AFFINITIES OF THE GENUS SIPHONIA. 819 



middle zone of flint which we described previously (p. 817). This 

 appears to be a late stage of silicification in arrested development. 



I have described the foregoing mineral changes in some detail and 

 with great exactness, because I believe that the obscurity which 

 prevails on nearly the whole subject of the mineralization of organic 

 remains is, in great part, owing to our ignorance of the precise 

 mineral replacements which have taken place, and consequently 

 will not be dispelled till we have made ourselves masters of these. 

 It is not part of my purpose now to theorize on the facts I have 

 brought forward ; that would require a paper to itself ; and I shall 

 content myself, therefore, with briefty summarizing in the following 

 Table the various mineral changes which I consider I have demon- 

 strated to have affected siliceous sponge-spicules in the course of 

 their fossilization : — 



Mineral changes of Sponge-spicules ivhicli were originally composed of 

 Colloidal Silica in combination ivith Organic Matter. 



1. Transformed into the crystalline state, with corresponding 

 elevation of refractive index. Ex. Stauronema, SipJionia, &c. 



2. Replaced by 



a. Calcic carbonate (calcite). Ex. PJiaretrospongia, SipJionia, 

 Stauronema. 



b. Eerric sulphide (iron-pyrites). Ex. SipJionia, certain spicules 

 in Cambridge " coprolites." 



c. Glauconite. Ex. Ventriculites and SipJionia, but merely as 

 an accompaniment of a coprolitic replacement. 



d. Phosphate and carbonate of lime (coprolite). 



3. Dissolved, leaving empt}^ casts. Ex. Ventriculites, SipJionia, 

 Eubroclius, &c. 



5. Classification. 



The agreement in general and mimite structure between the 

 various species we have now described is so complete, that no 

 one can doubt their generic identity. The genus to which they 

 belong we have shown to possess a skeletal structure agreeing in 

 all essential particulars with that of the Lithistidse (0. Schmidt), 

 or the Lithistina of Carter's Pachastrellidoe ; and with this family 

 group our Sipkonia must consequently be associated. 



\Ve now proceed to a closer determination of the affinities of this 

 genus. Amongst the various known Lithistids is one, Discodermia 

 poly discus, Bocage*, which is distinguished from the rest by the 

 fact that the arms of its chief spicules remain short and round, 

 whether bifurcated or not, till they break up into their terminal 

 bunch of rounded apophyses, just as we described in the case of 

 SipJionia. The two genera SipJionia and Discodermia thus show 

 the same distinctive kind of spicule ; and they so closely resemble 

 one another in their other characteristics that a description of the 



* Journ. cles Sci. Math. Phys. et Nat. Lisbonne, No. iv. pi. xi. f. 1, 1869. 

 This would belong to O. Schmidt's genus Corallistes, but differs from his C. 

 polydiscus; Bowerbank erroneously includes it with Dactylocalyx as J), poly- 

 discus, Bk. (Proc. Zool. Soc. Jan. 28, 1869). 



