STRUCTURE OF FLOWERS. 95 



petalous flowers. When this multiplication is limited to 

 one or two inner rows, it happens then either that the 

 inner corolla bears the stamens as usual, or that they are 

 absent. In this last case we may say that the corolla 

 results from the simple transformation of the stamens 

 into petals ; but in the first it is evident that there has 

 been a multiplication of the usual rows. The same 

 phenomenon is also met with in poly petalous flowers, as 

 the Pink, &c. 



5th. The Stamens very frequently present this mul- 

 tiplication of rows, especially in the genera where the 

 number of rows is naturally considerable ; thus on com- 

 paring together several flowers of the same species of 

 Poppy, we find that the number of their verticils is very 

 variable. 



6th. Finally, the Carpels, which are less numerous 

 and more central, rarely present this accidental multipli- 

 cation ; however, from time to time we find double rows 

 among the Ranunculaceae and Rosacea^ with verticillate 

 carpels. I have met with a very remarkable example of 

 this accident in Gentiana purpurea : I give a figure of it 

 in PI. 18, figs. 6, 7. Here we see two rows of ovuliferous 

 carpels ; the outer of four, the inner of two. 



But if all the floral organs can accidentally present 

 the multiplication of the rows of which they are usually 

 composed, is it not likely that this phenomenon may be 

 constant in certain plants, perhaps in certain families ? 

 And are not the genera, such as Nymphcea, Mesem- 

 bryanthemum, &c , where the parts of the flower are 

 presented in a very great and indefinite number of rows, 

 evident examples of this opinion ? 



