252 W. J. BOLLAS ON PHAKETROSPONGIA STRAHANI. 



terized by lattice-like (" gitterformig ") tissue, and which clearly be- 

 longs to the existing Hexactinellidoe. The other great fossil group, 

 distinguished by vermiform ( ; ' wurmformig") fibres, or the Yermi- 

 culata, is sometimes regarded as closely related to the Lithistina. 

 The structure of the genus Pharetrospongia, which is repeated with 

 modifications in various species of sponges belonging to the genera 

 Scyphia and Marion, together with the fact that Siphonia, Polypo- 

 tJiecia, and other fossil genera are true Lithistid sponges, shows, 

 however, that the Vermiculata are not at all a natural group, but 

 simply a mixed assemblage of such sponges as are not Hexacti- 

 nellids, and which belong, some to the Eenierida, others to the 

 Lithistina, and the rest probably to various other orders of recent 

 sponges. 



The separation of the fossil from the recent sponges, which has 

 been made since D'Orbigny by almost every palaeontologist who has 

 written on the subj ect, now that we know more about the structure 

 of both of them, appears to have been a great mistake in classifi- 

 cation. 



The simple fact is that we do not want a separate classification 

 for the fossil sponges, any more than we require a separate classifi- 

 cation for the fossil Foraminifera ; and hence all such elaborate 

 attempts as that, for instance, made recently by Pomel, are works 

 of supererogation, which the modern spongiologist is obliged to ig- 

 nore if he sets any value on economy in nomenclature and time. 

 What we do now greatly need is knowledge of the exact structure 

 of the fossil forms ; that once definitely determined, but little diffi- 

 culty need be expected in the work of correlation ; the grouping 

 of the petrified sponges will then follow naturally upon the lines 

 already laid down for existing species. 



Supplementary Note, April 2Q, 1877. 



It has been suggested to me that the taxonomic position of 

 Pharetrospongia cannot be regarded as demonstrated till its spicules 

 have, in some instances at least, been found to exist in a siliceous 

 state. This, however, is a view which I must confess myself unable 

 to accept ; the arguments as to the affinities of the genus which 

 have already been adduced, are, to my mind, and will be, I believe, 

 to most spongiologists, conclnsive. At the same time, since one 

 might naturally expect to find some of the spicules of a siliceons 

 sponge still retaining their original composition even in a fossil 

 state, and since it is impossible to demonstrate a fact too fully, I 

 have now made a fresh search by a method I had not previously 

 employed to see if the evidence which had been demanded was 

 forthcoming, and with affirmative results. 



My first experiments were conducted on specimens which had 

 been filled in with chalk-marl. The polished face of a thin section 

 taken from one of these was etched with dilute acid, and, after 



