15 



many of the more highly developed plants of the order there are four placen- 

 tae opposed to each other by pairs, and forming the inner edge of each side of 

 the replum, which itself terminates in the stigmas. To this replum is attached 

 on each side a deciduous plate, or valve as it is called, which has no vascular 

 connection with either the replum, stigmata, or pedicel. In consequence of 

 this singular arrangement of parts, it has been found extremely difficult to un- 

 derstand the exact nature of the Cruciferous pist ilium, or to reduce it to the rules 

 which are known to govern the formation of other compound pistilla. 



" According to Mr. Brown, and, after him, to M. Decandolle, the pistillum of 

 Cruciferae is to be understood to consist of two confluent ovaria, united by their 

 placenta?, two lamella? from each of which project into the cavity of the ova- 

 rium, and, meeting in the centre, coalesce and form the septum. This, how- 

 ever, does not remove the difficulty of the stigmata being opposite the placen- 

 ta?, instead of alternate with them. I am not aware that any explanation of 

 this point has been published by Mr. Brown ; but M. Decandolle {Thtorie EU- 

 mtntaire, ed. 1. p. 133) accounts for it thus. He assumes that there are seve- 

 ral kinds of simple pistilla, some of which are not to be found in an isolated 

 state, but the possible existence of which he conceives to be demonstrated by 

 certain compound pistilla, that cannot be reduced to their simplest state without 

 the admision of such a position. Among these supposititious simple pistilla, 

 is one called the Siliqnelle, ' which is formed originally of three pieces, the two 

 lateral producing ovula on their inner surface, and the outer (intermediate) 

 one bearing no ovula ; pistilla of this description make up the fruit of Nym- 

 phaeaceae, Papaveraceae, and Cruciferae. When two pistilla of this kind are 

 united by the external edge of their lateral pieces, they form those fruits which 

 are said to have intervalvular placenta? ; each of these double placenta? is 

 elongated into a style or stigma, simple in appearance, but in reality formed by 

 two half styles grown together.' 



" To maintain this theory, it is necessary to assume, in the first place, the ex- 

 istence of a simple pistillum, of structure not only entirely hypothetical, but op- 

 posed to all we know of vegetable organization ; and, in the next place, that 

 the stigmata of the order, although so simple in appearance that no trace what- 

 ever of composition can be found in them, are, nevertheless, each composed of 

 two half stigmata in a state of cohesion. 



" To us this explanation has always been unsatisfactory. It was difficult to 

 believe that rules of structure, well ascertained to be uniform in other plants 

 should be deviated from in Cruciferae, especially when the irregularity obser- 

 vable in the arrangement of other parts of their flower was taken into account. 

 It always appeared more probable, that the anomalous nature of the pistillum 

 depended upon some irregularity corresponding to that of the stamens, than 

 upon peculiar laws appertaining to Cruciferae alone. 



" This seems to be at length proved by Eschscholtzia, the fruit of which is 

 so similar to that of Cruciferae, that the uniformity of the laws under which 

 they are both formed is not likely to be disputed. In this plant the pistillum is 

 unilocular, with four stigmata, of which the two opposite ones are smaller than 

 the two others. Upon opening this pistillum we find that there are two parietal 

 placentae corresponding with the smaller stigmata, and that there are no pla- 

 centa? opposite the larger stigmata ; in other words, that it is formed of four 

 simple pistilla, two of which are opposite and ovuliferous, with their placenta? 

 in the usual place, alternating with themselves ; and two nearly abortive, des- 

 titute of placenta?, consequently not ovuliferous, and so nearly suppressed by 

 the superior energy of their two neighbours, that their existence would have 

 been unknown but for the sigmata which indicate their presence. This is one 

 way of understanding Eschscholtzia ; but as the ovula are not inserted in the 

 placenta? in a double row, but rather confusedly arranged in several rows, it 



