444 M. A. DeCandolle on the Characters which distinguish 



to say, in the same direction, and with a not veiy different 

 intensity. 



It may be objected that the addition to Germany of very dif- 

 ferent countries, such as Istria, throws too much weight into the 

 scale of the Leguminosse and Conijx)sit£e. There will always be 

 some analogous circumstance in the consideration of a very ex- 

 tensive country, but the following proportions show that without 

 quitting Germany the same facts may be observed. M. Fiirn- 

 rohr compares the pi'oportions of the families in the environs 

 of Ratisbon and in Germanj'^ proper, that is to say, not including 

 Switzerland, Istria, and the province of Prussia. The propor- 

 tions arc as follows, when put into the form here adopted : — 



Ratisbon. Germany. 



(1063 Phanerog.) (290G Phanerog-.) 



Species. Proportions. Species. Proportions, 



Legiimir.asce. ... 58 0-054 1/7 0061 



Compositae .... 115 0108 352 0121 



Gramine* .... 80 0073 205 0070 



Authors sometimes compare the proportions of the families 

 in regions as extensive as the whole of Europe, or even still 

 larger. If they happen to compare one of these immense regions 

 with the environs of a town, or with a small island, the error 

 resulting from the relative area of the species may rise to 4 or 5 

 per cent., or probably even more in some exceptional regions and 

 for certain families. The influence of this cause will be par- 

 ticularly great in countries where the species change rapidly from 

 one district to the other, as for instance at the Cape, Brazil, 

 Mexico, &c. 



Notwithstanding this cause of error and that arising from the 

 unequal degree of frequency of the species, it is certain that for 

 countries of nearly similar extent, and for families in which the 

 average area of the species is not very different, these proportions 

 possess some interest and deserve comparison. 



We may also ascertain that certain families have the greater 

 part of their species collected together in a particular region of 

 the globe, without taking any trouble about the proportion which 

 they bear to the whole of the Phanerogamia in each region. 

 This is a mode of looking at the question which sometimes leads 

 to different results. 



As a general rule, two characters which it is essential to know 

 may be derived from the study of the famihes : — 



1. In every country certain families predoininafe as regards 

 the proportion of their species. This is the case with the Gra- 

 minese and Compositae in Europe, the Leguminosaj in the West 

 Indies and most countries in the neighbourhood of the Equator, 

 and the Proteacese or ^Nlyrtacese in Australia. 



