Every here and there throughout the islands there are groups of bodies of very peculiar form projecting from the surface of the limestone where it has been weathered.  These have usually been regarded as fossil palmetto stumps, the roots of trees which have been overwhelmed with sand and whose organic matter has been entirely removed and replaced by carbonate of lime.  Fig. 1 represents one of the most characteristic of these from a group on the side of the road in Boaz Island.  It is a cylinder a foot in diameter and six inches or so high; the upper surface forms a shallow depression an inch deep surrounded by a raised border; the bottom of the cup is even, and pitted over with small depressions like the marks of rain-drops on sand; the walls of the cylinder seem to end a few inches below the surface of the limestone in a rounded boss, and all over this there are round markings or little cylindrical projections like the origins of rootlets.  The object certainly appears to agree even in every detail with a fossil palm-root, and as the palmetto is abundant on the islands and is constantly liable to be destroyed by and ultimately enveloped in a mass of moving sand, it seemed almost unreasonable to question its being one.  Still something about the look of these things made me doubt, with General Nelson, whether they were fossil palms, or indeed whether they were of organic origin at all; and after carefully examining and pondering over several groups of them, at Boaz Island, on the shore at Mount Langton, and elsewhere, I finally came to the conclusion that they were not fossils, but something totally different.

[Illustration:  FIG. 2.  CALCAREOUS CONCRETION IN AEOLIAN LIMESTONE, BERMUDAS.]

[Illustration:  FIG. 3.  CALCAREOUS CONCRETION IN AEOLIAN LIMESTONE, BERMUDAS.]

[Illustration:  FIG. 4.  CALCAREOUS CONCRETION, BERMUDAS.]

[Illustration:  FIG. 5.  CALCAREOUS CONCRETION IN AEOLIAN LIMESTONE, BERMUDAS.]

The form given in Fig. 1 is the most characteristic, and probably by far the most common; but very frequently one of a group of these, one which is evidently essentially the same as the rest and formed in the same way, has an oval or an irregular shape (Figs. 2, 3, and 4).  In these we have the same raised border, the same scars on the outside, the same origins of root-like fibres, and the same pitting of the bottom of the shallow cup; but their form precludes the possibility of their being tree-roots.  In some cases (Fig. 5), a group of so-called “palm-stems” is inclosed in a space surrounded by a ridge, and on examining it closely this outer ridge is found to show the same leaf-scars and traces of rootlets as the “palm-stems” themselves.  In some cases very irregular honey-combed figures are produced which the examination of a long series of intermediate forms shows to belong to the same category (Fig. 6).

[Illustration:  FIG. 6.  CONCRETIONS IN AEOLIAN ROCKS, BERMUDAS.]