34 EXCURSIONS IN MADEIRA 



bronze yellow colour led me at first to expect it might be magnetic 

 pyrites, (fer sulfure ferrifere, H.^ but its slight effect on the 

 needle, and its chrystallized form, corrected the impression. Pyrites 

 have been cited in basalt, and there has been no doubt of the fact, 

 but I believe specimens have not yet reached any cabinet. I 

 descended again for a short distance, and turning to the north, 

 pursued my course to the Coural along the edges of the beautiful 

 ravines, which appear on the left, in going by the more direct road. 

 A slight, crazy railing occasionally edged the precipices, along 

 which the narrow path ascends ; but more frequently, there was 

 not even this fancied security to assure the passenger, while he 

 contemplated the awful depths beneath. 



A conglomerate of fragments of porous and decomposing basalt, 

 above the upper or compact basalt, of inconsiderable depth, and 

 not general, was the only additionaldeposit to what I had observed 

 at the water -side. Basaltic dikes, intersecting the tufa, frequently 

 disclosed themselves in the sides of the ravines, which were highly 

 cultivated with the convolvulus batata, and an arum, which I shall 

 describe presently. The streams which usurp the bed of the tor- 

 rent until the rainy season, flowed through thick tufts of water- 

 cresses ; the honeysuckles twisted round, and hung from one ches- 

 nut tree to another ; the brambles were bending under the weight 

 of their berries, and the wild strawberry was pvishing forth its 

 compact foliage from the banks, which were lined with the most 

 elegant ferns, whilst the sides of almost all the precipices were 

 covered with vines. Here I found a plant much resembling the 

 phy sails alkakengi, but which I think must be admitted as a new 

 genus ; as the capsule, seeds, and corolla all differ : the p. alkakengi 

 is too bitter to be eaten, but the Madeira genus makes tarts of an 

 agreeable, gooseberry-like flavor". This is, I believe, the richest vine 



^ The Herschelia, which I have ventured to erect into a genus, has, I believe, been 

 figured by Curtis, as the phi/salis edulis : it bears very closely upon both atropa and 



