Sponges from South Australia. 207 



the lowest forms and building upwards, instead of following 

 the arrangement given in my classification, which is the re- 

 verse, I shall pursue this course after having premised the 

 following introductory remarks : — 



It is desirable, when about to introduce anything new, to 

 state such generalities as may not only facilitate its under- 

 standing, but, if possible, curtail length of description, in 

 order that the student may of himself be able to supply the 

 former, which otherwise might become tedious repetition in 

 the latter. 



Twenty-six years have passed since my observations on the 

 growth of the young Spongilla from the statoblast were pub- 

 lished, and although much has been done since in structural 

 description, yet a key to a great part of this might be found 

 in my illustration of a minute but entire specimen of Spongilla, 

 whose component parts were then drawn to the same scale, 

 " in order that their relative proportions might be preserved 

 as much as possible" for diagrammatic purposes in teaching; 

 but I question if such use was ever made of it (' Annals,' 

 1857, vol. xx. pi. i. fig. 1). 



In this figure, it may be observed that the parenchyma, of 

 which the body of the sponge is composed, is surrounded by 

 a dermal membrane, which is kept at a certain distance off by 

 the projection of conical bundles of spicules issuing from 

 the parenchyma, thus leaving a hollow between the two, 

 subsequently termed by Dr. Bowerbank the " subdermal 

 cavity." 



This " dermal membrane," or cuticle, I named at the time 

 " investing membrane," and the interval between it and the 

 parenchyma the "cavity" of the "investing membrane." 

 1 also then stated that the " investing membrane " was 

 pierced by "apertures" or pores (ib. p. 22), and that the 

 parenchyma was traversed throughout by the branches of the 

 excretory canal-system, which ended in a single tubular vent 

 that passed directly through the cavity of the investing mem- 

 brane (I. c. pi. i. tig. l,g)j to terminate a little distance beyond 

 the latter, with which it was intimately connected at the point 

 of its passage, so that the contents of the excretory canal- 

 system could not get into the cavity of the investing mem- 

 brane. Thus all the main features of sponge-structure were 

 foreshadowed. 



Now it is evident, as shown in the figure to which I have 

 alluded, that if the " investing membrane " should sink in 

 towards the parenchyma by collapse" or otherwise, the points 

 of the bundles of spicules which support it would become 

 conical, and hence the features thus produced have been termed 



15* 



