Sponges from South Australia. 43 



gathered by human hands, for the dredge cannot reach tliem, 

 on the undersides of rocks. At the same time dried speci- 

 mens also may differ considerably, inasmuch as the washed- 

 out beach-worn specimen differs greatly from those which 

 have been carefully soaked in fresh water and dried for preser- 

 vation immediately after they have been taken, alive and 

 growing, from their place of attachment, as Mr. Wilson's 

 dried specimens also evidence ; next to the wet state, this 

 is the best and most convenient condition for museum-collec- 

 tions. But still, for perfect description the wet state is also 

 necessary. 



Thus alterations in diagnosis and classification may have to 

 be continued so long as knowledge of the subject increases. 



Having, again, as far as my material would allow, studied 

 the groups 10, 11, and 12, viz. the Cavernosa, Compacta, and 

 Laxa, after the manner detailed in the ' Annals of 1882 

 (vol. ix. pp. 347-356) chiefly froni dried specimens (as in 

 my original classification), it seemed to me desirable to add 

 another group to these sponges under the name of " Subcom- 

 pacta " {ib. p. 358), and I must refer the reader to this paper 

 for what I then stated on the subject ; but since I have had 

 the advantage of examining wet or undried specimens, the 

 principle of this classification, viz, the degree of consistence 

 on which I then based my divisions, as may be inferred from 

 the names of the groups, has undergone much modification. 

 But, as I am not prepared to alter the arrangement in extenso 

 now, I would merely observe that it also seems to me desir- 

 able that these four divisions should be considered parts of one 

 group only, for which I would propose the name of " JSuberi- 

 tina ;" in support of which it miglit be stated that, altliough 

 the pin-like skeletal (with or without a flesh-spicule) for the 

 most part prevails in these sponges, and there are some to 

 which 1 have alluded at the conclusion of the paper last men- 

 tioned, in which the skeletal .spicule is not pin-like, that still 

 appear to me to find their most appropriate place in this 

 group, yet, however this may be, the pm-like skeletal with or 

 without a fiesh-spicule of a spinispirular form appears to be the 

 prevailing character of the spiculation in the " Suberitina," 

 and as the latter gradually diminishes in size and number 

 from the genus Spirastrellaj Sdt., to those Suberitina in which 

 there is no longer any trace of it, I would place the whole in 

 this group. 



In the genus LalruncuUa of Bocage, to which I have also 

 alluded in the paper last mentioned (p. 354), the spinispirular 



