STRUCTURE OF FRUIT. 145 



also when they are reduced so as to occupy only the 

 base or apex; for these are still the same two organs, 

 but much shorter than ordinarily. Since the placentas 

 of a carpel are necessarily double, and each of them has, 

 in the regular state, an equal right to bear seeds, it 

 results: — 1st, that the natural number of seeds in a 

 carpel ought to be always even when there is no 

 abortion ; but they are rarely placed exactly at an 

 equal height along each suture ; on the contrary, they 

 are situated alternately : this disposition is very evident 

 in legumes and long follicles ; but when the carpel is so 

 short as only to present one ovule on each placenta, 

 there are presented some cases which deserve notice : — 



1st. The two ovules may arise, as in long carpels, one 

 above the other, at so great a distance that both arrive 

 at maturity : this is what happens in the legumes of the 

 dispermous Leguminosas, and then the two seeds are 

 clearly horizontal. 



2d. These two alternate ovules are sometimes so close, 

 that one of them becomes abortive, the other only 

 arriving at maturity ; in this case, it happens either that 

 it is the upper one which is abortive, and then the 

 lower one finding more space towards the higher part of 

 the carpel, takes an erect position, or it is the lower 

 one which is abortive, and then the superior one finding 

 more space towards the base of the carpel, takes a 

 pendent position. It is very probable that it is to this 

 double cause that we must refer the diversity of direction 

 of the seed of the monospermous Ranunculaceae, which 

 are either ascending or pendent, whilst those of the 

 polyspermous ones are horizontal ; if one were to find a 

 Ranunculus or Clematis, the carpel of which presented 

 the two ovules arrived at maturity, they would be either 

 horizontal, or one ascending and the other pendent. 



3d. The two ovules may be very near together either 



VOL. II. L 



