20 II. ORDERS AND GENERA. 



from tlie general plant, as well as from the flowers and 

 fruit, or even in j)reference to the flowers and fruit. Had 

 the genera continued to be so instituted, — as was the case 

 before the sexual sj^stem of Linneus fixed attention so 

 exclusively on the flowers, — the ordinal groups of the 

 present time would probably have differed considerably 

 from those now adopted by systematists, and in such case 

 might likely enough have been found more in accordance 

 with geographic grouj)s. 



But the inconveniences which most interfere with the 

 investigations of the Phyto- geographer, in present ar- 

 rangements about orders and genera, do not arise much 

 from a want of naturahiess ; that is, from want of suffi- 

 ciently close resemblances between the included plants, 

 or from too slight differences between the included and 

 excluded plants. Nor do they arise very much from the 

 want of uniformity in the principles upon which orders 

 and genera — the major and minor groups — are consti- 

 tuted. They are found chiefly in the want of greater 

 uniformity of " vahxe " among the groups to which the 

 same abstract term is ajpplied. 



Botanico - statistical comparisons between different 

 countries, and more especially between dissimilar cli- 

 mates, are very usually attemj^ted by reckoning the pro- 

 portions of orders, or the numbers of their included 

 species. Yet such comparisons cannot be satisfactorily 

 made by reference to ordinal groups, unless these groups 

 themselves are formed on some uniform princij)le, so as 

 to render them equivalents of each other in character or 

 structure. At present this appears to be impracticable ; 

 for the best systematists fail in the attempt. Perhaps 

 the best approximation yet made towards the desiderated 

 uniformity, is seen in the " Alliances " of Dr. Lindley. 

 Yet those alliances do appear forced and arbitrary in 



