282 FAMILY NUMMULINIDA. 



outer surface of each whorl, except where (in well-preserved specimens) the portion last 

 formed is as yet unconsolidated by it. For a careful examination of sections taken in 

 different directions makes it clear that whilst the internal portion of the spiral lamina that 

 forms the outer wall of each chamber is continuous with the nearest lamella of the adjacent 

 septum on either side (as is shown in the portion of fig. 1 which traverses the median plane), 

 the substance of the external portion is no less continuous with that of the calcareous nucleus. 

 The whole thickness of the spiral lamina formed by the coalescence of these two lamellae is 

 generally traversed by minute tubuli, passing in a radial direction from one surface towards 

 the other; but these have by no means either the closeness or the regularity which distin- 

 guishes the tubular structure in Operculina and Cj/cloclt/pcus, and the shell-substance is in 

 many parts so destitute of tubuli as to be of almost glassy transparence. The farrowing of 

 the external surface (^ 473) is seen in vertical sections not to be produced by mere superficial 

 excavations, but to proceed from a plicated arrangement of the spiral lamina as shown in 

 fig. 1 ; and this is related to the prolongation of the posterior margin of each segment into a 

 series of " retral processes" (fig. 9, 1-, /•)> corresponding to those of P. cri.yia (^ 175). They are, 

 however, much less elongated in this type, simply giving a crenulated margin to that angle 

 of the segment, which contrasts remarkably with the smooth unbroken aspect of its anterior 

 border. The spiral lamina which forms the outer wail of the chamber, being modelled (so to 

 speak) upon the surface of these retral processes, presents internally a corresponding series of 

 grooves, which are deepest towards the posterior margin, and become rapidly shallower in 

 passing towards the anterior margin, of each chamber, as is shown at a, a}, a", fig. 8 ; these 

 grooves are not, however, as in P. cri-ypa, completed into tubes for part of their length by an 

 additional lamella of shell given off from the septum (1| 472) ; but they are sometimes shown, 

 in sections which happen to traverse them, to be extended into ca;cal prolongations by back- 

 ward inflexions of the septa at their junction with the spiral lamina. The communication 

 between the successive segments of the same whorl is established by a number of minute 

 processes or stolons of sarcode (fig. 7, c, fig. 9, c), which pass at regular intervals between 

 their internal margins through a series of pores that can be distinguished along the inner 

 border of each septum (fig. \, c, c, c) close to its junction with the preceding convolution. I 

 have not detected in any instance, either in sections of the shell or in the siliceous casts 

 which so exactly represent the sarcode-body, any other communications between the chambers 

 or their contained segments; and lam therefore satisfied that Prof. Max. Schultzc must 

 have been misled by appearances when he stated (xcvii, p. 65) that various other parts of 

 the septal plane are marked by similar pores, — more particularly as his figures of the 

 decalcified body do not show that any other threads or stolons of sarcode pass between its 

 segments, than those just described. 



477. So far, then, the structure of this comparatively gigantic type of Polustomella accords 

 very closely with that of the more delicate species previously described. I have now, however, 

 to give an account of a remarkable feature in its organization, namely, its highly developed 

 canal-system ; which, though not entirely wanting in P. crisjxt, is so imperfectly presented there 

 that Prof. Williamson may well be excused for having overlooked it, especially when it is 

 borne in mind that at that period the existence of such a system in Foi-aminifera was alto- 

 gether unknown. The general arrangement of this canal-system may be most readily aj)pre- 



