190 NOTICES OF BOOKS. 



A Natural Order may, M. de Candolle remarks, be characteristic in 

 two senses, — ^by containing an unusually great number of species as 

 compared to what other countries do, or by containing a great number 

 as compared with the other Natural Orders in the same country ; and 

 he treats the question under both aspects. To tabulate his materials 

 he divides the globe into thirteen regions^ which are considered natural. 

 These are— 1, the North Polar, or Arctic ; 2 and 3, the North Temperate 

 regions of the Old and New World ; 4 to 7, tropical America, Africa, 

 Asia, and Polynesia ; 8, New Holland and Tasmania ; 9, New Zealand 

 and adjacent islets; 10, South Africa; 11, Kerguelen's Land, the 

 Crozets, etc., and Tristan d*Acunha ; 12, Chili, Buenos Ay res, and 

 South Brazil; 13, Patagonia and the Falklands.* 



Under these divisions the names of the Natiural Orders eminently 

 characteristic are enumerated.f 



Chapter 24. On the variety of vegetable forms in diiTerent countries, 

 and in the world at large. — An extremely valuable table is given, 

 showing the ascertained and probable total number of species in nearly 

 one hundred different countries, grouped approximately according to 

 their areas, together with the latitude and area of each. The column 

 of ascertained species apparently follows the highest estimates attain- 

 able, the British species being taken at 1520, and those of the Russian 

 empire at 63664 The estimated probable number of species is greatest 

 for South Africa, amounting, according to Drege, from 16,000 to 20,000 

 in the countries including the Cape district and from the Gariep River 



.um 



country included in the list as to excite surprise, if not incredulity. 



Under the head of " Comparison of the great divisions of the Globe," 

 M. de Candolle states that America appears to have more species than 

 any area of equivalent extent, which he attributes to the direction of 

 its mountain-chains. Africa appears to be poor in species, except at 

 its south extreme, where they are very numerous, — a circumstance he 

 can only account for either by supposing that the Flora was originally 

 that of a country with more marked differences of climates in its diiferent 



• Of these we should not regard the 1st, 7th, 11th, and- 13th as at all worthy of 

 bcin^ ranked as regions, and the 9th as doubtful. 



t Stackhousie(B are omitted in the Australian list ; SapindacecB and Begoniacea 



can hardly be considered eminently characteristic of tropical America, considering how 



very many Asiatic species there are ; Antldesrnea are omitted in the tropical Asiatic 

 Flora. * 



% A number which, as has been shown in this work. Vol. V. p. 320, is no doubt 

 capable of \ZTy great reduction. 



