Geology, and Climate of the Island of Madeira. 351 



of compact basalt ; and in some places there appear masses of 

 the latter rock in the columnar form. The rocks are of less 

 height as we advance eastward ; but the line of the coast is at 

 intervals marked by masses of basalt, and in one place a hill of 

 ^onie height remains, the last high ground on the south coast, 

 on the summit of which is a chapel of the Virgin, from which 

 the rock takes its name, Monte de Nossa Senhora da Piedade. 

 On the north side of the island, the cliffs retain a great eleva- 

 tion, for some distance beyond this point. 



Immediately to the west of this chapel mount we land in a 

 small bay, where is the only piece of sand beach I believe in 

 the island. The distance to the opposite coast may here be 

 about a mile or more, the ground sloping gradually from the 

 north coast, where the cliffs are from two to three hundred feet 

 in elevation. Over the whole of this extent, and for a consi- 

 derable space to the eastward and westward, the pyrogenous 

 rocks are covered with a deep bed of calcareous sand, contain- 

 ing vast quantities of what has been described as remains of 

 former vegetation. The formation is thus described by Mr 

 Bov.ditch : — " The sands have been in some degree fixed or 

 bound by the numerous branches of forest trees which they 

 have enveloped ; for these branches (which have preserved their 

 lateral twigs) are so numerous, that 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 configura- 

 tion of the wood like a cast. In some instances, the wood has 

 entirely perished, and the envelopes are found void like tubes, 

 but most frequently the wood is found within, as a distinct 

 mass, and has become sufficiently siliceous to scratch arrago- 

 nite. Sometimes imbedded in the envelopes of the wood, but 

 generally in the loose sand of the surface, were innumerable 

 fossil shells, intermingled promiscuously, two species terrestrial, 

 the third belonging to a marine genus. The Delphinula ap- 

 proaches the D. sulcata of Lamarck, only known in the fossil 

 state, and found at Grignon. Both helices belong to the group 

 LameUatcc of De Ferrussac'*8 subgenus Heltcosti/la, Those sheila 



