from the iiniial Sintcliire of Seeds. ■'^<^- 14.> 



miprpgnatcd ; it is still possible to conceive a case in which a 

 ripe seed may l)e con^siclere(l as truly naked while retaTnino; its at- 

 tachment to the parent plant; and this not subsequent to germi- 

 nation, but even preceding the formation of the embryo. For if 

 •we suppose, as the immediate etfect of impregnation, a swellinn^ 

 of the ovuluni without a corresponding enlargement of the ova- 

 rium, the consequence will obviously be a premature rupture of 

 the ovarium, and the production of a seed provided with its pro- 

 per integuments only. 



I am not aware that such an economy has hitherto been de- 

 scribed ; I have observed it, however, in several plants belonging to 

 very different families, and of essentially different structures. 



The first of these is Leoiitice thalictroides of Linneus, Caulophtjl- 

 lum thalictroides of Miehaux, who has founded his new genus on 

 a difference of fruit, the nature of which he has entirely misunder- 

 stood. It is remarkable that its real structure should have escaped 

 so accurate an observer as M. Kichard, through whose hands it 

 is generally understood Michaux's work passed previous to its 

 publication ; but the fact may at least serve to show how entirely 

 unexpected such an economy must have been even to that excel- 

 lent carpologist. 



My observations were made in the summer of 181?, on a plant 

 of Leontice thalictroides, which flowered and ripened fruit in the 

 royal gardens at Kew. An examination of the unimpregnated 

 ovarium proved it to be in every respect of the same structure 

 with that of the other species of Leontice; and essentially the 

 same with the whole order of Berberides, to which this genus be-* 

 longs. A careful inspection of the fruit, in different states, 

 proved also that the " Drupa stipitata" of Miehaux is in reality 

 a naked seed, that in a very early stage had burst its pericarpivmi; 



VOL. MI. V the 



