NOTICES OF BOOKS. 215 



opinion of the richness of these countries enormously exceeds that which 

 our materials warrant us in forming ; and further, it is evident that his 

 estimate is not framed in accordance with the Linnaean sense of the term 



\ 



species; for instance, amongst the 1500 plants which he assigns to 

 Britain there are, besides a great number of non-indigenous species, a 

 still larger percentage of species that Linnaeus would have considered 

 as varieties,* 



On the proportion of Genera to Species i7i different countries. — The 

 uncertainty of the value of generic characters, or rather our opinion 

 that most genera are arbitrarily defined groups, renders it exceedingly 

 difficult to arrive at any definite conclusion upon this subject. Upon 

 the whole, M, de Candolle concludes that the proportion of species to 

 genera is smaller in islands than on continents, and we have no doubt 

 that he is correct. We should further expect that the proportion 

 would rise in direct proportion to the area of the island, just as it does 

 in proceeding from small areas of continents to larger ones. 



Chapter 25. On the division of the surface of the globe into Natural 

 Regions. — This Chapter is chiefly devoted to a long and very interesting 

 discussion of the merits, or rather demerits, of some of the botanical 



difficulties 



amve 



that shall be recognized by any two of more botanists, quite impossible. 

 Some remarks upon the objects to be obtained, and the errors to be 

 avoided, are also good ; but M. de Candolle gives no new attempt of 

 his 



own. 



' Por our own part, we believe that the materials do exist for an ap- 

 proximate determination of a limited number of tolerably well defined 

 botanical regions, and which may be characterized by the predominance 

 in number of species of certain Natural Families, by the features these 

 form in the landscape, and by the absence of others. The materials 

 however want tabulation, and the real reason why no attempt has 

 hitherto been successful is because no one has taken pains to ascertain 

 a sufficient number of facts in distribution. It is the number of errors 



• Nyman's careful Catalogue of European Plants, just publish^, contaiug 9700 

 flowering plants, including a very large proportion that are doubtful, besides others 

 which are certainly not species in the Linnsean sense. Allowing that there M;e as 



iscovered species in Europe as there are 

 enonnoos allowance), it would follow, a 



there 



•an extravagant assumption in the present state of our knowledge. 



