STRUCTURE OF FLOWERS. 87 



a question of purely artificial order ; but it is important, 

 as regards both the natural order of classification and the 

 physiology and comparative anatomy of plants, to fix 

 our ideas upon this subject, and to be able to compare 

 together organs truly analogous. 



Those who have paid any attention to this subject 

 have thought that the floral envelope, when it is single, 

 is not a corolla, because it is often green and foliaceous, 

 and frequently adherent to the ovary, which true 

 corollas never are, and because the corolla appears more 

 disposed to be abortive than the calyx ; I only truly 

 know Nemopanthes of which it might be said with any 

 reason that it has a corolla and no calyx, but this is 

 simply because the latter is reduced to a circular rim. 



Jussieu, uniting in the definition of calyx the con- 

 ditions of Tournefort and Linnaeus, has established that 

 the floral envelope, when single, is always a calyx. 

 This opinion cannot be called in question when it occurs 

 in Dicotyledonous plants which belong to families 

 usually furnished with calyx and corolla, but which are 

 devoid of one of these organs ; in this case it is 

 evidently the petals which are absent, as, for example, 

 in Clematis, the apetalous Capparideas, Caryophyllese, 

 Rutaceas, Rosaceaa, and Ficoids. The analogy with 

 neighbouring genera evidently shows it; and if one 

 wished to maintain that some of these organs could not 

 be calyces because they are coloured, I would call to 

 mind that the calyx, and even the bracts, of Hortensia 

 and Salvia splendens are as much coloured as the most 

 brilliant corollas ; I would add, that these single en- 

 velopes perform the part of true calyces, both in 

 bearing the stamens in calyeiflorous plants, and not 

 doing so in thalamiflorous ones, and in their being fre- 

 quently adherent to the ovary, Sec. 



The question is more difficult when it takes place in 



