6 SIDNEY F. HARMER. 



cells. It extends into the lateral regions of the embryo, where it 

 passes further forwards than in the middle line; its anterior 

 border is thus markedly concave, as figured by Barrels (1) in many 

 genera of Cheilostomata and Ctenostomata. The mantle-cavity is 

 lined by a very high epithelium. 



The function of the pyriform organ is by no means clear. 

 RepiachofF(15) states that in Tendra a mass of cells is segmented 

 off from the embryonic hypoblast in front of the mouth, and sup- 

 poses that this mass represents the hypoblastic vesicle described 

 by Hatschek in the embryos and stolons of Pedicellina. It ap- 

 pears to me probable, however, that the region which corresponds 

 in Alcyonidium to that including Repiachoff's supposed hypoblastic 

 vesicle in Tend^-a is occupied by a mass of nervous tissue which 

 constitutes the brain of the larva. In spite of this, it is still 

 possible to compare this region with the " dorsal organ " of 

 the Entoprocta, since there are reasons {cf. 5 and 6) for doubting 

 the existence of hypoblastic elements in the "dorsal organ" 

 and for regarding it, on the contrary, as an important nerve- 

 centre. 



Fig. 4, which illustrates the structure of the region in question 

 in Alcyonidium, is a transverse section through the anterior part of 

 an embryo of about the same age as that represented in fig. 3, the 

 ciliated ring {c.r.) and mantle-cavity {rn.c.) having precisely the 

 same arrangement as in that figure. In the middle of the ventral 

 surface is seen the cup-shaped depression which constitutes the 

 pyriform organ; at the sides of the latter, as far as the ciliated 

 ring, the epiblast is thick, with finely granular protoplasm and few 

 yolk-spheres. 



As in the preceding figures, there is no definite body-cavity, 

 although certain irregular spaces occur at intervals in the 

 mesoblastic structures. The middle of the section is occupied by 

 a large development of fine fibrils, bounded laterally by masses of 

 nucleated protoplasm without yolk-spheres, and these masses 

 appear to be continuous with the dorsal epiblast at the two sides of 

 the middle line. I shall provisionally assume that the structures 

 just described are of nervous nature, and that they represent the 

 brain of the embryonic Alcyonidium. 



