202 CEYLON PEARL OYSTER REPORT. 



I hope, appear in the course of the subsequent pages ; in the meantime I may give 

 the following summary of the conclusions at which I have arrived : 



In the first place it is pretty obvious that the Aplysillidse (constituting a large part 

 of the so-called " Hexaceratina ") form the starting point of the evolutionary series 

 within the order. The primitive character of such genera as Aplysilla and 

 Darwinella is clearly indicated by the simple canal-system, the large sac-shaped 

 flagellate chambers, and the very simple skeleton of branched spongin-fibres, supple- 

 mented in Darwinella by detached spicules of spongin. The presence of these 

 so-called spicules at first sight seems to lend colour to LENDENFELD'S views as to the 

 relationship between the Aplysillidse and Hexactinellida. It is very difficult to see, 

 however, how the horny spicules in question can have anything to do with the 

 siliceous spicules of the Hexactinellida ; their shape is extremely variable and they 

 are probably best regarded simply as isolated portions of the general spongin 

 skeleton, secreted by groups of spongoblasts which, for some unknown reason, have 

 become isolated from their fellows. 



Altogether the Aplysillidse agree very closely in structure with the Myxospongida, 

 especially with the genera Halisarca and Hexadclla, and it is not impossible that 

 the curious fibres of Halisarca may represent a rudiment of a horny skeleton. That 

 the spongin skeleton in the Aplysillidae has been developed quite independently of 

 that of the Monaxonellida, and with no relation to a pre-existing siliceous skeleton, 

 admits, I think, of little doubt. The character of the skeleton, consisting in the 

 simplest cases of a thin basal lamina of spongin, from which slightly branched fibres 

 spring vertically upwards and end in surface conuli, without anastomosing with one 

 another to form a network, is quite different from what we find in typical horny 

 Monaxonellida, in which the spongin is originally deposited as a cement which binds 

 together the spicules of a reticulate skeleton, and in which, consequently, when the 

 spicules disappear, the spongin is left in the form of a network of horny fibres. A 

 very similar network of horny fibres appears, however, to have been independently 

 evolved in the higher Euceratosa. 



This difference in the arrangement of the horny skeleton in the one case in the 

 form of a network and in the other case in the form of separate tree-like fibres has 

 given occasion to MINCHIN (12) to divide his " Grade" Keratosa into two orders, viz., 

 " Dictyoceratina (= Monoceratina, LDF.)" and " Dendroceratina (= Hexaceratina, 

 LDF., pars}" the latter group including only the family Aplysillidoe. MINCHIN'S 

 distinction cannot, however, be maintained as a basis of classification, for, as 

 LENDENFELD himself recognised, there are undoubted Aplysillids (e.g., Dendrilla 

 elegans, LENDENFELD) which possess a reticulate skeleton, and in the present report 

 I propose the new genus Megalopastas for such forms, of which two species occur in 

 Ceylon waters. 



The importance of the genus Megalopastas lies in the fact that it forms a 

 connecting link between the Aplysillidee and Spongeliidae, and thus completely breaks 



