58 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. [Apr. 28, 



and Azores. If the identification be correct, the abundant presence 

 of these two species may perhaps allow us to infer (though cer- 

 tainly not with absolute confidence) that the conditions of climate of 

 the island in the time of the S. Jorge deposit were not very widely 

 different from those now existing. This conclusion will be strength- 

 ened if the leaf referred to Vaccinium Maderense (No. 15) really 

 belonged to that species ; for that Vaccinium appears to be at pre- 

 sent confined to the island of Madeira*. Besides the Pteris, two 

 other Perns of the leaf-bed seem to be very probably identical with 

 species now very abundant in the island, the Davallia Canariensis 

 and the Woodwardia radicans. The present geographical range of 

 the Davallia is tolerably well defined, and not very extensive, 

 though more so than that of the laurels and whortleberry above 

 mentioned ; it is found in the Madeira and Canary islands (not in 

 the Azores), in the north-western part of the African continent, the 

 south-west of Spain and Portugal. Of the Woodwardia, the distri- 

 bution is so peculiar and so difficult to explain, that no conclusions 

 can safely be drawn from it. 



The presence of the JErica arborea, if the leaves No. 17 be rightly 

 referred to that species, is another point of agreement between the 

 recent vegetation of the island and that preserved in the deposit of 

 S. Jorge ; for the abundance of that heath is a well-known and 

 striking feature of the Madeira forests. The Erica arhorea has a 

 wider range than the laurels and whortleberry already mentioned ; 

 it is apparently general in the countries bordering on the Mediter- 

 ranean. 



2. Certain other forms in our list, such as the Corylus australis of 

 Heer, the leaves No. 20 and No. 21, and the Adiantum ? jpsychodes^ 

 are distinctly different from any now existing, at least in Madeira 

 or the neighbouring islands, and appear to belong to extinct species ; 

 but it is not impossible that their extinction may have taken place 

 within the human period. 



3. None of the fossils can be said to belong distinctly and posi- 

 tively to tropical families. The leaves No. 20 and 21 may perhaps 

 indicate something of a tropical character of vegetation, but by no 

 means unequivocally. 



4. Those forms, among the fossils, which are different from the 

 present vegetation of Madeira do not show any marked analogy to 

 any other existing Flora. The extinct vegetation preserved in some 

 of the Tertiary strata of Europe shows a striking resemblance to 

 the existing vegetation of North America ; this certainly cannot be 

 said of the S. Jorge plants. 



5. Webb and Berthelot remark that the general character of the 

 forest- trees of the Canary Islands is to have smooth, glossy, coriaceous 

 leaves, undivided, and either entire or at most finely serrated at the 

 margin. The same observation is equally applicable to the forest- trees 

 of Madeira. Now in this respect the fossil leaves of S. Jorge show, on 

 the whole, a considerable analogy with the recent vegetation ; for 



* Unless the Vaccinium cylindraceum of the Azores be the same species, as 

 Mr. Watson thinks. 



