PHANEROGAMIA : FLOWERS. 435 



pieces, from which two facts it follows that the subsequent separation of 

 the parts is wholly independent of the original composition. But as the 

 word blending has hitherto been applied without any meaning, according 

 to the arbitrary fictions of particular botanists, nothing opposed the 

 equally arbitrary twaddle about the conditions referred to in the para- 

 graph. The kind and manner of these separations has not the slightest 

 connection with the original composition of the germen out of individual 

 parts, carpels, &c., and every conclusion from the number of subse- 

 quent parts as to the number of original constituent portions merely 

 exhibits the total unacquaintance of the concluder with the nature of 

 plants, and of this process in particular. Here, as so frequently in 

 the vegetable organism, in the originally homogeneous cellular tissue, 

 which, where actual blending has taken place, becomes so closely con- 

 nected that the boundary is soon wholly obliterated, layers of a very 

 different kind of cells are developed, which exhibit great diversity, 

 partly in the consistence of the substance forming their walls, partly 

 in the relatively advanced condition of the deposition of thickening 

 matter upon them. Similar cells are generally more firmly adherent 

 to each other than to cells of a dissimilar character, and thence it happens 

 that the different layers separate so readily, as, for instance, in the 

 succulent part of the fruit of the Almond, Plum, Walnut, &c., from 

 the woody. In most cases, however, thin plates of very thin-walled 

 and speedily decaying tissue are formed for this purpose, and are lace- 

 rated by the slightest tension, and thus give rise to a solution of conti- 

 nuity. Even in the situations where originally distinct parts have been 

 blended, the separation seldom (or never ?) happens in such a way that 

 the blended parts become again loosed from one another, but so that the 

 cells are torn and destroyed ; and thus, even in these cases, the circum- 

 stances of the process are by no means arrived at and expressed, when 

 it is said that the valves are the original carpels ; it is here shown, 

 moreover, that all these solutions of continuity, in the entire plant, 

 full under one and the same law, that of the morphologically defined 

 dehiscence, which is wholly different from and independent of the mor- 

 phologically defined formation and connection of organs. 



I will particularly cite the application which has been made of that 

 incorrect view to the -Geraniacece and Umbelliferce. In these the fruit 

 separates into distinct parts from a stalk-like mass of cellular tissue, 

 remaining longest in connection with the summit of this, and, as it 

 were, suspended from it. By the favourite method of guessing, this 

 stalk is explained to be a prolongation of the floral axis, to which the 

 carpellary leaves are attached, and from which they again separate when 

 the fruit is mature. In the first place, it must be observed, that in 

 the Umbelliferce the whole germen is never formed of carpels, but of 

 the axis itself. In the Geraniacece, on the contrary, there are ori- 

 ginally five perfectly free carpels, having no trace of a prolongation of 

 the axis between them, which become blended, and subsequently de- 

 hisce in such a manner that an internal portion of each carpel remains 

 in the axis, while the external portion becomes gradually loosened from 

 below upwards. The internal portion contains a liber-bundle, with the 

 canal of the style. Now in the Umbelliferce two little liber-bundles are 

 found in the middle of the false septum of the germen, which remains, 

 with a portion of their enveloping cells, in the axis of the fruit, while 

 the two portions of the fruit in like manner tear away from them from 

 below upward. Sometimes those liber-bundles separate from each 



F F 2 



