PHANEROGAMIA : FLOWERS. 303 



venture to assume, from some investigations (but few, I own), that they 

 do exist everywhere in rudiment, as, for instance, in Iberis. 



As, on the one hand, very crowded inflorescences become stunted, 

 especially in the internal part, so in the excessive development of bracts, 

 the flowers in their axils are frequently abortive, chiefly in the outer 

 parts of a very crowded inflorescence (sterile bract, bractea sterilis). 

 To this are referable the common calyx (calyx communis, anthodium, &c.) 

 of the Composite, the similar circle of leaves which close the mouth of 

 the Fig, the outer scales of the Grasses (gluma Juss., calyx Linn., lepi- 

 cena Rich., teg men Palisot, glumce valvce Link), which either both or 

 one, sometimes the upper sometimes the lower have no flower in the 

 axil. Link acutely remarks on this point, that the Grasses have a 

 compound flower in respect of this, or more correctly an inflorescence, 

 similar to that of the Composites. We may apply to all these combina- 

 tions of bracts the general term blossom-envelope (Bluthenhiille), which 

 will then comprehend the involucrum of the Umbelliferce, the calyx 

 communis of the Composite, the cupula of the Cupuliferce, the invo- 

 lucrum of the Euphorbiacece, the gluma of the Grasses, &c. ; and, by 

 joining a clearer and more accurate distinctive name to the definition, at 

 once free us from a vast waste of terminology. 



From these considerations the general law may be expressed, that, 

 excepting as a terminal flower, the solitary flower always, and only, 

 appears ^in the axil of a leaf, or in the place corresponding to that axil. 



As in the branch-buds the distinction was required between principal 

 and secondary bud, so also here ; of which condition I believe no one has 

 hitherto thought : yet such secondary buds distinctly exist, for instance, 

 in the inflorescence of Apocynum androscsmifolium, hypericifolium, &c. 

 It is very difficult to say whether the peculiar condition of the in- 

 florescence, e. g. in Penstemon, belongs here, where, instead of one 

 (terminal) flower in the bifurcating division of the peduncle, there are 

 two, of which one with a longer pedicel projects beyond the other. 

 In like manner, the position of the flower of Helianthemum variabile 

 seems to me to be diverted toward one side of the pedicel, because it 

 originates from a secondary bud, while the terminal bud does not become 

 developed. 



Another peculiar condition exists in the bract of the Limes. The 

 axillary bud, formed every year to persist through the winter, has two 

 opposite lateral bud-scales quite to the outer side, one of which remains 

 in this condition. But a bud is formed in the axil of the other, develop- 

 ing the same year, and becoming blended with the bud-scale, which 

 grows up with it, forms the peduncle, and thus exhibits a very decided 

 example of prolepsis, which outstrips the homologous organic parts of 

 the plant at least by three years. An actual blending of the pedicel 

 with the bracts, like this, occurs also in the male flowers of many Cu- 

 puliferce, e. g. in Corylus, and in the flowers of Saururus. 



Finally, it must be noticed, that it very often happens, especially in 

 the peduncle, that the substance expands to a greater extent in the points 

 which are not the bases of the parts seated on it, and swells up around 

 those bases. Thus the parts attached upon it appear to have their base 

 sunk in little pits (e. g. in the receptaculum foveolatum of the Composite), 

 or completely imbedded in little cavities in the uniform mass, as in the 

 female flowers of Dorstenia. This condition occurs naturally most 

 frequently on very thick peduncles, developed into a woody or fleshy 

 substance. 



