84 OKHAMANDAL MARINE ZOOLOGY REPORT PART II 



sufficient, and in several respects misleading. Thus he says that " the terminal orifices 

 are rarely ciliated ; but when they are so furnished the ciliary fringe is composed of a 

 prolongation of the layer of large acerate spicula." This really applies only to the 

 openings accidentally produced by breaking across of branches, the true oscula are 

 provided with a special fringe of slender, hair-like oxea. Again, he says that there 

 are no " defensive spicula projected from the oscular surface," by which he evidently 

 means no gastral quadriradiates ; such spicules, however, occur in his specimen. The 

 statement that the apices of the subdermal and subgastral triradiates are cemented 

 together where they meet by " keratode " hardly deserves contradiction, the " kera- 

 tode " being, of course, simply the dried remains of the soft tissues. 



Under the circumstances it seems desirable to give a completely new description, 

 based upon Mr. Hornell's spirit-preserved material. I may say, however, that I cannot 

 find any character in which his specimens differ from the type. 



The sponge colony (Figs. 3, 3a, 36) consists of very numerous, rather slender, 

 cylindrical branches, for the most part ascending vertically and lying close together. 

 The branching is very irregular and takes place by the formation of lateral buds 

 at varying levels. Each branch terminates, when fully developed, in a distinct 

 osculum, but the younger buds are blind. The oscula appear to be naked, but 

 are in reality provided with an inconspicuous fringe of slender, hair-like oxea. 

 The projecting portions of these spicules appear always to be broken off. Just 

 within the osculum is a transverse membranous sphincter. All the colonies have 

 evidently been attached below in life, and in one case a portion of the substratum 

 is still present in the form of a barnacle shell. 



The largest colony, or piece of a colony, in the collection measures about 

 32 mm. in height by 38 mm. in greatest breadth, and is composed of about 50 

 branches measuring up to about 19 mm. in length and 2-5 mm. in diameter. 

 There is considerable variation, both as regards the length of the branches and 

 the compactness of the colony, in different specimens. The surface of the branches 

 is longitudinally striated (Ute-like) owing to the presence of the large dermal 

 oxea, but otherwise smooth. The colour in life was white, in spirit it is dirty 

 white. 



The canal system is typically syconoid, the thimble-shaped radial chambers 

 extending at right angles through the wall of the sponge from gastral to dermal 

 cortex. 



The preservation of the material is not good enough to enable me to make 

 any detailed histological observations, but the nuclei of the collared cells are apical. 



The gastral cortex is fairly thick, and its skeleton is made up of the facial 

 rays of gastral sagittal triradiates and quadriradiates, and the oral rays of sub- 

 gastral sagittal triradiates. The dermal cortical skeleton is very strongly developed 

 and made up of dermal triradiates, the outer rays of subdermal pseudosagittal 



