270 FAMILY NUMMULINIDA. 



a considerable interval being sometimes left between them, so as to become continuous with 

 that which forms the posterior face of the next septum, as is well shown in many of 

 MM. D'Ai'chiac and Ilaime's beautiful figures. Hence it appears that each segment had its 

 own proper and complete investment, and that the marginal cord is really to be considered as 

 representing the " intermediate" or "supplemental" skeleton of Calcarina (•[[382). Where, on 

 the other hand, the septal lamellae abut against the " spiral lamina," they become absolutely 

 continuous with its innermost lamella, the structure of w^hich is precisely the same as that of 

 the numerous minutely-tubular lamellae, of w hich (as in Opcrc/ilina, ^ 442) its whole thickness 

 is made up. 



461. This minutely- tubular structure is as distinctly seen in thin sections of Nummu-- 

 lites whose texture has not been altered by fossilization, as it is in Opercidina ; but sometimes 

 the texture of tlie shell is so altered by fossilization that the tubular structure is replaced by 

 a minute prismatic arrangement. The explanation of this is, I believe, to be found in the 

 mode in which the shell is originally formed, which is doubtless the same as in Opercidina, 

 (^ 442). — The shell-substance over the septa (as shown at h, h, fig. 6) is not traversed by 

 tubuli, and is thus more transparent than the rest ; and it is of this dense hyaline substance 

 that the septal bands are composed, which are visible over the whole surface of the " assiline" or 

 " explanate" Nummulites. These bands are often broken up externally (as in Operculbia, 

 *{ 436) into rows of tubercles ; and such tubercles form the bases of columns of non-tubular 

 substance, which are traceable into the superjacent investments formed by the later convolu- 

 tions of the spiral lamina, so that they may even be continued through several of these to the 

 external surface, as shown at e, e, '-, fig. S. I have not met in the portion of the spiral lamina 

 that bounds the median portion of the chambers of an)'- Nummulite, with those intermediate 

 spots of non-tubular substance which are very common among OpcrmliiKe (^ 437) ; and 

 though such intermediate spots frequently occur in the portion of the spiral lamina which 

 invests the preceding convolutions, they are really formed on the basis of the septal tubercles 

 of the included whorls.- — ^The " marginal cord"which is shown in horizontal section at n, a, 

 figs. 1, 3, 4, 6, 9, and invertical section at a, a', figs. 3, 7, and of which the furrowed surface 

 is seen at a", fig. 7, bears a close general resemblance to that of OpercuHiia (^ 444) ; but 

 differs from it, as will be presently showii, in the disposition of the canal-system (^ 464). 



462. When a typical Nummulite is divided by a section passing through its centre per- 

 pendicularly to the median plane, it presents an aspect (as shown in fig. 10) which remark- 

 ably difi"ers from that of a typical OjiermUna. For although the spiral lamina in the latter, as 

 well as in the former, is usually continued to the centre of the spire, yet in Opercidina it is 

 generally applied so closely to the lateral surfaces of the included whorls, that the cavities of 

 the chambers are but little or not at all extended inwards by alar prolongations ; and if, as 

 sometimes happens (Fig. XLII, a, e, p. 255), such prolongations should exist, they converge 

 uniformly towards the centre of the spire in which the alar prolongations of their septa all meet 

 (see Plate XVII, fig. 1). In NumnmVma, on the other hand, it is the rule (the exceptions being 

 chiefly found in the ' assiline' or ' explanate' group) for the investing portion of each convolution 

 of the spiral lamina to be separated from the lateral surfaces of the convolution it includes, 

 by the alar extensions of the chambers of its own marginal portion, which extensions are 



