the Cape Cephalodiscug. 245 



C. dodecalophus to which I have referred in my ' Siboga ' 

 Report (2, p. 102), the first record of the characters of 

 the larvae was given by Andersson (5), who observed free 

 planula-like larvae in the living condition. I have myself 

 given some account of the early development of C. levinseni 

 and C. gracilis. Somewhat later, Andersson (6) published a 

 fuller account of the observations referred to in his preliminary 

 note, recording a number of facts with regard to the early 

 development of G. dodecalophus, G. incequatus, Orthoecus 

 solt'dus, and 0. densus. Schepotieff (7) has described the 

 early development of C. indicus ; while Braem (8) has pub- 

 lished a paper, based on the results recorded in the memoirs 

 above cited, in which he attempts to demonstrate that the 

 similarity previously noticed between the larvae of Gephalo- 

 discus and those of Ectoproct Polyzoa is an indication of 

 genetic affinity, and that it is accompanied by fundamental 

 resemblances in structure between the two types of larvae. 

 I may remark, in passing, that I am not satisfied that Braem's 

 comparisons will hold good ; and in particular this author 

 does not seem to me to have taken sufficiently into account 

 the fact that the larvae of Ectoprocta show no trace of the 

 Balanoglossus-Wke disposition of the ccelomic cavities which 

 appears to be demonstrated in the larvae of Cephalodiscus. 



The accounts which have been given of the development 

 in the memoirs already noticed are in close agreement with 

 one another; and the following facts appear to be clearly 

 established : — 



(i.) The egg of Cephalodiscus is of considerable size — 

 reaching a diameter of 680 /j, in Orthoecus soltdus — and it 

 contains a large quantity of yolk. 



(ii.) Segmentation is complete and leads to the formation 

 of a gastrula-like stage, although the mode of origin of the 

 inner layer has not been definitely established. The lumen 

 of the archenteron is very small, in correlation with the large 

 amount of yolk. 



(iii.) The yolk is present in the cells of the inner and outer 

 layers in the earlier stages of development; but it later 

 disappears from the outer layer, persisting in large quantity 

 in a central mass representing the wall of the archenteron, 

 the cavity of which remains very small. 



(iv.) The wall of the archenteron is continuous with the 

 outer layer, in the later stages observed, near the posterior 

 pole of the embryo. This region may be regarded as the 

 blastopore, and it seems probable that it gives rise to the 

 anus. 



(v.) The free-swimming larva and the later embryos possess 

 five body-cavities arranged like those of the adult— namely, 



