■ 115 



rect course of these canals, seeing that the same object could be 

 attained in so much easier a way through the ordinary pseudopodian 

 tubes. These anastomosing canals obviously maintain a free com- 

 munication between the various parts of the comparatively thick 

 shell. 



Fig. 7 represents a superficial section of part of one of the lateral 

 parietes of the same Nonionina, viewed as a transparent object. It 

 exhibits the densely crowded foramina (seen also in portions of fig. 

 6), as well as the horizontal direction assumed by the translucent 

 spaces (fig. 4, d, 5, c, and 7, a) surmounting the septa. These details 

 appear to throw new light upon the structure of some species of 

 Nummulites. 



The next group, to the structure of which I would direct attention, 

 is that comprehended in D'Orbigny's genus Orbiculina, including 0. 

 complanata, and previously designated Orbitolites and Marginipora. 

 I have been much surprised to find that notwithstanding what has 

 been accomplished by M. D'Orbigny in connexion with this group, it 

 is still necessary to establish their affinity with the Foraminifera ; 

 many of the leading British zoologists continuing to regard them as 

 true Bryozoa. That they belong to the former of these two families 

 I have not the most remote doubt ; their mode of growth, and 

 the details of their internal structure, alike revealing their close 

 alliance with the well-known Orbiculina adunca. 



One of the most common species is the 0. complanata. It not 

 only occurs recent in the Mediterranean, in the Cuban seas, and in 

 those surrounding the Philippine Islands, but is found in a fossil 

 state in the foraminiferous marls of the Calcaire Grossiere. A verti- 

 cal section of this species is shown in fig. 8, of a portion of which 

 section a still more highly magnified representation is given in fig. 9. 



The central or primordial cell of this object is a large spherical 

 cavity (fig. 8, a), partially separated from an adjoining one (8, c) by 

 an imperfect intervening septum (8, b). The subsequently-added 

 cells are arranged round these two as a central nucleus ; at first they 

 appear in oblique rows, and afterwards in concentric circles. In the 

 vertical section, each cell right and left of the central point repre- 

 sents one of these rows and circles. The reason why there are more 

 cells in fig. 8 to the left of the central point than to the right, arises 

 from the circumstance that the early growths do not form complete 

 circles, as is seen in fig. 10, shortly to be referred to more fully. 

 The consequence of this arrangement is, that the primordial cell 

 does not occupy the geometrical centre of the disk. 



Fig. 9 gives an enlarged view of these cells, as seen in the vertical 



