1867.] CLAEK MATJEITnjS. . 187 



leaving only the hard envelope, to which, thin portions of the inside, 

 bearing the same appearance as that of the "svood of the supposed 

 exogens, is seen adherent. These stumps are very near together in 

 many places ; and numerous examples of the adherence of two stems 

 are seen, as may often be witnessed in living trees in this island. 

 Everything seems to indicate that a very thick forest, of trees of 

 various sizes and kinds, once occupied this spot. Those trees which 

 have disappeared, leaving only the cavities they occupied, may have 

 been of species less favourable to the process of fossilization than those 

 which remaia. In every case, the vegetable substance, how different 

 soever in size and colour it may have been, has been replaced by car- 

 bonate of lime. 



Pass Island, about two miles from Egret Island, is composed of 

 the same calcareous rock as the latter, but is not covered by so 

 thick a stratum of sandstone, x^o stumps remain on it ; but some 

 cavities show where trees have been ; and a large prostrate trunk 

 seems to indicate that one had been blown down and fossilized there ; 

 this presents the same appearance as those of the other islands. 

 Yast numbers of fossilized roots cover the surface of this island, being 

 particularly numerous and apparent on the face of a cliff on the 

 seaward side. No roots are found in Egret Island ; and those found 

 in the other islands are free from the perforations which pierce the 

 trunks. Some of them appear to have rotted in great measure be- 

 fore becoming fossilized, the bark alone remaining to indicate their 

 size and shape. 



Vakois Island stands a few rods from Pass Island, with which it 

 is connected by the reef. Its structure is the same as that of the 

 others ; but there is less sandstone upon it, and the calcareous rock 

 is crossed by more numerous veins of carbonate of lime. There are 

 two smaU caverns in it, formed by the unequal inclination of masses 

 of stone, showing the same formation within as without. At the 

 opening of one of these caverns is a large fossil stump, apparently of 

 some tree of the fig family, and a great number of roots of the same, 

 all of which appear sound. 



Booby Island, similarly situated with the last two, on the edge of 

 the reef, is about six acres in area, and differs much from the others 

 in having a considerable depression across the middle. It once con- 

 tained many fossil stumps, imbedded in loose sea-sand, which covers 

 the south end of it to a depth of from 10 to 15 feet. Most of those 

 which remain appear to belong to some species of fig, to which their 

 roots, which are very numerous, bear a striking resemblance. 



EosslL corals, of species now existing around it, imbedded in sand- 

 stone, in which are included fragments of shells, are also found here; 

 and under a mass of calcareous rock is a vein of carbonate of mag- 

 nesia (?) in a moist state. There are several other small islands 

 within a short distance of those mentioned, of similar formation and 

 offering the same phenomena. One of these is called " I'lle aux 

 Fous," and is situated on a projecting point of the reef, which ex- 

 • poses it more than any of the others to the action of the surf. It 

 is only about three rods square, and was once covered with a deeper 



VOL. XXm. PART I. p 



