30 EXPLANATION OF THE CLASSES 



claws or bases of the petals, decides what plants ought 

 to be referred to this class.* 



The class Polyandria (from nolvg, many, and arijp) 

 differs only from the preceding in the insertion of the 

 stamens, which may be, if we abolish Dodecandria,from 

 eleven to one thousand ; these are always situated on 

 the base or receptable of the flower, and fall off with 

 the petals. But in the Rose and many orchard fruits 

 of the preceding class the stamens adhere to the per- 

 manent calyx. 



In the next class Didynamia (or of two powers, 

 in allusion to the unequal length of the stamens, which 

 are only four in number), the proportional length is 

 the essential character, two being longer than the 

 other pair. In such flowers, also, there is almost uni- 

 versally an irregularity in the form of the corolla, 

 which is always monopetalous ; and, in fact, you will 

 immediately perceive in the Didynamous class of Lin- 

 naeus, the labiate and personate groups with which 

 you are already acquainted ; so that here, as in sev- 

 eral other instances, the artificial and natural method 

 of arrangement agree together. 



Your Cruciform flowers form, also, Linnaeus's next 

 class of stamens with different proportions in length, 

 which he terms Tetradynamia. These have four 

 stamens longer than the other two, which gives rise 

 to the name of the class. The flowers are remark- 

 able, in having, contrary to the usual symmetry in 

 the structure, six stamens, and only a calyx and 

 corolla of four parts ; yet two of the six stamens 



* Calycandria, in allusion to the insertion of the stamens in 

 this class, would have been a preferable name to that of Icosandria, 

 so commonly deceptive ; and such a term, which I had long thought 

 of. has been employed by my friend Dr.' Darlington, in his Cata- 

 logue of Plants growing round Chester, Pennsylvania, 



