382 FORAMINIFERA inosculate, so as to break up the alar prolongations into subdivisions, the partitions between which (as seen in tig. 26) show themselves in vertical sections in far greater numbers than the proper septa. TIG. 26. Vertical section of portion of NummuUte, showing the investment of the earlier whorls by the alar prolongations of the later: a, marginal cord; b, chamber of outer whorl; r, c, whorl invested by a: d, one of the chambers of the fourth whorl from the margin ; e, e , marginal portions of the enclosed whorls; /, investing portion of the outer whorl; g. g, spaces left between the investing portions of successive whorls ; h, h, sections of the partitions dividing these. And by a further increase of this subdivision, a complete reticula tion is formed between the successive whorls of the spiral lamina, by the complex inosculation of the septal prolongations. Now, although it would not be diffi cult to arrange a collection of Nummulites under a num ber of specific types founded on these characters, if account were only taken of the specimens showing the most strongly marked differences, yet when due attention is given to the intermediate forms, they are found to present such a transitional gradation as to make it im possible to draw any absolute differential lines between the numerous components of the series. The canal system oi Numiiiulina bears a close re semblance to that of Opera*- FIG. 27. Internal cast of two chambers, a, a, Una: and it is sometimes of Awmmfrfifl^ the radial canals between , " . , , .. r , ,. them passing into 6, marginal plexua brought into beautiful dis tinctness in fossil specimens by the infiltration of ferruginous or other fossilizing material. And sometimes fragmentary "internal casts" present themselves, which admirably display the distribution of parts of the system (fig. 27). We find in the larger forms of the Nummuline type the same tendency to become " labyrinthic," by the subdivision of the prin cipal chambers into chamberlets, " that ve have seen in the Mili oline. Just as Pene- roplis is converted by such subdivision into a spiral Orbiculina, and this, again, by the early exchange of the spiral for the cyclical plan of growth, into the cycli cal Orbitolitc, so Opcr- culina, without any other departure from the ordinary Numnm- line type, becomes Heterostngina (fig. 28) ; whilst the substitution of the cyclical for the spiral plan of growth produces the discoidal Qydodypcus (fig. 29), the largest existing Foraminifer yet discovered, its diameter sometimes reaching 2} inches. This is the recent representative of the very important fossil type Orbitoules, examples of which present themselves very abundantly in many localities in association with Nummulites (fig. 25, c), which they so closely resemble as to be easily mistaken for them. Some of them are very thin flattened disks, of the pro portion shown in vertical section in fig. 30 ; while others have two or three times the proportionate thickness. Eacli disk has a central plane of chamberlets (figs. 31, 32, b, b) resembling that of Cyclodypeus, arranged with a general but not constant regu larity in concentric, Circles ; but whilst in the recent typo this central plane is enclosed above and below by successive lamellaj of tubular shell-substance not separated from each other by any FIG 28. ffetensttfftna. intervening spaces, the superposed lamella; of Orlitoides are com posed of chamberlets of irregular form (figs. 31, 32, a, a), which are piled up, one upon another, communicating vertically by minute FIG. 29. Cycloclirpeui. apertures in their shell-walls. This arrangement obviously cor responds with that of the most reticulate forms of Nummulites. FIG. 30. Vertical section of Orbiloides; a, primordial chamber. Another important link of affinity between Orbitoides and Cyclo- clypeus is furnished by the disposition of the canal-system, which, allowance being made for the cyclical plan of growth of the former, as contrasted with the spiral plan of the latter, is essentially the same in the two types. This, moreover, is beautifully shown in " internal casts " of Orbitoide.-t found by Professor Ehrenberg in the Greensand ; which, though fragmentary, display a canal- system precisely resembling that which the writer had worked out in Cyclodypeus by the examina tion of sections of the shell. It has been by the study of such " internal casts," recent and fossil, taken in connexion with the micro scopic appearances shown in thin sections of its calcareous layers, that the Foraminiferal nature of the Serpentine-lime stone of Canada has been elucidated, of Which Some of Fir. 81.- Horizontal section of disk of 1.1 3 , -I j e ,i r Qrbitoul?.i showing at ft, 6 the chain- the deepest beds ot the Lau- bt .,. ll!ts of ,<, nH , dian ,, ]nne , aild at rentian formation {the earliest . tllose of tlie superposed iameii. series of stratified rocks at present known) are composed, and which has its parallel in certain beds that lie near the base of the " fundamental gneiss " of central Europe. In order that the evidence of the organic origin of this rock a point of the highest palreontological and petrological interest may be properly understood, regard must be had to what is even now going on in the depths of the ocean. FIG. 32. Chamberlets of the superposed lamella 1 a. and of the median plane b, of Orbitoides, more highly magnified. The animals that form the enormous mass of globigerina ooze at present in course of deposit over vast areas of the sea-bottom increase and multiply by limited gemmation ; that is, the size of each individual shell, which depends upon the number of segments of which it is composed, is restricted by the cessation of continuous gemmation after n definite number of segments has been produced. An inde finite extension of the type, however, is provided for by the detachment of further segments, which, increasing by gemmation, in their turn give origin to new shells. In the
