REPOKT ON THE BOTANY OF JUAN FF.KXAN I »EZ AND MASAFUERA. 7 



at that season. He published l some particulars of this collection, which was deposited in 

 the Paris Museum of Natural History. Although the time was short, he met, ho says, 

 with many interesting plants, including three or four species of tree-ferns, which were 

 overrunning more and more ground, a new genus of Drimys, a Myrtus, which he thought 

 might be Myrtus ugni, Molina [doubtless Myrtus selkirUi], an Urticaceous tree, called 

 "Manzano," a fine Sophora, two species of Gfaaphalium, a Campanula \ Wahlenbergia~], 

 a Zanthoxylum, an Arbutus [Pernettya], and even two species of Piperacete. He also 

 observed that the Juan Fernandez resin, so famous throughout Chili and Peru, though still 

 unknown to science, was exuded by a new genus of Composite, near Senecio. Among 

 other plants was a very fine series of ferns, which were very common in the island ; " but 

 my most important botanical discovery," he adds, " was five or six species of a genus belong- 

 ing to the Cichoriacese, all of them woody, and ten to twelve feet high." In this dis- 

 covery, as appears from Bertero's sketch of the vegetation of the island, published three 

 years earlier, he had been anticipated; indeed, Don had already published the genus 

 Dendroseris on specimens of a species collected in Masafuera by Cumino-. 



Mr Philibert Germain, formerly conservator of the Natural History Museum at 

 Valparaiso, paid a visit to Juan Fernandez in 1854, and was there during the latter part 

 of October and the beginning of November, as we learn from Dr E. A. Philippi. 2 There 

 are only a few of Germain's plants at Kew, but, from what Philippi says, he must have 

 made an extensive collection. As already observed, there is no direct evidence in Philippi's 

 writings that he has ever visited the island ; yet some of the information in the Berner- 

 kungen does not read like second-hand. In the place cited he enumerates 137 3 vascular 

 plants, belonging to forty-three orders, or an average of about three species to each order. 

 Among them are thirty-six species of ferns, constituting 2 6 '3 per cent, of the whole ; 

 twenty-three Coinpositse =16 per cent. ; aud ten Gramineaa = 7 per cent. Philippi's 

 list includes, besides several certainly introduced plants, such as Rumex acetosella, Aira 

 caryophyllea, and Anthoxanthum odoratum, descriptions of twenty-eight proposed new 

 species, several of which are regarded as varieties in this work, whilst a few others are 

 synonyms of previously described species, as subsequently ascertained in part by hilippi 

 himself. A few notable omissions are inexplicable, such, for instance, as Bcdbisia 4 (Rhetino- 

 dendron) and Lobelia tupa, except that they were overlooked in the haste of compilation. 

 We extract a few particulars relating to the condition of the vegetation at that date. The 



1 Apercu sur les recherches d'histoire naturelle faites dans l'Amenque du sud, et principalerucnt dans lc Chili, 

 pendant les annees 1830 et 1831, Annates des Sciences Naturelles, xxviii. pp. 369-393. 



2 Benierkungen iiber die Flora der Insel Juan Fernandez, Botanische Zeitung, September 1856, pp. 625-636, 

 641-650, S18-819. This article was originally read before the Universidad de Chile on July 12, 1856, and 

 published in Spanish in the July number (1856) of the Amies de la Universidad J>' Chile. There is also a 

 French version in the Annates des Sciences Naturelles, serie 4, vii. p. 87. 



3 Since the publication of this list, Dr Philippi has proposed and described a number of new species of Juan 

 Fernandez plants in various periodicals. 



4 Subsequently mentioned by him in his general remarks on the vegetation. 



