PHANEROGAMIA : FLOWERS. 323 



organs, namely, the style and stigma. Then, to create all possible 

 combinations, comes the interesting structure of Siphonodon celastri- 

 neus, described by Griffiths in the Calcutta Journal, where the carpels 

 form the cavity of the germen and the canal of the style ; while the 

 axial organ forms the spermophore, the conducting cellular tissue, and 

 the large umbrella-shaped stigma. 



All those structures grouped together under the name of discs in 

 the paragraphs are doubtless of the same nature ; the development 

 distinctly shows them to be flat or concave expansions of the internodes 

 entering into the flower ; sufficient analogies to which occur in the flat un- 

 doubted internodes of many Compositas (e. g., Helianthus, fig. 184.), in 

 the hollow ones of Ficus (fig. 190.), and so forth. The assertion of the 

 axial nature of all the structures mentioned is, in remarkable manner, 

 inductively proved by the following considerations ; that one leaf 

 actually springs from another, grows forth from it, contradicts the 

 definition, and is therefore impossible. The assumption of a calyx- 

 tube, therefore, of blended foliar organs in the Onagracece, Rosacece, 

 &c., from which free petals should spring, was altogether an uncon- 

 sidered and absurd fiction. An axial structure must necessarily have 

 been presupposed here ; and then it was completely unwarrantable, and 

 clumsy prolixity, to superadd an imagined coherence of sepals into a 

 tube, and adherence of this tube to the discoid receptacle, for instance, 

 in Rosa (fig. 191.), Geum (fig. 185.), &c. (while not the most distant 

 attempt to demonstrate such a condition was thought of). In these 

 cases (the so-called Calyciflorce) the sepals are no more grown together 

 than the petals, and they stand quite free upon the border of the in- 

 ferior (fig. 185.), surrounding (figs. 188. 191.), or superior disc. But 

 the course of development speaks with equal decision in favour of 

 the view propounded in the paragraphs, since often the structures, and 

 especially the inferior germen, exist complete, or almost so, and at least 

 clearly perceptible, before even a trace of the leaves which grow out 

 of them is evident. (See, in regard to this, the development of the flower 

 of Canna exigua, Plate III., with the explanation.) 



In all discs, the sudden change of texture, and commonly also a dis- 

 tinctly projecting rim, show the boundary of the disc, and the point of 

 connection between it and the foliar organs standing upon it ; and by 

 these marks, in the generality of Calyctflorce, the calyx is characterised 

 as composed of free foliar organs "not coherent by growth. I recall 

 it to recollection here that, in these forms of the axis, the middle of the 

 lower or outer surface corresponds to the lowest portions of the axis, 

 the lower or outer and the upper or inner surfaces, together, to the sides of 

 the axis, and the central point of the upper or inner surface to the apex 

 of the axis. On this axis the individual foliar organs, or circles of 

 leaves, may be very variously arranged, as is seen by-comparing together 

 figs. 188. 191. 193. It is not necessary that they should be all attached 

 around in one zone, for in the discoid axis a condition is possible 

 which, as in the elongation of the axis, separates the individual foliar 

 organs or circles from one another, and corresponds to one or more 

 developed internodes. It is usual for all internodes of the flower to 

 concur in the formation of the discs where this exists ; but there are 

 cases where this is not so, and where the different internodes assume 

 very different forms. Thus the Rosacece are divided pretty strictly into 

 two groups, according as the disc is quite flat or concave (Rosa, San- 

 yuisorbca, figs. 188. 191.), or the internodes between calyx and stamens 



v 2 



