GENUS NUiMJlULIXA. 2Gy 



median plane (Plate XVIIl, fig. 5), wc sec that the mimbcr of wliorls is usually much greater 

 than in Opcrcdinn, and that their increase in breadth is by no means so rapid, especially between 

 later convolutions. The progressive increase is tolerably regular in the earlier whorls ; but 

 it afterwards not unfrcqucntly happens that the spire enlarges rather suddenly, whilst some- 

 times it contracts again to a breadth inferior to that of the convolution it embraces. The division 

 into chambers, also, is far from being regular, the distance of the septa from each other being 

 subject to great variation even in the same whorl ; and chambers much smaller than the rest, and 

 apparently ahortkc (resembling tliose shown in Fig. XXXIX, ]). 253), being not unfrequently 

 seen. It is worthy of remark that \\hen the septa thus fail to divide the whole breadth of the 

 convolution, it is invariably on its outer side (so far as my observation extends) that they are 

 defective ; thus indicating that their development does not commence from the marginal cord 

 (as supposed by Mr. Carter, xxiii«), but from the internal edge of the spire, — as might 

 naturally be inferred from the position of the aperture through which the mass of sarcode 

 projects itself, whereon the new chamber is moulded. It is occasionally observed that a con- 

 volution subdivides into two, sometimes of equal, sometimes of unequal breadth ; this abnor- 

 mality is chiefly observable in such large and somewhat contorted forms as N. (/yzehensit and 

 N. disfaiis. When the subdivision is equal, each portion soon acquires the dimensions of the 

 original convolution, and continues to maintain a distinct course, sometimes undergoing 

 further subdivisions. But it often happens that the supplemental convolution is so much 

 smaller as to be a mere appendage to the principal, into which it returns after having retained 

 its distinctness through a course of greater or less extent. I am not acquainted with any 

 other helicoid type of Foraminifera in which such subdivisions occur ; and I am disposed to 

 connect it with the peculiar distribution of the canal-system to which I shall presently direct 

 attention (5[ 464) as tending to favour growth in a radial direction corresponding to that of 

 Cijdoclnpeus, in which type a multiplication of incomplete annuli is so common as to be 

 the rule rather than the exception. In specimens whose chambers are occupied by trans- 

 parent calcareous infiltration, it not unfrequently happens that the central portion of each 

 crystalline aggregation (as shown in the upper part of fig. 5) is deeply tinged with dark brown 

 matter, which has every appearance of being the carbonaceous residue of the original sarcode- 

 body. 



4G0. The form and position of the apertural fissure which establishes the communication 

 between successive chambers, are precisely the same as in Operculina (If 440) ; and in this 

 type also we meet with " secondary pores"* scattered irregularly over the surface of the 

 septum, as shown in figs. 7, 8, 10. Each septum, in all well-preserved specimens, can be 

 distinctly seen to be composed of two lamcllce, between which the canal-system is interposed, 

 as shown in fig. 4 ; and it generally happens that the lamella which forms the anterior face of 

 one septum can be traced along the inner edge of the "marginal cord" (as shown in fig. 4), 



* These " secondary pores" were first observed by nie in N. hevigata (xii, p. 34f) : I then be- 

 lieved, however, that they do not pass through both layers of the septa, but only establish a com- 

 munication between the chambers and the iutcrseptal passages — a notion which was supported by 

 Mr. Carter (xviii), but of the fallacy of which I afterwards became convinced by my own study of 

 the alUed recent types Opercidiim and Cycloclijpeus. 



