172 Prof. A. Dendy on 
referable to Tragosia infundibuliformis. Two of these I 
examined microscopically. ‘The first was labelled, in Bower- 
bank’s writing, “ Halichondria infundibuliformis Johnst.’’, to 
which had been added, apparently in Carter’s writing, “'T. S.” 
and “* Jsodictya inf. Bk.,”’ suggesting that Mr. Carter re- 
garded this as the type-specimen of the species as accepted 
by Johnston and Bowerbank. The second was labelled 
“* Tsodictya infundibuliformis Bk.’ and “ Halichondria in- 
Jfundibuliformis Sowerby.” Both of these specimens had 
the typical external form, and both contained trichodrag- 
mata. There can be no doubt that they are specifically 
identical with the specimens obtained by Professor Herd- 
man from the Minch. 
I think we may now state confidently that the common 
British species upon which Gray’s genus 7ragosia was founded 
contains trichodragmata as constant and characteristic con- 
stituents of its spiculation, and may thereby be distinguished 
much more sharply from Phakellia than was formerly 
possible. The fact that these spicules have been com- 
pletely overlooked by previous observers may probably be 
accounted for by their having examined only dry specimens, 
in which the shrinkage of the soft tissues makes it much 
more difficult to recognize them. 
I have already, on more than one occasion [1916, 1921 4], 
called attention to the curiously sporadic distribution of this 
type of microsclere amongst the Tetraxonid Sponges, and 
suggested that it has probably arisen again and again in the 
course of evolution through parallel mutation. I have also 
doubted its value for purposes of generic distinction, and 
been, I fear, somewhat inconsistent in this respect. Thus 
I have refrained [19218] from separating Cinachyra eury- 
stoma, which possesses trichodragmata, from the remaining 
nineteen species of the genus, which possesses none, while, 
in the same memoir, I refused to admit into the genus Awi- 
nella a species (“ Thinacophora”’ durissima) which possesses 
trichodragmata, and also suggested that the presence of these 
spicules may be used as a means of distinguishing the genus 
Mycale from Aigagropila and Esperella. I have suggested, 
in short, that each case should be treated on its merits, 
without attempting to lay down a general rule. Where a 
generic diagnosis is quite inadequate, as in the case of 
Tragosia, the presence or absence of trichodragmata may 
well be taken into consideration. 
Hallmann [1916-17], a few years ago, published a 
revision of the genera of so-called Axinellide containing 
microscleres, in which he mentions a considerable number 
