DENDY NON-CALCAKEOUS SPONGES 133 



In 1887 the same species was recorded by Mr. Carter from the Mergui Archipelago 

 and some interesting particulars added as to its mode of growth, as follows : " Its 

 growth is more remarkable than in that [the Honduras] example, for it is laminar, 

 and extends in a horizontal direction for several square inches ; the superficial stratum, 

 which is comparatively thin and buff-yellow in colour, changes to black or dark brown 

 in the cancellated cavities to be presently mentioned for half an inch downwards, where 

 it rests on granite. The explanation of this abrupt termination is that the lower 

 portion is mingled with a layer of coral which has been cancellated by the excavating 

 habit of these sponges, which exhibit an apparent fondness for calcareous material, 

 whether in a mineral or organic form." 



Mr. Carter thus clearly recognises that his Suberites coronarius may, at any rate 

 under some circumstances, be an excavating sponge. A specimen in Mr. Hornell's 

 collection (R.N. XXVI. 10), also excavating and encrusting a piece of coral, agrees 

 very closely with the Mergui specimen, and its examination, I think, fully justifies the 

 transference of the species to the genus Cliona. In all probability the massive West 

 Indian specimens stand in exactly the same relation to the excavating Indian Ocean 

 specimens as does the massive " Raphyrus griffithsii " to the excavating Cliona celata 

 of European seas, or to the encrusting form of the latter described by Topsent [1900]. 



Mr. Hornell's specimen consists, in the first place, of a thin crust (about 1 mm. 

 thick), with smooth outer surface. The underlying coral has been eroded and largely 

 disintegrated by the sponge, giving rise to the " cancellated " structure described by 

 Mr. Carter, which, when teased up and examined microscopically, is seen to consist 

 of a mixture of coral fragments and sponge. 



In the outer portion of the sponge the skeleton consists of dense, irregular wisps 

 of tylostyles running towards the surface, where they form a thick dermal pile with 

 outwardly directed apices. In the cancellated portion the tylostyles seem to be quite 

 irregularly scattered. The spiculation consists of tylostyles and spirasters only ; I 

 have seen no oxea such as sometimes occur in Cliona (fide Topsent). The tylostyles 

 are straight or very slightly curved, usually very sharply and gradually pointed. They 

 have well developed, almost spherical heads. When fully grown they measure about 

 0-35 by 0-012 mm., with head 0-014 mm. in diameter. The greatest diameter of the 

 shaft is at about one-third of the distance from head to apex. The spirasters are very 

 slender and measure about 0-02 mm. in a straight line from end to end. Many of them 

 have the typical " semicircular " form described by Carter, but many show a very 

 obvious spiral twist. They resemble bent fragments of broken fretsaws. It is not 

 easy to assure oneself that the short spines are really capitate, though they may have 

 the appearance of being so. 



As might be expected, the spiculation, as well as the mode of growth, of the Okha- 

 mandal sponge seems to agree more closely with that of the Mergui than with that of 

 the West Indian specimens. Thus the heads of the tylostyles are more nearly spherical 



