138 "ENDEAVOUE" SCIENTIFIC KESULTS. 



assumption of an erect habit of growth, or that a return to a 

 primitive condition on the part of more or less highly evohed 

 species has occurred again and again within the group ; other- 

 wise it is impossible to account for the existence of groups of 

 genera — e.g. Hymeraphia, Microciona, Clathria and Ophlita- 

 spongia; Leptosia, StyJostichon, Ectyodoryx and Lissoden- 

 doryx ; Hymenancora, Plumohalichondria, Ectyomyxilla and 

 Myxilla; Dragniatyle, Tedauia and Acheliderma ; and others — 

 each of which comprises a natural series connecting "promyx- 

 illine" with Ectyonine and Dendoricine forms. Whilst it is 

 probable that each of the tw^o possible explanations contains 

 some portion of the truth, it would seem that the former is 

 more satisfactorily in accord with the bulk of the evidence, 

 although it involves the assumption that, amongst the Myxil- 

 lina?, sponginous fibres have originated independently many 

 times over. If, howe\er, as is scarcely to be doubted, such 

 fibres have arisen as linear upgrowths of the basal spongin- 

 lamina, there is no great improbability in such an assumption 

 pro\ ided that there already existed in the common ancestor an 

 incipient tendency, such as has been postulated, towards fibre- 

 formation. 



Taking into account onl\ those features which belong to the 

 skeleton, but neglecting, for the time being, any considerations 

 in reference to the microscleres, it may be said that nearly all 

 of the diversities of spiculation and structure which occur in 

 the Myxillina^ are ascribable to (a) modifications in the form of 

 the megascleres and the division of either or both of the origi- 

 nal groups of megascleres into two or more kinds; {h) 

 differences in the arrangement and constitution of the fibres, 

 and in connection therewith different combinations of the 

 megascleres in respect to their particular location in the skele- 

 ton ; or (c) the loss by atrophy of one or more groups of megas- 

 cleres formerly present. 



Of verv common occurrence has been the differentiation of 

 the basical megascleres into two kinds, a larger, in the case of 

 which the primitive spination has usually undergone partial 

 reduction or become entirely lost, and a smaller, in which it has 

 almost invariably been retained ; these may be distinguished 

 as principal and accessory basical megascleres respecti\ely, or 

 simply as "principals" and "accessories." The resultant tri- 

 megascleric condition is that which is characteristic of normal 

 "Ectyoninse." In many Myxillina', however, the "basicals" 

 are onlv incompletely differentiated and in such cases, unless 

 (as in Crella) there be a determinable difference of function 

 between the spicules which lie at one end of the series and those 

 which lie at the other, it is diilicult to decide whether they 

 should be regarded as belonging to one or to two groups {e.g. 

 species of Stylostichon, Clathrissa, etc.). The maximum 



