302 MORPHOLOGY. 



end of the axis, or in the axils of its leaves. In the terminal flower the 

 axes of the flower and of the stem are identical ; therefore a pedicel is 

 only to be distinguished when true articulation warrants a division of the 

 axis, or the true leaves suddenly pass into bracteoles. In a gradual 

 transition no distinction is possible. The leaf, where the axillary bud 

 becomes a flower, is called a floral leaf ; if it deviates importantly in 

 form and substance from the common leaves of its plant, it is called 

 a bract. But this transition from floral leaf into bract is not sudden, 

 as both, in a rudimentary condition, are exactly similar foliar organs, 

 so we often find on one and the same stalk all intermediate forms 

 between them ; and, for instance, in Veronica fruticulosa. Delphinium 

 Ajacis, Epilobium angustifolium, Verbascum thapsus, &c., no one can 

 declare where the floral leaves cease and the bracts commence : thus the 

 distinction between many solitary axillary flowers and a spike or raceme 

 becomes already a very uncertain one, which in the perfect plant cannot, 

 indeed, be strictly sustained in the examples referred to. But the devia- 

 tion from the common leaf often goes still further : the leaflets (bracts), 

 distinct and green in a rudimentary condition, e. g. in the Dahlias, 

 become little dry shreds of membrane in their perfect development, 

 and are called palece*; or they are altogether stunted, so that no trace of 

 them can be perceived in the perfect inflorescence (as in the Composites 

 to which is ascribed a receptaculum nudum). In like manner we find 

 a stunting and final disappearance of the bracts in the UmbellifercB and 

 Boraginacece. Among the former, in which the entire complication of 

 bracts in the simple umbel is usually called an involucel*, in the com- 

 pound one involucre^ Scandix Pecten, Astrantia caucasica, Bupleu- 

 rum, and Eryngium have true floral leaves, which pass gradually into 

 bracts, such as alone exist in Daucus hispidus and Hasselquistia cordata, 

 Oreomyrrhis eriopoda. In Petroselinum sativum and Heracleum spe- 

 ciosum the bracts of the compound umbel are already stunted ; in Caucalis 

 pulcherrima, wholly gone : in Chcerophyllum aromaticum, also, the bracts 

 of the simple umbel are small ; in Anthriscus, quite stunted internally : 

 finally, in Pastinaca, Anethum, and Pimpinella, almost all have dis- 

 appeared. In the Boraginacece the floral leaves pass gradually into 

 bracts in Cerinthe, mLycopsis the bracts are arrested in development up- 

 wards, finally, absent in Symphytum. 



The Cupuliferce exhibit another peculiarity : one or more circles of 

 bracts (e. g. in Fagus\ or bracteoles (e. g. in Quercus), become blended 

 with each other, and continue to grow with the ripening fruit. This has 

 been called a cupula.\ Something similar occurs in the bracts ofJEuphor- 

 biacece, where ten bracts are usually blended together, the free apices of 

 the five inner being usually different and bent inwards, while in the 

 outer the perfectly free apices, or the bases of them, become developed in 

 a fleshy (glandular) manner. Both phenomena fall completely within 

 the definition of the spathe. In the Cruciferce there appear to be scarcely 

 any exceptions to the rule that no bracts exist ; and yet I believe I may 



* Altogether superfluous terms. 



f Better partial and general involucre. 



\ Link (Elem. Phil. Bot. ed. 2. vol. ii. p. 109.) says the cupula does not yet exist 

 in the flower. He has apparently never looked into a flowering Cupuliferous plant. 

 Moreover, there is not here any peculiar part with leaves grown to it, as he says, but 

 the cupula originates solely from blended leaves. The cupula has no similarity to the 

 succulent coat of the seed of Taxus, and is not, as Link says, proper to the AmentacecE, 

 since, in the true AmentacecE, it does not occur at all ; only in the Cupuliferce, which 

 hence derive their name. 



