26 n. ORDERS AND GENERA. 



genera, and to the average number of species in an order ; 

 as if such numbers and averages expressed some real 

 knowledge regarding the peculiarities of their vegetation. 

 But such comparisons are vitiated and almost valueless 

 ■when made by numerical figures only. The absence of 

 one single well-defined order, in any one countrj', or at 

 anj^ given altitude, might imply more real peculiarity in 

 its vegetation, than the absence of a score or two of other 

 orders, such as approximate to generic groups by the 

 paucity or the pettiness of their distinctive characters, as 

 well as by the inferior numbers of their included species. 



The vegetation of an island or a mountain, from which 

 the three first-mentioned orders (page 21) were quite ab- 

 sent, — namelj', Filices, Gramina, Cyperoides, — would dif- 

 fer immensely more from the average vegetation of the 

 Earth, than would that of another island or another 

 mountain, from which only the three last - mentioned 

 orders were absent, — namely, C eratophyllacees, Callitri- 

 cliacea;, Elatinacea. The total absence of a score of 

 such orders as these latter three would be less note- 

 worthy than the absence of any one of the first-named 

 trio. So likewise, it may be said, a tract of country in 

 which the Gramma were wanting, and their place sup- 

 j)lied by CompositcB, would be a far more remarkable bo- 

 tanical province, than would be any similar tract, in 

 which the Gentianacece were wanting, and their place 

 supplied by Primulacece. 



But within the restricted area of Britain we need give 

 little present attention to the fallacies that arise from 

 statistical comparisons based on orders or genera. Ex- 

 cept in the direction of altitude, and of climate as 

 varied by altitude, the differences of vegetation must be 

 described here by species, rather than by generic or 

 ordinal groups ; although no doubt several genera, and 



