19 



I am by no means sure that it would not be more consonant to analogy to 

 consider the parts of their flower divided upon a binary plan ; thus under- 

 standing the outer series of the supposed petals as calyx, and the inner only 

 as petals; while the parts now called sepals are perhaps more analogous to 

 bracteae ; an idea which their arrangement, and the constant tendency of the 

 outer series to become saccate at the base, which is not uncommon in the calyx 

 of Cruciferse, but never happens, as far as I know, in their petals, would seem 

 to confirm. Of this, some further evidence may be found in the stamens. 

 These are combined in two parcels, one of which is opposite each of the divi- 

 sions of the outer series, and consists of one perfect 2-celled anther in the 

 middle and two lateral 1 -celled ones: now, supposing the lateral 1-celled 

 anthers of each parcel to belong to a common stamen, the filament of which 

 is split by the separation of the two parcels, an hypothesis to which I do not 

 think any objection can be entertained, we shall find that the number of 

 stamens of Fumariacese is 4, one of which is before each of the divisions of 

 the flower ; an arrangement which is precisely what we should expect to find 

 in a normal flower consisting of 2 sepals and 2 petals, and the reverse of what 

 ought to occur if the divisions of the flower were really all petals, as has been 

 hitherto believed. 



The economy of the sexual organs of Fumariacese is remarkable. The 

 stamens are in two parcels, the anthers of which are a little higher than the 

 stigma ; the two middle ones of these anthers are turned outwards, and do 

 not appear to be capable of communicating their pollen to the stigma; 

 the four lateral ones are also naturally turned outwards, but by a twist of 

 their filament their face is presented to the stigma. They are all held 

 firmly together by the cohesion of the tops of the flower, which, never 

 unclosing, offer no apparent means of the pollen being disturbed, so as 

 to be shed upon the stigmatic surface. To remedy this inconvenience, the 

 stigma is furnished with two blunt horns, one of which is inserted between 

 and under the cells of the anthers of each parcel, so that without any alter- 

 ation of position on the part of either organ, the mere contraction of the 

 valves of the anthers is sufficient to shed the pollen upon that spot where it 

 is required to perform the office of fecundation. 



This order offers every gradation, from monospermous to polyspermous 

 fruit, and between indehiscence, as in Fumaria itself, and dehiscence, as in 

 Corydalis. 



Geography. Their principal range is in the temperate latitudes of the 

 northern hemisphere, where they inhabit thickets and w^aste places. Two 

 are found at the Cape of Good Hope. 



Properties. The character of Fumariacese is, to be scentless, a little 

 bitter, in no degree milky, and to act as diaphoretics and aperients. Dec. 

 The root of Fumaria cava and Corydalis tuberosa has been found to contain 

 a peculiar alkali called Corydalin. Turner, 653. 



Examples. Fumaria, Diclytra, Corydalis. 



XI. CAPPARIDE^. The Caper Tribe. 



CAPPARiDEiE, Juss. Gen. 242. (1789) ; Ann. Mus. 18. 474. (1811) ,• Dec. Prodr. 1. 237. 



(1824). 



Diagnosis. Polypetalous dicotyledons, with hypogynous stamens, con- 

 crete carpella, a 1-celled pedicellate ovaritmi, narrow simple parietal placentae, 

 a continuous enlarged disk, and reniform seeds. 



