STRUCTURE OF FLOWERS. 121 



it is called Chlorophylle) or otherwise coloured,* we 

 should easily understand that very slight physiological 

 modifications can cause changes of colour, and conse- 

 quently that petals were only simple degenerations of 

 foliaceous organs. 



What we have said of the leaflets, calyx, and involu- 

 crum, may as justly be said of the carpels, for the ovaries 

 are sometimes in a foliaceous state, sometimes coloured, 

 without there being any other essential difference in 

 their organization ; thus, very nearly related plants, as 

 Ornithogalum and S cilia, have the ovary green and of a 

 foliaceous appearance in the one, and coloured and 

 petaloid in the other. There are few families where 

 we do not find the same disparity between analogous 

 genera. 



But if the bracts are leaves, which no one has con- 

 tested ; if the sepals are leaves, as can hardly be doubted ; 

 if the carpels are leaves, as seems to me to have been 

 demonstrated above ; if all these organs, although of 

 foliaceous origin, be however more or less susceptible of 

 being coloured and becoming petaloid; how can the 

 petals themselves be different? Why cannot they be 

 leaves more constantly metamorphosed than the others ? 

 This supposition will presently acquire more force when 

 we pursue our researches in the opposite direction, i. e. 

 when we have examined if the organs usually petaloid, 

 can present themselves in a foliaceous state. 



I may mention, as an example of this, the carpels 

 changed into leaves, which are observed in Lathyrus 

 latifolius, in a variety of the Cherry, &c. which I have 

 already mentioned. But these examples may be doubt- 

 ful, because the carpels, in the ordinary state, are almost 



• Since writing the above, this supposition seems verified by the expe- 

 riments of Macaire, from which it appears to result that the coloure d 

 chromule only differs from the green by being more oxygenized. 



