302 MAMMALS. 



37. Venezuela. 



38. New Granada. 



39. Peru. 



40. Galapagos Islands. 



41. Bolivia. 



42. Guiana. 



43. Course of the Amazons and Rios Negro, and Madeira. 



44. North-east of Brazil (Maranham, Goyaz, Piauhy, Bahia, and Ceara). 



45. Western Brazil and Paraguay (Mato-grosso, Cuyaba, Paraguay, Chaco). 



46. South-east of Brazil (Minas, Rio, Saint Paul, Saint Catharine, Saint Pierre). 



47. Uruguay and Plata (Banda Oriental, Buenos Ayres as far as Chili). 



48. Chili and Juan Fernandez. 



49. Patagonia, Tierra del Fuego, and the Falkland Isles. 



50. The Antarctic Archipelago of the Shetland Isles, Georgia, the Southern Sandwich Isles, &c.* 



If we are to regard tlie eight regions first above mentioned as truly representing Decandolle's 

 idea of the regional distribution of plants (of which, however, I am by no means sure), then I 

 would with the greatest deference venture to demur to liis conclusions. The idea embraced in 

 tliem is not very different from the system proposed by Meyer and adopted in Johnston's "Physical 

 Atlas," where the earth is sejjarated into provinces, according to latitude and longitude. 



Although Dr. Joseph Hooker has perhaps done more than any other living Botanist for the science 

 of Geographical Botany, especiallj^ in the way of supplying original material, I am not aware 

 that he has anywhere expressed a definite opinion upon the great Botanical Regions, or their limits- 

 The nearest approach to this which I am acquainted with in his writings is an incidental comiDarison 

 in his Indian Flora of the plants of that country with the typical floras of other regions. f In 

 this he notices the following types as characteristic of the regions in which they occur ; viz. 



1. The Australian type. 



2. The Malayan Archipelago type. 



3. The China and Japan type. 



4. The Siberian type. 



5. The European type. 



6. The Egyptian type. 



7. The Tiopical African typo. 



8. Tho American type. 



It docs not appear whether these are all the regions into which he would divide the world ; nor 

 whether he regards them as all of equal value, but I think we may assmne it to represent pretty 

 nearly the main ideas which he entertains on the subject. So regarding it, it appears to me that 

 of all the above authors, or any others that I have met with. Dr. Sclatcr and he take the 

 most comprehensive and philosophical view of tlic subject ; and although I difler to a certain 



* Decandolle. op. cit, p. 478. 



+ Hookkr's "Indian Flora."], p. 103. 



