^46 FUNCTIONS OF 



totally subverted by every flower with separated organs, 

 whose stamens could circulate nothing to germens on 

 a different branch or root; a difficulty which the judi- 

 cious Tournefort perceived, and was candid enough 

 to allow. 



Both the conjectures just mentioned vanish before 

 one luminous experiment of Linneeus, of all others 

 the most easy to repeat and to understand. He re- 

 moved the anthers from a flower of Glaucium phoeni' 

 ceimiy Engl. Bot. t. 1433, stripping off the rest of 

 that day's blossoms. Another morning he repeated 

 the same practice, only sprinkling the stigma of that 

 blossom, which he had last deprived of its own sta- 

 mens, with the pollen from another. The flower first 

 mutilated produced no fruit, but the second afforded 

 very perfect seed. " My design," says Linnaeus, 

 *' was to prevent any one in future from believing that 

 the removal of the anthers from a flower was in itself 

 capable of rendering the germen abortive." 



The usual proportion and situation of stamens with 

 respect to pistils is well w^orthy of notice. The for- 

 mer are generally shortest in drooping flowers, longest 

 in erect ones. The barren blossoms stand above the 

 fertile ones in Carejr, Coix, A?'um, &c., that the pollen 

 may fall on the stigmas. This is the more remark- 

 able, as the usual order of Nature seems in such 

 plants, as well indeed as in compound and even um- 

 belliferous flowers, to be reversed, for the pistils are 



