496 M. Ad. Brongniart on Vegetable Morphology. 



mucli in form and dimensions, by their longer support and nar- 

 row partition from those of Brassica ; but on opening them it was 

 seen that no true membranous partition existed, the thickened 

 borders of the carpels approached each other, and were in more 

 or less complete contact in their whole extent, or soldered together 

 only in part. The margins of the two different carpels were, on 

 the contrary, very intimately imited together in their whole ex- 

 tent, in most cases. 



The margins of these carpels gave origin to foliaceous lobes 

 occupying the position of ovules, reflected upon the interior of the 

 carpels, continuous with them by their base, disposed nearly in 

 the same plane, and divided into two or three acute teeth, so as 

 closely to resemble the margin of a pinnatifid leaf. Each of these 

 lobes was traversed by a little nerve, and its subdivisions by se- 

 condary nerves. The connexion of these little folioles together 

 at their base, their position in the same plane parallel to the axis 

 of the siliqua, showed plainly that they were not small distinct 

 leaves, but portions of a single lobed leaf. The juxtaposition of 

 this lobed margin with the similarly lobed margins of the other 

 carpellary leaf, the union even of the longitudinal vascular bundles 

 of the two into a single median bundle, produced the appearance 

 of a single pinnatifid leaf applied to the interior, within the suture 

 of the carpellary leaves, so that one might have imagined that the 

 pistil was formed of four leaves in the form of a cross, two form- 

 ing the valves or walls of the ovary and the other two the pla- 

 centse ; but the complete examination of the monstrosity rendered 

 this supposition unlikely to be true. 



Besides the parts already indicated, there existed always in 

 these pistils two little short cylindrical branches terminating in 

 tubercles or papilla3 representing young rudimentary leaves. These 

 originated in the axil of the carpellary leaves, and were only their 

 axillary buds elongated in the form of a slender axis. Also, from 

 the centre of the ovary, between the two carpels and from the 

 disunited base of the partition, a rather larger cylindrical axis 

 often arose, bearing on its summit little leaves united into a ca- 

 pitulum ; this was evidently the prolongation of the principal axis 

 of the flower. Thus all the modifications that could present them- 

 selves on a branch bearing two opposite leaves were found in the 

 interior of this pistil, without one of them taking part in the for- 

 mation of a placenta. In the few carpellary leaves which, in a 

 state of transition to free expanded leaves, began to separate from 

 each other towards the top, the carpels united at bottom still pre- 

 sented their ovuliform lobes, which then appeared to depend more 

 completely upon the carpellary leaves, and did not tend at all to 

 form a second pair independent of them. It is remarkable, that 

 in proportion as the carpellary leaves more completely assumed 



