1S8 FAMILY GLOBIGERINIDA. 



(separating the chambers from each other) or secondary (partially subdividing the chambers), 

 are obviously formed by a doubling-in of the outer wall, so as to make each septum consist 

 of two laminae (fig. 13, «, a) ; this is seen also in sections of the incomplete septa (fig. 13, b), 

 as well as of the ridges which may be considered as rudimentary septa (fig. 18, h, b, b). The 

 two layers sometimes separate from each other, as shown in these figures, so as to leave 

 intraseptal spaces ; and these form a tolerably regular canal-system, which may be traced 

 throughout the network of ridges that covers the inner wall of each principal chamber, and, 

 through the primary septa, into the ring that surrounds the vertical canal (fig. \\,r/,g'). 



312. Notwithstanding these marked peculiarities in the general plan of conformation of 

 Carpenteria, a comparison of specimens in difierent stages of evolution, and the removal from 

 the older specimens of one whorl after another until the original nucleus is arrived at, make 

 it evident that the early condition of this organism essentially accords with that of the Globige- 

 rine type of Foraminifera, — its approximation being the closest to Eolalia in general form, 

 but its tendency being rather towards Globiyerina in the circumstance that its chambers do 

 not seem to communicate directly with each other, but that each has a separate external 

 orifice directed towards the umbilicus. Various aspects of this first-formed portion of the 

 shell, thi'ee of them showing the animal substance contained in the chambers, are seen in 

 Plate XXI, figs. 1 — 5. Now supposing that a Globigerina were to grow in such a manner, 

 attached by one of its surfaces, that the walls of its successively-formed chambers came into 

 mutual contact, and that these chambers were so shaped and so piled one on the other as to 

 give to the entire shell a conical form, each chamber opening by its own separate orifice into 

 an umbilical funnel, we should have the essential type (so far as its shell is concerned) of 

 Carpenteria ; and this is really the mode in which the latter type is superinduced upon the 

 former, as the development of the organism advances. It is further interesting to observe 

 that the great size of the chambers which form the superficial whorl of Carpenteria, has every 

 appearance of being due to the deficiency of that complete segmentation, in the later stages 

 of growth, which characterises the earlier ; for every one of the lociili marked out by the 

 ridges projecting into the interior corresponds so closely both in size and general aspect with 

 an entire chamber of the earlier whorl, that the areolation of the outer wall may be regarded 

 as a sort of attempt at that complete subdivision of the cavity, which we shall see to be fully 

 carried out in Heterosteyina. 



313. The animal substance which occupies the chambers (figs. 1 — 4) is peculiai'ly 

 spongeous in its texture ; not only possessing (according to the evidence afforded by the dry 

 specimens which alone I have had the opportunity of examining) far more consistence than 

 the sarcode-body of the Foraminifera, but being supported, in the larger chambers at least 

 (fig. 11), by sponge-like spicules (fig. 17), whose form resembles that of the simplest spicules 

 of Halichondria, and whose composition is siliceous. — It may be undoubtedly urged that as 

 the surface of dead Coral and the valves of living as well as dead Shells are frequently covered 

 with Sponges, any multilocular organism growing on such surfaces might be so penetrated by 

 the sponge, that all its chambers would be filled by the parasite. The following considerations, 

 however, seem to me strongly to militate against the applicabihty of such an explanation to 

 the present case : — 1st. That neither on the surface nor in the substance of the specimen of 



