THE VOYAGE OF H.H.S. CHALLENGER. 
IU4 
sided affinities that the above mentioned subdivision of them could not be expressed 
graphically in a satisfactory manner. 
“ The second chapter of the Report is devoted to the characters of the Keratosa gene- 
rally regarded as of generic importance, and it renders obvious that, firstly, these charac- 
ters are really of more subordinate consequence than those discussed before, and that, 
secondly, most of the genera hitherto distinguished are created on the basis of thoroughly 
relative and often very ambiguous characters. 
“ The third chapter is that dealing with the description of the Challenger forms, 
and though not interrupted by any considerations of a more general character, it 
neglects no opportunity to point out that the classifier, following the principles used 
at the present time, is on the whole ‘ to be compared with a man wandering in the 
dark.’ 
“ The ‘ concluding remarks ’ summarize the general results of the Memoir, and after 
having discussed the systematic position of the Keratosa with respect to other groups 
of Porifera, Dr. Polejaeff asserts that the Keratosa form probably in this type or subtype 
nothing more than a single family, and that accordingly they are to be subdivided into 
genera directly. In order to have no heterogeneous genera among them, he proposes to 
enlarge the definition of the genus, and for the instances when some undoubtedly good 
species are constant to a relative character to unite them into a subgenus. According to 
this the family in question would consist of the genera Darivinella, Simplicetta (including 
the subgenera Aplysilla and Dendrillci), Ianthella, Spongia (including the subgenera 
Coscinoderma, Euspongia, Hippospongia (?), Cacospongia, and Stelospongos ) , Phyllo- 
spongos (including the subgenera Phyllospongia and Carteriospongia), and Velaria 
(including the subgenera Aplysina, Verongia, and Luff aria). He believes, however, that 
it would have been premature to follow up in his Report the arrangement just mentioned. 
Sure as he is that this arrangement is in the main natural in every direction, he confesses 
that by its adoption only the simplest part of the problem would have been cleared up, 
its most important part consisting in the task of proving actually which of the subgenera 
just mentioned are really to be regarded as subgenera (i.e., groups which, although 
connected by numerous intermediate stages with their systematic neighbours, still present 
in their organisation a new principle fit for a further development) and not as species 
and even varieties. This latter question is, according to him, to be decided (perhaps 
exclusively) by the methods of comparative physiology. Many passages are devoted to 
the definition of the idea of the genus. In his report on the Challenger Calearea Dr. 
Polejaeff regarded the generic character to be a 1 character of sufficient constancy, and 
together with this, allowing numerous modifications either in the direction of a further 
development, or in the direction of different variations.’ His experience whilst examining 
the Challenger Keratosa leads him to the suggestion that at least in the instances where 
but one character — and not a series of them as is the case with the Calearea — decides 
