12 SIDNEY F. HARMER. 



cells filling up most of the interior of the embryo. This has been 

 shown, for instance, by Barrois (4) and Ostroumoff (13) for Cyclo- 

 stomata, hy Vigelius (16) for Bugula and by Repiachoff (15) for 

 Bowerhankia. ■ 



The discovery of a well developed (though probably not func- 

 tional) alimentary canal in the larvae of the Ctenostomata {Alcyo- 

 nidium) and of the Gyclostomata (as shown by Ostroumoff) relieves 

 us from the necessity of supposing that Cyphonautes is really an 

 archaic larva : — an assumption which is very difScult to reconcile 

 with current views as to the highly specialized character of the 

 -Cheilostomata. 



Cyphonautes may perhaps be regarded as a much modified 

 type of larva in which the alimentary canal has been preserved in 

 a functional form (owing perhaps to a longer larval life than in 

 other Polyzoa ?), whilst the oral face has become transformed into 

 an atrium in which are situated the pyriform organ and the 

 internal sac. 



It is at present hardly possible to affirm that the characters of 

 the Alcyonidium-laxva, are retained throughout the group of the 

 ■Ctenostomata, since the accounts given by Barrois (1, 3 and 4) and 

 Repiachoff (15) of the larvae of this division of the Polyzoa are by 

 no means concordant. 



Barrois (3) has stated that the Ctenostomata (in which group it 

 may be assumed that he does not include Alcyonidium) are 

 characterized by the absence of the internal sac. Repiachoff (15) 

 lias, however, given a description, with figures, of the development 

 of Bowerhankia which may perhaps tend to show that this structure 

 is not really absent in the larva of this genus. Repiachoff's 

 description and figures are extremely difficult to understand 

 thoroughly, but if we accept his statements, the larva of Bower- 

 hankia is very different from that of other Ctenostomata. A 

 comparison of Repiachoff's paper (and especially of the series 

 of sections of the larva figured on PI. IV) with my own pre- 

 parations of Alcyonidium leads me to suspect that Repiachoff's 

 identifications of the surfaces of his larvse were not accurate. 



A further investigation of the larva of Bowerhankia is needed 

 to clear up its structure. I will at present merely point out that 



