206 



THE FLOWER. 



378. The obvious relationship of Cruciferae to Fumariaceae, 

 their agreement in the rare peculiarity of having the two carpels 



side by side instead of fore and afc 

 (median), and the characteristic 

 anomaly which the androecium pre- 

 sents (i. e. the tetradynamy), would 

 give reason to expect that its prob- 

 lems might be solved by chorisis. 

 Indeed, the doctrine was applied to 

 this, long before its application to 

 the other order. Beginning at the 

 centre (Fig. 395, &c.), the pistil is 

 of two carpels, right and left ; alter- 

 nate with these is a pair of stamens 

 on the side next the axis, matched 

 by another pair on the opposite 

 side of the pistil, the four longer 

 and interior stamens ; alternate 

 with these, and lower in insertion 

 a single stamen on each side ; next> 

 four petals, of somewhat various overlapping in aestivation, 

 which essentially alternate with the two single stamens and the 

 two pairs ; lastly, four sepals, alternating with the four petals as 

 a whole, the anterior and posterior overlapping the lateral ones 

 in the bud: Now the median (i. e. the anterior and posterior) 

 pairs of stamens occasionally have their contiguous filaments 

 conjoined, as in Fig. 397. If this were at all constant, the 

 inference would undoubtedly be that the case is one of chorisis, 

 and that the flower as to its essential organs is dimerous. This 

 is apparently the best explanation to be given. It assumes that 

 the chorisis is normally complete in the androecium of Cruciferae, 

 instead of incomplete, as in Fumariaceae. 1 And this view is 

 confirmed by the fact that the median stamens are simple and 



307 



1 The hypothesis here adopted, as to the androeeium, is that of Steinheil 

 (1839), and of Eichler (in Flora, 1865, 1872, and Bliithend. ii. 200), replacing 

 that of Kunth, 1833, &c., employed in former editions. The rejected view 

 makes the flower 4-merous up to the pistil, and the stamens all of one circle, 

 alternating with the four petals, the median stamens (as in our view) doubled 

 by chorisis. Krause and Wretschko (cited as above by Eichler) would 

 have the floral circles 2-merous and 4-merous by turns ; the calyx of two 

 2-rnerous circles (which it plainly is); the corolla of one 4-merous circle 



FIG. 394. A cruciferous flower. 395. Diagram of such a flower, with position of 

 axis marked above it. 396. Tetradynamous stamens and the pistil. 397. A common 

 monstrosity of the same, two of the four inner stamens combined into a common 

 2-antheriferous body. 



