134 POPULAR GEOGRAPHY OF PLANTS. 



it is to prove seemingly different plants to be identical, by 

 detecting in each the one link which connects them. 



An instance is given of the manner in which species have 

 been erroneously multiplied, in the case of two different 

 kinds of Oxalis ; of the varieties of one of them, Oxalu cor- 

 nwulata, "no less than seven or eight species have been 

 made, most of them supposed to be peculiar to New Zea- 

 land," whilst to the name of this plant no allusion whatever 

 is made. As similar instances appear to be of common oc- 

 currence, and as the so-called species have been catalogued 

 as such in ordinary works of reference, the result has been 

 that the number of species of known plants compiled from 

 these sources has been commonly estimated at about 

 100,000; whilst in the essay here quoted it is concluded, 

 "from a multiplicity of data/' that half that number is 

 much nearer the truth. 



Such distinctions as depend on " differences of habit, co- 

 lour, hairiness, and outline of leaves, and minute characters 

 drawn from other organs than those of reproduction," are 

 pronounced to be generally untrustworthy guides in deciding 

 upon the species of a plant, as they are the result of external 

 causes, " such as difference of climate, soil, etc., and easily 

 obliterated under cultivation." Botanists are further re- 



