POLYZOA. 247 



conveniently selected for description, as having been more completely 

 worked out by Barrois than perhaps any other form. The seg- 

 mentation commences in the normal manner by the appearance of 

 two vertical furrows followed by an equatorial furrow, which divide 

 the ovum into eight equal spheres. The stage with eight spheres is 

 followed, according to Barrois, by one with sixteen, formed in a re- 

 markable manner by the simultaneous appearance of two vertical 

 furrows, both parallel to one of the original vertical furrows, so that 

 the segmentation spheres at this stage are arranged in two layers of 

 eight each. In the next stage segmentation takes place along two 

 fresh vertical planes, similar to those of the last stage, but at right 

 angles to them, and therefore parallel to the second of the two pri- 

 mitive vertical furrows. At the close of this stage there are thirty- 

 two cells arranged in two layers of sixteen each, and when viewed from 

 the surface each of these layers presents a regularly symmetrical pat- 

 tern. Up to the stage with sixteen cells the two poles of the egg, 

 separated by the primitive equatorial plane of segmentation, remain 

 equal, but during the stage with thirty-two cells a peculiar change 

 takes place in the character of the cells at the two poles. At the one 

 pole, which will be spoken of as the oral pole, the four central cells 

 become much larger than the twelve peripheral cells. 



The stages immediately following are still involved in much ob- 

 scurity, and have been described very differently by Barrois in his 

 original memoir (No. 298), and in a subsequent note (No. 307) \ In 

 the latter he states that the four large cells of the oral face become 

 enclosed by the division and growth of the twelve peripheral cells. 

 They are thus carried into the interior of the ovum ; and there divide 

 into a central vitelline mass the hypoblast and a peripheral meso- 

 blastic layer. 



The eight peripheral cells of the aboral pole divide vertically, and, 

 owing to the eight central cells at the aboral pole dividing trans- 

 versely so as to form a protuberance on the aboral surface, they con- 

 stitute a transverse ring of large cells round the ovum, which become 

 ciliated and constitute the main ciliated band of the embryo, cor- 

 responding to the ciliated band at the edge of the vestibule of the 

 entoproctous larvas. They divide the embryo into an aboral and an 

 oral region. The central part of the aboral projection forms a 

 structure which I shall speak of as the ciliated disc. It probably 

 corresponds with the ciliated disc in the Entoprocta. An invagina- 

 tion is next formed on the oral surface, which gives rise to a sack 

 opening to the exterior (fig. 131, st.). This was originally held by 

 Barrois to be the stomach; but Barrois now prefers to call it 'the 

 internal sack.' To my mind it is probably the stomodsenm. The 

 embryo has become in the meantime laterally compressed, and, at 



1 The note (No. 307) refers in the first instance to the changes in the larvze of the 

 Chilostomata, but the similarity of the larvae of the Ctenostomata to those of the 

 Chilostoinata renders it practically certain that the corrections, in so far as they apply 

 to the one group, apply also to the other. 



