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



the matter is quite different, and it depends very often exactly on our conceptions as to 

 whether the given group forms a class or a family whether we subdivide it directly into 

 genera or preliminarily into orders and suborders. Now we have seen that there exist 

 in science quite opposite opinions as to what the genus ought to represent, and that 

 many naturalists find no absolute distinction between genus and species. We have also 

 seen that in the Keratosa, to use the words of Prof. F. E. Schulze,^ " die Entscheidung 

 der Frage, ob eine Anzahl verwandter Formen als Arten einer Gattung oder als Varietaten 

 einer Art hinzustellen sind, oft besonders schwderig erscheint." Be that as it may, the 

 given forms must be classified and introduced in the system, and the hesitation of the 

 classifier must come to an end. On what now may his final decision depend ? That it 

 may dejjend on the inspiration of the moment is undeniable, but I think the importance 

 of this latter factor must not be exaggerated. Every conscientious investigator will 

 always search for more positive arguments, and it is plain that if he regard the corre- 

 sponding group as an order subdivisible into families, he will bring his hesitation to 

 an end by creating a new genus ; and, on the contrary, if he believe the group to be only 

 a family, he will describe the series of forms in question as a species with varieties. 

 Accordingly, and as I remarked before, from this point of view a reliable answer to the 

 question put on the preceding page is the most desirable. 



Numerous and very conflicting opinions have been expressed on the problem of the 



affinities of the horny sponges. I begin with that of Oscar Schmidt. This naturalist does 



not deny the close relationship of his Ceraospongise with the Monactinellida, and namely 



with Chalinidae, moreover, he regards them as forming no larger systematic unity than a 



family; but he considers ^ them to present an independent natural group, and is decidedly 



against any introduction of true Chalinidas into it. Hyatt goes still further. He regards 



the Keratosa as forming an indej^endent order, a very significant statement, since the 



naturalist just named, when writing that " the characteristics of the order Keratosa are 



more clearly defined than those of any other among the class Porifera," &c. (Revision, 



&c., vol. i. p. 399), was of the opinion that the whole group of Porifera form nothing 



more than a class of Infusoria.' There are in his valuable memoir on the North American 



Porifera no further explanatory observations in this direction, but it is plain that the 



words above quoted admit of but one explanation, namely, that the Keratosa are to be 



regarded as a group systematically equivalent to the groups Calcarea and Silicea. Gray * 



and Bowerbank,^ in harmony with Grant, call the Keratosa also an order, but they 



class within it the true horny sponges with sponges producing "proper spicules." 



Finally, Carter," agreeing on the whole with Gray and Bowerbank, differs from them in 



this point, that he considers the Keratosa to represent two orders, without foiming, 



' Zeitschr.f. wiss. Zool., Bd. xxxii. p. 612. 



2 Spong. d. adriat. Meeres, II. Suppl., p. 9 ; Spong. d. Kuste v. Algier, p. 36. 



' Revision, &c., vol. ii. p. 481. * Proc. Zool. Soc. Land., 1867, p. 503. 



^ Monograph Brit. Spong., vol. i. p. 205. " Ann. and May. Nat. Hist, ser. 4, vol. xvi., 1875, p. 13?. 



