NOTICES OF BOOKS, 187 



value, and between which he himself elsewhere discriminates : these are^ 

 the numerical proportions of the species and Natural Orders, which 

 afford strictly the botanical features of a country; and the relative num- 

 ber of individuals, and their habits and appearance, which determine 

 the physiognomy of its vegetation. Different classes of naturalists, and 

 above all, naturalists with differently constituted minds, will attach 

 more or less importance to one or other of these subjects ; and the 

 fact of the first not satisfying M. de Candolle's love of the exact, or 

 rather of its not fulfilling his idea of what is exact, is rather to be at- 

 tributed to his not taking the same interest in one branch of speculative 

 inquiry, which he does in others equally barren of direct results, and 

 equaUy exposed to innumerable sources of grave error. Were it not 

 that no amount of prospective labour has deterred M. de^CandoUe from 

 the full and complete investigation of the earlier-treated subjects in his 

 work, we should be inclined to suspect that the complexity of the phe- 

 nomena, the difficulty of correlating the principal facts, and the multi- 

 plicity of detached observations requiring investigation, had influenced 

 his judgment as to the relative value of this branch of the inquiry, 



and of those that precede it. 



Chapter 22. On the comparison of the relative proportion of Mono- 

 cotyledons to Dicotyledons iji different countries.— Under this subject 

 (which however is considered as of doubtful importance) M. de Can- 

 dolle gives a tabular statement of the number and proportions of Dico- 

 tyledons and Monocotyledons in sixty-eight different countries. This 

 is a document of great value, however little it may be avaUable in 

 the present state of our knowledge for soMng the problem for which 

 it is collected and arranged. The principal laws deduced from it are. 

 that in temperate regions the proportion of Monocotyledons decreases 

 relatively to the Dicotyledons in approaching the tropics ; and that, 

 ceteris paribus. Monocotyledons prevail relatively to Dicotyledons m 

 humid countries, and the reverse in -dry countries. A very careful 

 investigation follows of the available data for determining the same 

 proportions in mountain countries, and some good observations on the 

 results of comparing mountain regions with analogous climates in this 



respect. x r -i • 



In concluding the chapter M. de CandoUe repeats that he finds it 

 impossible to attach any real importance to the proportions presented 

 by the two great classes of flowering plants, not only for the reasons 



