790 ON THE VOLCANO OF TAAL, 



the banks of the River Carony, in the midst of the thickest 

 forests, ahnost everywhere that Indian tribes are found who have 

 had no relations with European settlements, we meet with plant- 

 ations of manioc and bananas." Humboldt suggests the hypothesis 

 that several species or constant varieties of the banana have been 

 confounded, some of which are indigenous to the new world. 



Desvaux studied the specific question, and in a really remark- 

 able work, published in 1814,* he gives it as his opinion that all 

 the bananas cultivated for their fruits are of the same species. In 

 this species he distinguishes forty-four varieties, which he arranges, 

 in two groups ; the large-fruited bananas (seven to fifteen inches 

 long) and the small-fruited bananas (one to six inches) commonly 

 called fig bananas. R. Brown, in 1818, in his work on the Plants 

 of the Congo, p. 51, maintains also that no structural difference in 

 the bananas cultivated in Asia and those in America prevents us 

 from considering them as belonging to the same species. He 

 adopts the name Musa sapie^itum, which appears to me prefer- 

 able to that of M. paradisiaca adopted by Desvaux, because the 

 varieties with small fertile fruit appear to be nearer the con- 

 dition of the wild Musce found in Asia. 



Brown remarks on the question of the origin that all the other 

 species of the genus 3Iusa belong to the old world ; that no one 

 pretends to have found in America, in a wild state, varieties with 

 fertile fruit, as has happened in Asia ; lastly, that Piso and 

 Marcgraf considered that the banana was introduced into Brazil 

 from Congo. In spite of the force of these three arguments, 

 Humboldt, in his second edition of his essay on New Spain 

 (II. p. 397), does not entirely renounce his opinion. He says 

 that the traveller Caldcleughf found among the Puris the tradition 

 that a small species of banana was cultivated on the borders of 

 the Prato long before they had any communication with the 

 Portuguese. He adds that words which are not borrowed ones 

 are found in American languages to distinguish the fruit of the 



* Desvaux, Journ. Bot. IV. p. 5. 



+ Caldcleugh, Trav. in S. Amer., 1825, I. p. 23. 



