140 EXCURSIONS IN MADEIRA 



they are spread over the surface, like a network of stoloniferous 

 roots. It is scarcely possible to set the foot on the ground 

 without treading on them. Both the branches and the trunks 

 (which stand on their roots in their natural position) are encased 

 in a thick, hard sheath of agglutinated sand, which has followed 

 the external configuration of the Avood, like a cast. In some 

 instances the wood has entirely perished, and the envelopes are 

 found void Uke tubes, but most frequently the wood is found 

 within, as a distinct mass, and has become sufficiently siliceous to 

 scratch arragonite. V. fig. 30 and 31. The tallest fragments of trunks 

 reach about a foot above the surface of the sand; how far beneath it I 

 cannot say : there were two of these as thick as my body. Some- 

 times imbedded in the envelopes of the wood, but generally in the 

 looser sand of the surface, were innumerable fossil-shells, inter- 

 mingled promiscuously; two species terrestrial, the third belonging 

 to a marine genus. 



The delphinula, fig. 33 a, b, approaches the d. sulcata of Lamarck, 

 only known in the fossil state, and found at Grignon. Both 

 helices belong to the group lamellatfB of De Ferrussac's sub-genus 

 helicostyla. The smaller species, fig. 32, is globose ; but the larger, 

 fig. 34, a,b,c, (which is one inch and a quarter in its greatest 

 diameter, and yVths deep) has the last whorl compressed, or 

 flattened. There are several helices still smaller than the former, 

 with the umbihcus exposed; but this is merely because the 

 plate which covers the columella is not entirely developed, and 

 I have not the least doubt of their being young shells of the 

 first-mentioned species. These shells are perfectly distinct from 

 the existing helices of Madeira, which I have already figured 

 or described, and there is not one to be found in this neigh- 

 bourhood. All the branches and wood appear to belong to the 

 same sort of tree (of which there seems to have been a small forest 

 on that spot), and that evidently a dicotyledon, but more than this 



