46 DAWSON AND HINDE 



horizontal spicules seem to be triacerate in form, and much shorter than those of the 

 vertical system, though of very different lengths. They are sometimes in bundles and 

 sometimes solitary. 



In parts of the substance, apparently within the reticulate wall, may be seen a few 

 cruciform spicules, and flocculent patches apparently of very small spicules, which seem 

 to have been mostly internal and most abundant toward the base, but cannot be 

 distinctly ma.de out. 



The wall is very delicate, and consists of quadrate or oblong areas formed by slender 

 longitudinal and transverse strands or fibres, of which the former are the more prom- 

 inent. As in Protospongia, the quadrate areas are formed by the four transverse rays of 

 cruciform, or five-rayed spicules, but these are disposed so that their rays overlap each 

 other, and thus form fascicles of closely opposed parallel rays. The spicules in the trans- 

 verse strands of the wall are less thickly grouped together, and even in some of the larger 

 squares they may be arranged singly, whilst the smaller squares are generally bounded 

 by single spicules only. The longitudinal strands principally consist of cruciform (?) 

 spicules, but it is possible that elongated filiform spicules may likewise be present. There 

 are plain indications of a fifth or distal ray in many of the principal spicules of the wall, 

 shown by a very minute knob or blunted process projecting from the central node of the 

 transverse rays, which may represent a partially developed ray, or the broken stump of a 

 complete one. In some places, also, there is a continuous film of pyrites, probably 

 indicating a membrane of very minute spicules or an agglomeration of flesh-spicules, 

 now replaced by this mineral. 



The basal portion of these specimens is incomplete, but there are indications of an 

 extension of the longitudinal strands of the wall downward into a spreading tuft of 

 short anchoring spicules widening at their distal ends. 



This genus is mainly distinguished from Protospongia by the fascicular arrangement 

 of the spicular rays in the principal longitudinal and transverse fibres. The regular 

 quadrate areas of the body- wall also mark it off from Pledoderma and Phormosella, Hinde. 

 (See Brit. Foss. Sponges, Pait. I, PI. Ill, figs. 1, 2 and Part. II, p. 124-5, Pal. Soc, ISSe-SY.) 

 How far it may resemble Diclyophylon,^ Hall, and the other genera associated therewith 

 by Prof. Hall (35th Report of the State Museum, 1884, p. 165, pis. 18-21), it is impossible 

 to state, for the stnrctural features of this genus haA'^e not been sufficiently described, 

 and the characters assigned to the other genera are mainly those of external form, which, 

 as regards this group of sponges, are hardly of generic importance. 



The structures of Cyathophycus, as shown in these specimens, bear a great resem- 

 blance to those of the recent genus Holascvs, Schulze (Challenger Reports, Vol. XXI, p. 85), 

 based on sponges dredged from depths varying between 1,3*75 and 2,650 fathoms in the 

 South Atlantic and in the Southern Ocean." There is a striking similarity in the struc- 



' In the only species of the DictyospongidBe in which I have seen structure, that named by Whitfield Uphan- 

 lenkt Dawsoni, Am. Journ. of Science, Aug., 1881, and Bulletin Am. Num. Nat. Hist, Dec, 1881, the spicules are 

 appparently filiform and arranged in broad longitudinal and transverse bundles crossing each other, and with 

 small, loose flesh-spicules in the meshes. The arrangement is therefore différent in details from tliat of Cyatho- 

 phycus, or, as it should now be called, Cyatho^pongia. The name Hydroccras proposed by Conrad, is liable to the 

 objection that it was intended to indicate affinity to cephalopo'l shells. — .1. W. D. 



^ Especially H. fihvlatus, Schulze. Chal. Rep. svj, fig. P. 



