120 CEYLON PEAUL OYSTER EEPOET. 



Coppatias, as we have already seen, is simply a reduced Stellettid in which the cladi 

 of the triaenes have become completely aborted, and the megascleres now consist 

 solely of oxea. From the oxeote to the tylostylote form appears at first sight a big 

 jump, but I think it is evident that the pin-headed type in Hymedesmia is simply the 

 result of the thinly encrusting habit. The megascleres in this genus are typically 

 arranged at right angles to the base of support, with their apices projecting oiitwards, 

 often beyond the surface of the sponge, an arrangement which is admirably adapted 

 Ixith to support the soft tissues and to protect the sponge from the attacks of parasitic 

 Crustaceans or other enemies. In this position the proximal end of the megasclere, 

 pressed against the hard base of attachment, is unable to elongate in the normal 

 manner, and siliceous material, which would otherwise have been devoted to its 

 elongation, is accumulated in a swelling or knob the " pin-head." 



Such is, I believe, the origin of the spirastrellid and suberitid tylostyle, and an 

 exactly analogous process appears to have taken place in the echinating spictiles of 

 the Ectyoninse, which are commonly swollen into a "head" at the end which abuts 

 against the skeleton fibre. The at first sight tempting idea that the head of the 

 tylostyle is the vestige of the cladome of a trisene is put completely out of court by 

 the fact that it lies at the wrong end of th'e spicule, viz., the proximal instead of the 

 distal end. 



The evolution of the typical spirastrellid spiraster from the euaster has evidently 

 taken place by elongation and bending of the centrum of the latter, and various 

 stages of the process may be seen in different species of Hymedesmia ; indeed, the two 

 forms of aster pass so insensibly into one another that it is impossible to base a 

 generic distinction upon their character alone. In the spirastrellid spiraster the rays 

 (actines) are commonly reduced to mere spines or minute blunt projections. 



Hymedesmia stellivarians, CARTER. 



1880, Hymedesmia stellivarians, CARTER (4). 



The single specimen forms a thin yellowish crust, attached to a small melobesian 

 nodule, in company with Paresperdla serratohamata, &c. In spite of the somewhat 

 meagre character of Mr. CARTER'S description and figures, I think there can be little 

 doubt of the correctness of the identification, especially as Mr. CARTER'S specimen 

 also came from Ceylon waters. 



All three forms of aster mentioned by Mr. CARTER are present, but they are not all 

 of the same size, as might, perhaps, be inferred from his description and figures. 

 Moreover, in our specimen the megascleres are usually a good deal longer in proportion 

 to their thickness than Mr. CARTER'S figure indicates, but they are also very variable 

 in dimensions. 



Under the circumstances it seems desirable to give the following particulars as to 

 the spiculation : 



(1.) Tylostyli ; only slightly curved, if at all; with oval head at one end and 



