800 PROF. A. DENDY AND MR. R. W. H. ROW ON 



one author, appears rarely if ever to be the case in true Sycettidse: 

 or their descendants. To this must be added the primitive type 

 of skeleton, composed exclusively of equiangular radiates, which 

 do not exhibit the characteristic arrangement met with in the 

 syconoid sponges, with their differentiated gastral cortex and 

 articulate tubar skeleton. In this connection we may especially 

 note the absence of subgastral sagittal triradiates (or quadri- 

 radiates), which form such a constant feature of the Sycettidse 

 and their derivatives. 



The Denclya line seems to have given off two branches, repre- 

 sented by the Leucascid* and Leucaltidag respectively. The Leucas- 

 cidfe are undoubtedly the more primitive of the two. The genus 

 Leucascus itself, indeed, might very easily be mistaken for a 

 homocoel sponge were it not for the presence of a distinct and 

 independent pore-bearing dermal membrane ; it retains the 

 elongated, branched, and more or less radially arranged flagellate 

 chambers of its Bendya-like ancestors. Within the family evolu- 

 tion has led to the development of a more highly differentiated 

 dermal cortex in Leucetta and Perichm-ax, accompanied by great 

 reduction in the size of the flagellate chambers and complication 

 of the inhalant and exhalant canal systems. In this way 

 has arisen that remarkable convergence between Leucetta and 

 Pericharax on the one hand, and the leuconoid Grantiidse on the 

 other, which has for so long prevented the appreciation of the 

 fundamental distinction which really exists between these forms. 

 The remaining genus in the family, Leucomalthe, is a highly 

 specialised and aberrant type, which is only included here 

 provisionally, vmtil we know more of its minute anatomy and 

 histology. 



In the Leucaltidae the distinctive peculiarity has been the 

 enormous development of the dermal cortex with its special 

 skeleton, and the accompanying reduction of the skeleton of 

 the chamber layer to a more or less vestigial condition, or even 

 its complete disappearance. In this family, again, as regards 

 canal system, we meet with the customary transition from the 

 long chambers and radial arrangement of the more primitive 

 forms (Leucaltis) to the spherical chambers and scattered 

 arrangement of the highest [Leucettusa). 



To this line of descent must also be relegated two out of the 

 three surviving families of " Pharetrones," namely, the Minchi- 

 nellidfe and the Murrayonidte. We found this conclusion upon 

 the basal position of the nucleus in the collared cells in 

 MinchineUa and Murrayona ; but it must be borne in mind that 

 as regards their general organisation also the members of these 

 two families differ very widely from Lelapia and Kebira^ the 

 only representatives of the Lelapiidte, the third surviving family 

 of " Pharetrones." 



We are therefore compelled to regard the so-called family 

 Pharetronidae as of diphyletic origin, and the resemblance, such 



