98 THE VOYAGE OF H.M.S. CHALLENGER. 



Findlay must have visited the island at a more rainy season. The mass of the island is 

 evidently volcanic with veins of granite (?), with gravel and iron (?), covered with a thiu 

 layer of sandy earth saturated with guano, loose, and full of rents, as though it had 

 recently undergone some violent shocks. In the middle is a hollow which appears to have 

 been the crater. 



"With regard to the vegetation, on the flat parts only one species of IAguerilla [Den- 

 droseris f\ existed ; it was usually eighteen inches high, but sometimes double that height. 

 In the ravines, where it was damp, there were other plants, of which specimens were col- 

 lected ; but no grass or edible plant was seen. Of living beings, only three kinds of sea- 

 birds were seen, and they were different from those of the Chilian coast, and so tame that 

 one had to be careful to avoid treading upon them in their nests. 



Ten nautical miles westward from San Ambrosio is the group of San Felix, which 

 consists of two islands and a rock. The main island measures two nautical miles from 

 east to west by half a mile broad, with a general height of 200 feet, rising in the north- 

 Avestto 400 feet. There is no beach except in two small spots on the north and west sides; 

 otherwise it is not difficult to reach the plateau. The moss of the island is identical with 

 that of San Ambrosio, but the thickness of the loose layer of earth is greater, being in 

 places as much as a yard. It consists almost wholly of fresh white guano, apparently of 

 excellent quality, 1 and many shiploads of it might be obtained. The vegetation was 

 limited to a single nettle ; and the birds were the same as in San Ambrosio. No drink- 

 able water was found. About half a mile west-north-west of the western point of San 

 Felix is a singular rock some 300 feet high with two pointed pyramids ; hence the 

 Spaniards have named it " Brigantine," and the English "Peterborough Cathedral." 



The specimens of plants were mere fragments, brought away by Commander Simpson 

 in his hat, and were never properly pressed ; still Dr Philippi was able to determine them, 

 with the exception of two. And two of the plants collected by Dr Coppinger are 

 evidently the same as two of those described by Philippi. The plants known are : — 



MKLVACEM 



Malva limensis, Linn. 



Malva limensis, Linn., Sp. PL, ed. 2, p. 9GS? Philippi in Pot. Zeit., 1870, p. 498. 



Sak Ambrosio. Simpson. 



Philippi hesitates to refer this unconditionally to Malva limensis, Linn., because the 

 carpels are furnished with small teeth or spines, instead of being smooth; but all the 

 specimens we have examined that are so named at Kew are more or less prickly. Malva 

 peruviana, Linn., is very closely allied, differing in its more evidently prickly carjoels. 

 Perhaps they are better regarded as forms of one species, common from New Grenada to 

 Chili. Malvastrum is the genus to which the plant is now generally referred. 

 1 A sample was subsequently analysed and found to contain very little ammonia ; hence it is of little value. ] 



