ALCYONIID^S. 261 



division. This probably represents au early stage of sporogenesis. Very minute 

 zoochlorelho (fig. 16, c.) are also seen in the canals, but I have been unable to determine 

 the intermediate stages of spore formation. 



Yellow Cells. 



The more or less rounded cells of a yellow colour which occur in the ectodermal as 

 well as endodermal tissues are to be found in many tropical forms, but are more 

 numerous in the specimens of Sarcophytv/m (Pratt, 1903, fig. 4) than in those of 

 Lobophytum and Sclerophyt.um. They appear to be very abundant in certain 

 Madreporaria, and have been described and figured by Duerden (1903, p. 439). 



As small portions of yellow granular matter, similar in all respects to the contents 

 of the yellow cells, may be frequently observed in the zoochlorellse, it is suggested 

 that the yellow cells represent an early stage of sporogenesis. The presence of yellow 

 granular matter may be due to the breaking up of the chromatophores, but this point 

 can only be satisfactorilv settled by a systematic study of the lifediistory of the 

 zoochlorellas in living corals. 



Generative Organs. 



Comparatively few specimens in the collection contain generative organs, and in such 

 instances they are entirely of a female character. Hickson (1895, pp. 376-380) has 

 already described the oogenesis, and to a certain extent the spermatogenesis, of 

 Alcyonium, and Ashworth (1898, pp. 207-209) has described the spermatogenesis of 

 Xenia. In the British Alcyonium digitatum the development of the sexual cells 

 occupies a very considerable period, as the ova begin to appear about April or May, 

 and are usually not sexually ripe until December or even later. I have observed the 

 discharge of ripe ova at Port Erin during the early part of April. In the British 

 species the period of the sexually mature condition is extremely short. Apparently 

 the colonies inhabiting a particular district discharge all their ripe ova and spermatozoa 

 almost simultaneously. In tropical members of the family, however, such is not 

 the case. 



In his description of the spermatogenesis of Xenia, Ashworth (1898, p. 207) 

 states that the spermatozoa closely resemble those of Alcyonium in development and 

 structure, but spermatozoa occur in a single specimen in all stages of development, so 

 that the discharge of ripe spermatozoa extends over a considerable period ; this he 

 attributes to the fact that as the coral lives on reefs in the shallow waters of tropical 

 seas, it is not subjected to any great variation in temperature and food supply. 



The development of the ovum in Sarcophytum, Lobophytum, and Sclerophytum 

 closely resembles that of Alcyonium, and is typically Alcyonarian in character 

 (tigs. 17 and 18). Frecpiently (in specimens of Sarcophytum latum, Sclerophytum 

 durum, and Sclerophytum gardineri), however, ova in several stages of development 

 may be observed on a single mesentery, the development proceeding from above 



