., vol. in, p. 355) and of 

 the notochord, which is supposed to have been primitively tubu- 

 lar, with the sheath of the proboscis, if true, would make the 

 presence of a vertical canal in the anterior end of the body 

 unnecessary. In that case the hypophysis could be considered as 

 connecting directly with the tube of the notochord (Lieberkuhn, 



