OF THE LINNiEAN SYSTEM. 29 



No flower being known constantly possessed of 

 eleven stamens, the eleventh class of Linnaeus con- 

 tained those plants which were said to have twelve, 

 and therefore entitled Dodecandria ; but as there are 

 scarcely any plants in existence with exactly twelve 

 stamens, all plants were comprehended in this class 

 possessed of any number of stamens from eleven to 

 nineteen inclusive. This slender distinction of num- 

 ber, however, where irregular and inconstant, and 

 more than ten, does not deserve to form the basis of 

 any particular class ; and all the plants of Dodecandria, 

 according to the insertion of the stamens, may be 

 conveniently distributed in one or other of the two 

 following classes ; for, without this generalizing, spe- 

 cies of one natural genus might be dispersed into two 

 different classes, as in Hudsonia, where some species 

 are Dodecandrous, and another Icosandrous ! 



All plants having more separate stamens than ten, 

 if we abolish Dodecandria, will belong to one of the 

 two following classes, in which, the mere number of 

 stamens is no longer of importance, being inconstant, 

 and the insertion or situation of the stamens alone 

 distinguishes the class : thus, in Icosandria they are 

 seated upon the calyx or corolla (as in the Apple and 

 the Rose) ; but in the class Polyandria, on the base 

 or receptacle of the flower (as in the Columbine and 

 Poppy). This difference of situation, in this system, 

 is only attended to in the flowers of these two classes, 

 which have many stamens. The name Icosandria 

 (from the Greek uxoov, twenty, and wvt^q, a mm, by 

 allusion a stamen), would indicate apparently a class of 

 flowers with twenty stamens ; in many of our orchard 

 fruits this is about the usual number ; but in the Rose 

 and Cactus there are many more, and their insertion 

 alone, either immediately on the calyx, or on the 

 3* 



