XCviii THE YOYAGE OF H.M.S. CHALLENGEE. 
distinctive ? To this I answer frankly, that as a distinctive character it has absolutely no 
existence ; the mesogloea (which, as a term, is preferable to the certainly objectionable 
mesoderm) is in some very few of the Chondrospongise a chondrenchyme, but in the vast 
majority it is a collenchyma (soft mesogloea) or sarcenchyme (also a soft mesogloea), the 
former precisely similar to the collenchyma, and the latter probably to the granular 
collenchyma of the so-called Cornucospongise ; this presumed distinction is thus of far 
less value than that supposed by Vosmaer, for it is not even fairly general, while the 
absence of spongin in the Spiculispongise is. Nor, I regret to say, can I agree with 
Lendenfeld in regarding the so-called Chondrospongise as generally corticate ; a large 
number certainly are, but a goodly number as certainly are not. 
The Spiculispongise and the Cornucospongise cannot be distinguished, either by the 
characters of the skeleton or of the soft parts ; they pass insensibly into each other, as 
Carter alleged long ago.^ 
It would therefore appear that a single order equivalent to the Hexactiuellida should 
be defined to receive the other members of the class, and there would certainly be great 
convenience in this proceeding, were it not for the existence of the Myxospongiae, 
Sponges without any skeleton at all. If it were possible to accept Vosmaer’s view 
that these are simply degraded forms which have lost a skeleton they at one time 
possessed, one might readily include them in a single group with the rest of the 
Micromastictora that are not Hexactinellids, but for this presumed degradation there 
appears to me to be no shred of evidence ; the Halisarcidse are characterised by great 
simplicity, both in the chamber-system and in the canal-system ; and in the course of 
their embryological development they give no signs of a retrogressive metamorphosis ; 
they may therefore with much greater probability be regarded as persistent simple forms, 
descended from askeletal ancestors which were the common parents of them and the 
spicular Sponges. If this view be taken of tbe Halisarcidse and their associates, it 
appears to follow that their position amongst the Sponges is so unique that they should 
be separated from the rest of the Micromastictora as a distinct order, and we then arrive 
at the following classification : — 
Class I. Megamastictora (with the single subclass Calcarea). 
Class II. Micromastictora. 
Subclass I. Myxospongise. Micromastictora which are askeletose. 
Subclass II. Hexactinellida. Micromastictora in which triaxon spicules con- 
tribute to the formation of the skeleton. 
Subclass III. Demospongiae.^ Micromastictora which possess a skeleton either 
of siliceous spicules or spongin, or of both combined, but the megascleres 
are never triaxons. 
^ Carter, Ann. and Mag. Nat. Hist., ser. 14, vol. xvi. p. 58 (sep. copy), 1875. 
2 Sollas, Sci. Proc. Boy. Duhl. Soc., N.S., vol. v. p. 112, 1886. 
