Ivi THE VOYAGE OF H.M.S. CHALLENGER. 



pores we have shown to be almost utterly untrustworthy except for specific distinctions, 

 and we are inclined to set very little value, except for purposes of a most general nature 

 upon the fundamental arrangement of the canal system, that is to say, whether it belongs 

 to Vosmaer's third or fourth type, for the simple reason that in nearly every Monaxonid 

 sponge as yet worked out (and we believe also in the Keratosa) it belongs to the former. 

 Still, although the same fundamental type of arrangement will probably be found to 

 prevail throughout the group, yet it is also probable that many minor differences in 

 the canal system of difi"erent genera will be discovered, and will prove to be of great 

 systematic value. As examples of such minor differences we may jwiut to the character- 

 istic arrangement of the larger inhalent and exhalent canals in the genus PhakcUia, and 

 the arrangement of the inhalent canal system in the genus Esperella, as already described 

 by us, in both of which genera the canal system differs in the respects mentioned from 

 the more common Halichondrine type found in Halichondria, &c. As another example 

 we may note the presence of distinct sjihincters or diaphragms in the larger exhalent 

 canals of the genus Spirastrella, which seems to be a fairly constant generic character. 



We come now to the consideration of the spicules, and of the two categories of these we 

 wiU take first the megasclera, which we shall find to be of very great service in classifica- 

 tion. The same form of megasclera is found to run through large divisions ; thus, in the 

 family Homorrhaphida3, comprising the Eenierines and Chalininge, we find only smooth 

 oxeote megasclera and no microsclera. In the Clavulina, again, the megasclera are almost 

 all tylostylote or stylote. But here we occasionally meet with startling exceptions, one of 

 the most striking being the genus Stylocordyla, a corticate sponge, agreeing in skeleton 

 arrangement, &c., with other Clavulina. but always with oxeote sjiicules. In theHeteror- 

 rhaphidse and Desmacidonidse the form of the megasclera is very variable, but we can 

 always use it for purposes of generic distinction. The size of the megasclera is of very 

 slight use save for distinguishing species, but here it is of the greatest value, especially 

 in groups like the Homorrhaphidse, where even the different genera have all the same form 

 of spicule. 



The microsclera, when present, are of still greater systematic value than the 

 megasclera ; indeed they form the best guides to the classification of the Monaxonida. 

 The reason of this is probably that they ai-e not subject to modification to suit the 

 external conditions of the sponge ; and, further, they are usually more complex in form, 

 and thus present more points of possible diS"erence than do the megasclera. Their use 

 in the economy of the sponge is probably very slight ; in many cases we can hardly 

 believe, from their minute size and irregular arrangement, that they have any, and hence 

 there appears to be no reason why they should undergo much modification with changing 

 conditions.^ The structures which we have hitherto considered are, on the other hand, 



^ " It is a strange result wliich we thus arrive at, namely, that characters of slight vital importance to the species, 

 are the most important to the systematist ; " Darwin, Origin of Species, ed. 6, p. 176. 



