23 



at right angles those parallel to the general conical surface, and likewise originat- 

 ing from each joint. As therefore, the first lines of separation generate a conical 

 surface pointing upwards, so this second set generate a conieal surface point- 

 ing downwards, and thus truncate, or rather excavate, by cutting into them 

 the apices of the first cones. When the base is in such a state that the 

 lamina: of which it is formed are disposed to separate (PL. n. fig. 15.) its de- 

 tached pieces, often present this excavated conical form, altogether resembling 

 models of the crater of a volcano. These second lines of separation do not, 

 however, extend to the exterior circumference of the base, except near its sum- 

 mit, but terminate in its interior, and are not all of equal lengths, but occur 

 insets, of which the lowest line is always longest, and those which succeed 

 4t, in proceeding upwards, gradually diminish for a series of from six to ten. 

 joints, until followed by the longest line of the set next above. Such are the 

 appearances ; the explanation I am inclined to offer of them is as follows. The 

 muscular membranes which are interposed between the columnar joints, may 

 irery probably have carried on them the vessels whence the exudation of the 

 calcareous matter forming the laminae of the base proceeded ; these membranes 

 may have continued to extend themselves in the direction marked by these 

 secoud lines. The original existence of such an interposed membrane will 

 readily account for the separation which, as we have seen, still takes place in. 

 this direction, in many specimens. And as we have already seen reason to be- 

 lieve that the exudation of the calcareous matter did not proceed continuously, 

 but was rapidly thrown out at certain periods; so we may suppose all the lines 

 of a single set to have been formed by the extension of the membranes during 

 one of these active periods, and their different lengths to indicate the size they 

 had attained when that action ceased. 



The .quantity of matter accumulated round the base, indicates the degree 

 of adhesion and support the animal required ; and, perhaps, bears also pro- 

 portion to the quantity of calcareous substance accumulated in the alimentary 

 canal, and from which it became necessary to free it when inconvenient to its 

 economy. 



Portions of the column (PL. i. fig. 1. and PL. n. fig. 2. to 14.) are not 

 ^infrequently encompassed by a thick spreading ring of calcareous matter, in 

 every particular resembling that which covers them at the base, and forms their 



