212 OllIGlN OF CULTIVATED PLANTS. 



admitted two species of this fruit tree of the family 

 of Myrtficcre, the one with elliptical or spherical fruit, 

 with red flesh, Psidhifu iiomlftrum ; the other with a 

 pyriform fruit and white or pink flesh, more agreeable 

 to the taste. Such diversity is also observed in pears, 

 apples, or peaches ; so it was decided to consider all the 

 Psidii as forming a single species. Raddi saw a proof 

 that there was no essential difference, for he observed 

 pyriform and I'ound fruits growing on the same tree in 

 Brazil.-^ The majority of botanists, especially those who 

 have observed the guava in the colonies, follow the 

 opinion of Raddi,^ to which I was inclined, even in 1855, 

 from reasons drawn from the geographical distribution,^ 



Lowe,^ in his Flora of Madeira, maintains with some 

 hesitation the distinction into two species, and asserts 

 that each can be raised from seed. They are, therefore, 

 races like those of our domestic animals, and of many 

 cultivated plants. Each of these races comprehends 

 several varieties.^ 



The study of the origin of the guava presents in the 

 highest degree the difficulty which exists in the case of 

 many fruit trees of this nature : their fleshy and some- 

 what aromatic fruits attract omnivorous animals which 

 cast their seeds in ])laces far from cultivation. Those of 

 the guava germinate rapidly, and fructify in the third 

 or fourth year. Its area has thus spread, and is still 

 spreading by naturalization, principally in those tropical 

 countries which are neither very hot nor very damp. 



In order to simplify the search after the origin of the 

 species, I may begin by eliminating the old world, for it 

 is sufficiently evident that the guava came from America. 



* Raddi, Di Alcune Specie di Pero Indiana, in 4to, Boloprna, 1821, p. 1. 

 ' Martius, Syst. Nat. Mcdicm Bras., p. 32 ; Bliime, Museum Lugd.- 



Bat., i. p. 71 ; Hasskarl, iu iVora, 18-14, p. 589 j Sir J. Hooker, Fl. of Brit. 

 Ind., ii. p 468. 



Gdogr. Bot. Pais., p. 893. 



* Lowe, Flora of Madeira, p. 266. 



* See Blume, t'bid. ; Dcscourtilz, Flo7-e M^dicale des Antilles, ii. p. 20, 

 in which there is a good illustration of the pyriform guava. Tussac, 

 Flore des Antilles, gives a good plate of the round form. These two 

 latter works furnish interesting details on the use of the gnava, on the 

 vegetation of tl;e species, etc. 



