146 VEGETABLE ORGANOGRAPHY. 



at the base or apex of the fruit ; in this case one is very 

 frequently abortive, and in this manner the monosper- 

 mous carpels of the Compositae and Dipsaceae are formed. 

 But when we find, now and then, one of the Composites 

 with the fruit having two ovules, we see that they are 

 both ascending, and it is very likely that, if we should 

 find two in the Dipsaceae, they would be both pendent. 



Whenever a carpel contains several seeds, they are 

 free and not united to the inner surface of it ; but when 

 it only contains a single seed, this one is sometimes free, 

 as in the utricles of the Amaranthaceas, or united by its 

 entire surface with the carpellary leaf, as in the fruit of 

 the Gramineae ; and then the carpel is so confounded 

 with the proper integument of the seed, that it does not 

 seem to exist ; it is in this case that seeds were called 

 naked, but there are none really devoid of the pericarp. 

 In fact, the style necessarily takes its origin from the 

 pericarp and not from the seed, and consequently, every 

 organ from which we see, at the period of flowering, a 

 style or stigma arise, is a true pericarp, whatever be its 

 appearance. 



Seeds may appear naked from three causes : — either 

 by the intimate union of the seed with the carpel, as in 

 the Gramineae : — or because, as in certain species of 

 Leontice, or in Slateria, the seed, growing rapidly, 

 breaks the carpellary leaf and is found exposed : — or 

 because, as in Reseda, the carpellary leaves not being 

 completely folded upon themselves, leave their ex- 

 tremity open, and consequently the seeds naked. But 

 we know that neither of these causes answers exactly to 

 what was intended by the term of naked seeds, and that 

 the pericarp exists or always has existed. 



The manner in which the seeds of a carpel ripen and 

 are dispersed, accords with the principles above declared. 

 In all carpels with a long placenta, i. e. when the seeds 



