the Cape Ceplialodiscus. 235 



area and tlie sliore, but no iiulications of these have as yet 

 been found in this locality, nor have any fragments of the 

 ccenceciuni been observed cast up on the shore. The oyster- 

 gatherers of the locality are familiar with the general 

 appearance of most fairly large objects at extreme low tide, 

 but none of them, with one doubtful exception, recognized a 

 specimen shown them as having been met with in the course 

 of their work. It may therefore be concluded that the CJape 

 Cephalodiscus grows in fairly deep water on rocky ground. 



Mode of Occurrence. 



The mode of occurrence of the animals on the rocky areas, 

 which they seem to frequent, is indicated by some interesting 

 specimens procured. Those which showed a region of 

 attachment had a basis of irregular shape, about 2 inches 

 or so in diameter, on which tiie bush-like structure of the 

 coencecium was supported. Tliis app^'arance is so charac- 

 teristic of the Cape species that the tialiermen call them " sea 

 busiies." Other specimens, however, showed an expanded 

 sheet of ccenoecial substance devoid of animals, and on this 

 several such "bushes" occurred. It would appear, there- 

 fore, that the rock or substratum may be covered by more 

 or less extensive sheets of coencecial substance from which 

 the branching stems may arise. 



It may be noted that the condition of the basal part of 

 these specimens with a single point of attachment is not 

 incompatible with the supposition that they may have come 

 from a more extended base. 



One specimen * was of particular interest, as it retained the 

 object on which it had grown and is the only case, so far as 

 I know, in which the basal object has been found, except 

 those described by Harmer (2) in the 'Siboga' Expedition 

 collection. This object was a compaiatively small piece of 

 sponge about 4 inches in length and of a compact hard 

 skeletal structure. The history of this has been probably 

 that a larva or larvse which, as I have noticed, have a tendency 

 to crawl upwards, had begun their coencecium-building on 

 the sponge or its dead skeleton, and this was gradually built 

 up to a comparatively large superstructure. 



That the animals occur in fair abundance in their natural 

 habitat is indicated by the frequency with which detached 

 specimens are found at a distance from rocky ground. 



* [This specimen has been presented by Dr. Gilchrist to the British 

 Museum. 5lr. 11. Kirkpatiick informs me that the Sponge referred to 

 belongs to the genus Chondrocladia, A^'yv. Thorn?., and the family Desma- 

 cidouidae. — S. F. H.] 



17* 



