1849.] CARPENTER ON THE STRUCTURE OF NUMMULINA. 27 



fig. 8 obtained by a vertical fracture, passing from the outer surfaces 

 towards the medial plane, in a somewhat conical form, contracting 

 as they descend ; but they are never directly traceable into the 

 chambers of the central plane. MM. Joly and Leymerie do not ap- 

 pear to have determined their mode of connection with these cham- 

 bers ; although in their ideal figure of the animal of Nummulite 

 (fig. 5), they represent pseudopodia passing out, it must be sup- 

 posed, through such apertures. I have satisfied myself, however, by 

 a careful examination of numerous specimens, that they always ter- 

 minate over the septa (as shown in fig. 8), and actually pass into the 

 interseptal spaces, which we have already seen to have numerous 

 communications with the chambers themselves. I am confirmed in 

 this view, by finding that wherever they penetrate the investing layer 

 of any whorl, the perforations pass between the two laminse of the 

 prolongations of the septa, which are double along their whole course 

 (as maybe seen in fig. 6). These laminae diverge to give them pass- 

 age (a, a), and then reunite, thus completely enclosing them, and 

 cutting them off from the vacant spaces between the investing whorls. 

 Thus a direct and continuous tubular connection is formed between 

 the interseptal spaces of the central plane, and the external surface 

 of the shell ; and as these spaces are connected, by numerous small 

 apertures in the septa, with the chambers between which they are 

 interposed, we see that no chamber, however deeply buried beneath 

 the investing whorls, is cut off from communication with the medium 

 inhabited by the animal. In the Nu7mnulina complanata (fig. 1/) and 

 other species, in which every investing whorl is in contact with the 

 one it incloses, except at its edge, the perforations have the form of 

 fissures, that correspond with the subjacent septa, towards which they 

 directly pass. These fissures are usually found to be filled with opake 

 matter ; and the dark bands thus formed in a transparent section 

 (fig. 12, a, a) are seen to be crossed by dehcate white lines, which 

 seem to indicate a division of the fissure into a number of tubes of 

 irregular form, — probably for the passage of pseudopodia. 



All my observations tend, therefore, to confirm the opinion gene- 

 rally entertained, that the Nummulites are members of the group of 

 Foraminlfera ; and that each chamber may have been tenanted at 

 the same time by a living segment, connected with those before and 

 behind it by means of one or more tubular prolongations, and ab- 

 sorbing its nourishment from without by means of filamentous pseu- 

 dopodia projecting through the system of passages leading from the 

 medial plane to the external surface. Every whorl, as we have seen, 

 retains its connection with the exterior by means of the vertical pass- 

 ages ; and as we do not find those of the inner whorls blocked up by 

 the investing layers of the outer, but as, on the contrary, they are 

 invariably continued through them, the inference appears to be justi- 

 fied, that the segments of the animal inhabiting the chambers of the 

 inner whorls did not lose their vitahty when thus more deeply inclosed. 

 It will obviously be only from the outer whorl, however, that the 

 marginal pseudopodia can issue ; and this would give to those seg- 

 ments an advantage which we may not unreasonably suppose them 



