28 THE VOYAGE OF H.M.S. CHALLENGER. 



columns often lie very close to each other, as in fig. 10, a section taken parallel to the 

 surfoce of the disk on a deeper plane (fig. 9) shows their diameters to be smaller, and the 

 intervals between them (fiUed-up by the shell-substance of the disk) to be wider ; and 

 another cross-section taken just above the plane of their junction with the annular cord 

 (fig. 7) shows the still further reduction exhibited laterally in fig. 8. Now it is this 

 deeper and slenderer pedicle of each columnar segment (fig. 8) that receives the stolon- 

 process from the sarcodic cord of the annulus next interior to its own ; and it is the 

 connection of this pedicle with the sarcodic cord of its own annulus that brings that 

 cord into continuous connection with that of its interior annulus. Thus, while both 

 series of columnar sub-segments of any one annulus are all connected together by its 

 annular sarcodic cord, the connection between the successive annuli is established by the 

 radial stolon-processes that pass from the upper and lower margins of each annular cord 

 to the upper and lower columnar sub-segments of the next annulus. 



When this arrangement has been rightly apprehended, there is no difliculty in under- 

 standing what is otherwise somewhat perplexing in the structure of the calcareous 

 disk. When the surface-layer of an empty disk has been removed by grinding or by the 

 action of acid, so as to lay open the chamberlets that lodge the columnar sub-segments, 

 these chamberlets are looked into from above (PI. III. fig. 12), not in the direction of their 

 axes, but in lines more or less oblique to them ; so that, instead of seeing downwards into 



the annular canals, we are really looking, in each charaberlet, 



A s ?^ L "Un against the oblique septum that separates it from the cham- 



^--.!.^^p.4;^a..|p^.._.^ berlet of the next interior annulus, — as is shown in the 



"^ tin (1 1 (■ (^-ac accompanying diagram (fig. 4), in which a b, a' b' are the super- 



/v- 'M -W- -ff- m' ' ficial planes of the disk, ch, ch, the chamberlets lying obliquely 



, jW/ ^ / /W /^m, to it, s, s, the septa that divide them, and o, o, o, o, the lines 



'^j,^^ ^'^ of sight. The pores, p, p', seen in each hollow correspond 



to the marginal pores, mp, m'p' , of the peripheral ring ; 

 being the outlet of the passages which lead into each chamberlet from the annidar 

 canal, ac, ac, of the interior ring, and which convey from its sarcodic annulus the 

 radial stolon-processes that originate the new columnar sub-segments. When, again, 

 we carry our section through the median plane of the disk, we lay open the con- 

 centric annular galleries (PL III. fig. 11); and along the concave (or inner) borders 

 of the septa that divide them, we see the small pores forming the entrances of the 

 passages just described, which lead to the chamberlets of the next annulus ; while along 

 their convex (or outer) borders of the septa (as shown also in the transparent section, 

 PI. IV. fig. 7) are seen the larger oblique passages, which are occupied by the pedicles 

 of the columnar sub-segments of their own annuli. W^hen this median stratum has been 

 removed, the chamberlets of the lower layer are laid open (PI. III. fig. 9, a) ; and these 

 being viewed, like those of the upper, in an oblique direction, but being seen from their 



