"28 THE GEOLOGY OF BEEMUDA. 



beneath for some inches in depth, though less hard than the superficial 

 <;riist, is more firmly consolidated than the surrounding rock. The ob- 

 jects accordingly appear, when the surrounding rock is removed by 

 weathering or otherwise, as irregular cylinders. It has often been 

 crudely supposed that these cylinders are petrifactions or casts of the 

 trunks of the palmetto ; but this is certainly erroneous. I believe, how- 

 ever, that this error is but a misconception of the truth. The base of 

 the palmetto stem is convex, with numerous small roots radiating from 

 its surface. Its form is the counterpart of the shallow cup, pitted with 

 little depressions, which is the characteristic feature of the bodies in 

 question. The true explanation of the formation of these bodies appears 

 to be simply this : the rain-water trickles down around the convex base 

 of the palmetto stem, and thence follows the little radiating roots. As 

 in the other cases already discussed, the course of the waters is marked 

 by a more perfect cementing of the grains of calcareous sand, giving 

 the rock in those parts a sub-stalagmitic character. When the tree 

 finally dies, and drops out of its socket, there is left a saucer-shaped 

 cavity, lined by a sub-stalagmitic crust, and an irregular cylinder of 

 somewhat hardened rock beneath it. Sir Wyville Thomson combats 

 the idea of the organic origin of these bodies, and calls attention to the 

 frequent irregularity of their form. He tells us that a perfect series of 

 gradations maybe traced from the regulitr circular form ("the most 

 <}haracteristic, and i^robably by far the most common")* to forms so 

 irregular that their organic origin is entirely out of the question. Now 

 in maintaining that the common and typical sort of these bodies are 

 produced by the rain-waters following the course determined for them 

 by the stem of the palmetto, I by no means deny that by accidents of 

 a totally different sort special channels for the percolating waters may 

 be determined, and "calcareous concretions" produced of all sorts of 

 irregular forms. Moreover, it would be the most natural thing in the 

 world that some of the concretions whose form is determined by other 

 conditions should considerably resemble some of the least regular and 

 perfect of those formed in the way I have explained. Admitting that 

 all the "concretions," regular and irregular, are the result of the un- 

 equal hardening of the stone by the cementing action of water, the regu- 

 lar saucer-shaped cavities already described are so frequent and so 

 characteristic that it is worth while to inquire what is the special con- 

 dition which has hardened the rock in precisely that form. That ques- 



* Thomson, ojj. dt., Vol. I., p. 308. 



