266 



THE FLOWER. 



of a congenital development of organs in union which, in the 

 development of a vegetative shoot, would lie leaves. This case 

 is represented by the combination of open carpellaiy leaves, as 

 the preceding one is by that of closed ones. As the edges 

 of the leaves must needs be turned in, to bear the ovules, 

 a compound ovary with parietal placentation may be likened 

 to the unopened calyx of a Clematis, as shown in Fig. iVii;. 

 257. Every gradation is found between axilc and parietal 

 placentation. Sometimes the placent;e are strict!}- on the pari- 

 etes or wall (Fig. 543, 5-17) ; sometimes borne inwards on 

 incomplete dissepiments (Fig. 5-1 <S) ; and sometimes they are 

 brought firmly together in the axis, as in Fig. 544, though sepa- 

 rable, and indeed separated in the fruiting stage. 



49G. A compound ovary with parietal placenta 1 is necessarily 

 one-celled (unilocvlar) ; except it be divided by an anomalous 

 partition, such as is found in Cruciferae (Fig. 3'J5) and in many 

 Bignoniacese. 



497. Normal placenta? are necessarily double : when parietal, 

 the two halves belong to different leaves ; when axile, to the same 

 leaf. These two halves may diverge or be widely separated, 



sometimes even at their origin, as in Aphyllon and some other 

 Orobanchaceae, in which a dicarpellary ovary has four almost 

 equidistant placenta 1 ; or in such cases the placenta 1 may be 

 regarded as infra-marginal instead of marginal. 



I'.t.s. The placentae of a two-several-celled ovary, such as in 

 Fig. 536, 537, &c., may be described in the plural number, 

 being one in each carpel ; or when consolidated into a central 

 column, and well covered with ovules, they may lie .-aid to form 

 one (compound) placenta. Then when the dissepiments early 

 disappear, or are aliortivc from the first, the result is a compound 

 o\ary of this class, 



I'.i'.i. With one Cell and Free Central Placenta. In Caryo- 

 phyllacea' (Fig. 5l'.t. ;>;>()) ;1 nd Portnlacacere. this evidently 

 results from the obliteration of the dissepiments (as many as 

 there are styles or stigmas), vestiges of which may be sometimes 



FIG. f>4f> l>i;iL.'r:iMi (u'ri>und-pl:iiO to illustrate froc i-i-ninil i>l.ioentation produced 

 by abortion of dissepiments. 547. Same of strict parietal placentation. 548. Same with 

 the placenta carried inward on imperfect dissepiments. 



