384 X. GENERAL EEJIARKS. 



orders. For example, the species of Ruhiacece and Ona- 

 gracece, one with another, occur in 48 counties ; while 

 those of Liliacece and Cavipanulacece, one wdth another, 

 are found only in 21 counties. These four orders are 

 represented in the flora of total Britain by 18, 13 ; 21, 14 

 species ; the first and third, the second and fourth, being 

 thus nearly on equality, with a slight superiority for Lili- 

 acecB and Campanulace(e. But in taking small sections 

 of Britain, provincial or comital, the Rubiacece and Ona- 

 gracece would usually assume the superiority, by reason 

 of their greater frequency and wider diifusion ; the census 

 being the joint result of both these conditions of distribu- 

 tion. Inequality of area would in this manner interfere 

 in some degree also with analytic comparisons, as well as 

 with those made between total or collective floras for 

 uneqvial spaces. 



A. De Candolle has adduced examples for the purpose 

 of showing that the number of species to the league, and 

 the number of species in proportion to orders and genera, 

 are much changed if calculations are founded upon floras 

 for very unequal spaces. And he remarks (Geog. Bot. p. 

 1172) that " a great many botanists, untrained to nume- 

 rical methods, have fallen into the error of compaiing 

 numerical proportions based on tracts too unequal be- 

 tween themselves." Some similar examples will be inci- 

 dentally adduced a few pages onward in this work ; 

 especially with respect to influence of areal extent on the 

 proportions which species bear to orders and genera. 



3. The Flora of Britain. 

 Its numerical value, geographical affinities, etc. 



1. The word ^ora is here to be understood as a col- 

 lective term, employed to express the whole group of 



