62 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. [Nov. 23, 



communication. Such a provision I believe to have been made by 

 the extension of bands of sarcode, through canals left in the inter- 

 mediate skeleton, from the lower to the upper tier of chambers. 

 For in such sections as happen to have traversed thick deposits of 

 the intermediate skeleton, there are generally found passages distin- 

 guished from those of the ordinary "canal-system" by their broad 

 flat form, their great transverse diameter, and their non-ramifica- 

 tion. One of these passages I have distinctly traced to a chamber, 

 with the cavity of which it communicated through two or three 

 apertures in its proper wall (PI. VIII. fig. 3 c) ; and I think it likely 

 that I should have been able to trace it at its other extremity into 

 a chamber of the superjacent tier, had not the plane of the section 

 passed out of its course. Eiband-like casts of these passages are 

 often to be seen in decalcified specimens, traversing the void spaces 

 left by the removal of the thickest layers of the intermediate 

 skeleton (PL IX. fig. 3). 



But the organization of a neV layer seems to have not unfre- 

 quently taken place in a much more considerable extension of the 

 sarcode-body of the pre-formed layer ; which either folded back its 

 margin over the surface already consohdated (in a manner somewhat 

 like that in which the mantle of a Ci/jorcea doubles back to deposit 

 the final surface-layer of its shell), or sent upwards wall-like la- 

 mellae, sometimes of very limited extent, but not unfrequently of 

 considerable length, which, after traversing the substance of the 

 shell like trap-dykes in a bed of sandstone, spread themselves out 

 over its surface. Such, at least, are the only interpretations I can 

 put upon the appearances presented by decalcified specimens. Eor, 

 on the one hand, it is frequently to be observed that two bands of 

 serpentine (or other infiltrated mineral) which represent two layers 

 of the original sarcode-body of the animal, approximate each other 

 in some part of their course, and come into complete continuity ; so 

 that the upper layer would seem at that part to have had its origin 

 in the lower. Again, even where these bands are most widely sepa- 

 rated, we find that they are commonly held together by vertical 

 lamellae of the same material, sometimes forming mere tongues, but 

 often running to a considerable length. That these lamellae have 

 not been formed by mineral infiltration into accidental fissures in 

 the shell, but represent corresponding extensions of the sarcode- 

 body, seems to me to be indicated not merely by the characters of 

 their surface, but also by the fact that portions of the canal-system 

 may be occasionally traced into connexion with them. 



Although Dr. Dawson has noticed that some parts of the sections 

 which he examined present the fine tubulation characteristic of the 

 shells of the Nummuline Foramuiifera, he does not seem to have 

 recognized the fact, which the sections placed in my hands have en- 

 abled me most satisfactorily to determine, — that tJie proper walls of 

 the chambers everywhere present the fine tubulation of the Nummuline 

 shell (PI. VIII. figs. 3, 4), a point of the highest importance in the 

 determination of the affinities of Eozobn. This tubulation, although 

 not seen with the clearness with which it is to be discerned in recent 



