Prof. W. Thomson on the “ Vitreous” Sponges. 117 
spicules, where these occur. I am inclined to regard the cor- 
ticate group, as limited by Prof. Schmidt, as of ordinal value. 
Prof. Schmidt’s last family are the Halisarcine :—“ Spongie 
molles, non fibrose, corpuscula calcarea vel silicea non conti- 
nentes ’—equivalent to the Halisarcina of Lieberktihn, an 
obscure group with neither horny fibres nor siliceous spicules, 
and consisting of little more than an extended sheet of sarcode. 
The definitions of the three orders in Dr. Bowerbank’s 
‘British Spongiade’ are sufficiently simple; but I cannot 
regard the ‘‘ KERATOSA”’ as a group equivalent to the “ CAL- 
CAREA ” and the “ SiLicEA.” 
The diagnoses of the suborders are based upon some impor- 
tant modifications in the arrangement of the spicules and horny 
matter, which do not, however, seem to be sufficiently definite 
for the purposes of classification. The groups are, on the 
whole, natural. 
Admitting the value of Dr. Gray’s two primary subclasses, 
the details of his classification seem to me unsatisfactory. The 
author divides the Siliceous Sponges into two primary groups 
—(1) those with membranous or unarmed ovisacs, and (2) those 
whose ovisacs are strengthened with siliceous spicules. [ 
doubt if we know enough of the nature of these peculiar 
bodies which we for the present call ovisacs, to found upon 
them broad distinctive characters. The present attempt to do 
so separates to the utmost the nearly allied corticate genera 
Tethya and Geodia, and eS Spongilla and Halichondria 
(Isodictyon) (between which, except m the one point of the 
structure of the ‘“ ovisacs,” it is difficult to define generic dis- 
tinctions) in different principal sections. Under the first two 
subsections of the MALAcOspoR&, Dactylocalyx and Spongia 
are associated, on account of the common character of possess- 
ing a network, while Aphrocallistes is divorced from its beau- 
tiful partner Huplectella. 'The third subsection, the ARENO- 
SPONGI#, is an excellent group, apparently of ordinal value. 
Subordinate to the subsections, Dr. Gray proposes seven 
orders and a host of genera, and very naturally anticipates the 
general denunciation of a system which complicates the no- 
menclature to bewilderment, and founds generic groups upon 
such “ imperfect materials” as a ‘“ bihamate spicule figured in 
Bowerbank’s ‘ British Sponges.’ ”’ 
The only classification which has any material advantage 
over the older classifications of Nardo, De Blainville, Johnston, 
and Lieberkiihn, seems to be that of Dr. Oscar Schmidt. Du- 
_ chassaing and Michelotti, Bowerbank and Gray, have each made 
valuable individual suggestions; but Dr. Schmidt’s grouping, 
taken as a whole, appears to be the most in accordance with 
our knowledge of the anatomy and physiology of the class. 
