10 OPvIGIN OF CULTIVATED PLANTS. 



African and southern species whicTi excited wonder, but 

 of which no trace remained after two or three winters. 



Some collectors and authors of floras are very careful 

 in noting these facts. Thanks to personal relations 

 with some of them, and to frequent references to their 

 herbaria and botanical works, I flatter myself I am 

 acquainted with them. I shall, therefore, willingly 

 cite their testimony in doubtful cases. For certain 

 countries and certain species I have addressed myself 

 directly to these eminent naturalists. I have appealed 

 to their memory, to their notes, to their herbaria, and from 

 the answers they have been so kind as to return, I have 

 been enabled to add unpublished documents to those 

 found in works already made public. My sincere thanks 

 are due for information of this nature received from 

 Mr. C. B. Clarke on the plants of India, from ^L Boissier 

 on those of the East, from M. Sagot on the species of 

 French Guiana, from M. Cosson on those of Algeria, from 

 MM. Decaisne and Bretschneider on the plants of China, 

 from M. Pancic on the cereals of Servia, from Messrs. 

 Bentham and Baker on the specimens of the herbarium 

 at Kew, lastly from M. Edouard Andre on the plants of 

 America. This zealous traveller was kind enouoh to 

 lend me some most interesting specimens of species 

 cultivated in South America, which he found presenting 

 every appearance of indigenous plants. 



A more difficult question, and one which cannot be 

 solved at once, is whether a plant growing wild, with 

 all the appearance of the indigenous species, has existed 

 in the country from a very early period, or has been 

 introduced at a more or less ancient date. 



For there are naturalized species, that is, those that 

 are introduced among the plants of the ancient flora, and 

 which, although of foreign origin, persist there in such a 

 manner that observation alone cannot distinguish them, so 

 that historical records or botanical considerations, whether 

 simple or geogra])hical, are needed for their detection. 

 In a very general sense, taking into consideration the 

 lengthened periods with which science is concerned, nearly 

 all species, especially in the regions lying outside the 



