1867.J DR. J. E. GRAY ON SVONGES. 493 



posed to divide sponges into those with calcareous and those with 

 siliceous spicules, and those without any spicules and having only a 

 horny skeleton, and to call them respectively Spongia calcarea, S. 

 silicea, and S. keratosa. 



Dr. Bowerhank and several zoologists have adopted this arrange- 

 ment, changing the names of one or more of the divisions. 



The division between the calcareous and the siliceous sponges is 

 very distinct and natural ; the separation between the siliceous and 

 keratose sponges, on the other hand, is very indistinct and separates 

 nearly allied genera. Some siliceous sponges are entirely formed of 

 siliceous spicules, covered with a very small quantity of sarcode ; in 

 others the spicules are enclosed in a very thin coating of horny or 

 keratose matter ; in others the horny matter increases in thickness, 

 and the spicules diminish in quantity until they almost entirely dis- 

 appear ; and sometimes the place of the spicules in the horny fibres 

 is supplied by sand or other siliceous matter, which the fibres take 

 up as they are formed ; and the skeleton of some sponges is entirely 

 formed of horny matter ; and the passage of these forms into each 

 other is so gradual that I believe it is better to unite the siliceous 

 and keratose sponges of these zoologists into one group or order. 



The spicula that form the main part of the skeleton of these sili- 

 ceous sponges are of three shapes : — 



Fusiform (acerate. Bower b.), more or less cylindrical, and pointed 

 at each end. 



Needle-shaped (acuate, Bowerb.), cylindrical, blunt at one and 

 sharp at the other end. 



Pin-shaped (spinulate, Bowerb.), cylindrical, with a more or less 

 spherical head and a tapering point. 



There are spicula of many other shapes which have been repre- 

 sented by Bowerhank, Quekett, Carpenter, Oscar Schmidt, and other 

 authors ; but they are for special purposes, are found in certain 

 parts of the sponge, and are peculiar to certain forms of sponges. 

 Many spicula oifer great variation in form in the same species of 

 sponge, and also in the different parts of the same specimen and 

 according to their state of development. 



These spicules may be divided into three series, thus : — 



(1) The spicula that Dr. Bowerhank calls retentive, and designates 

 as auchorate, (hi- or quadri-) hamate, umbonate, and hi- or tri- 

 rotulate, all belong to the same series, and each presents several mo- 

 difications, sometimes in the same species of sponge, and they all 

 gradually pass into each other. 



In the same way (2) the stellate spicules and (3) the three- 

 pronged spicules each belong to a separate series offering many 

 modifications. The stellate are usually scattered in the sarcode ; 

 and the three-pronged are what Dr. Bowerhank calls tension and 

 defensive spicules, supporting the outer surface of the sponge, and 

 sometimes for this purpose even extending beyond their surface. 



The forms of the spicules are characteristic of the different fami- 

 lies, if they are not always absolutely peculiar to them, thus : — 



The many-rayed stellate, with rays on all sides, and the three- 



