en the Geography of Plants, 



245 



this likewise proceeds from an erroneous method of calcu- 

 lation. 



The second chief defect of the Humboldtian method is this, 

 that in making use of particular Floras we consider only the 

 plants of certain degrees of geographical longitude, and not 

 the whole zone ; or, as the author expresses it, all those which 

 appear within the same limites isothermi. The author, indeed, 

 supposes that the most exact uniformity is always found within 

 these limits, (No. 1, p. xviii ; No. 2, p. 5), so that by means of 

 some families, the whole, and the number of plants belonging - 

 to the other families in any given district may be computed ; 

 but his own numbers entirely contradict this opinion. I will 

 merely offer a few remarks upon the following variations be- 

 tween the Floras of Germany, France, and North America, as 

 given by the author. 



Germany. 



France. 



Nortl 

 America. 



n 



Cyperoideae 



Junceae 



Labiatse 



Ericinae et Rhododendra 



Compositse 



Umbelliferae 



Cruciferae 



Caryophylleae 



Amentaceae. • . 



1 



is 

 4 



'A 



I 



i 



The endeavours to explain this striking difference by the 

 circumstance, that North America extends itself further to- 

 wards the north, as well as towards the south, than the 

 two cited European countries, and therefore its Flora has 

 a character partly polar and partly tropical. But this ex- 

 planation is not sufficient, for, in the first place, it appears 

 that these two particulars should, to some extent at least, 

 have reciprocally this effect, viz,f that North America should 

 possess a greater number both of polar and of tropical plants ; 



