THE COMPOUND PISTIL. 



293 



507 



506 



fectly, we have only to imagine two, three, or any number of pistil- 

 leaves (like Fig. 490), arranged in a circle, to unite with one another 

 by their contiguous edges, either witliout any intro- 

 flexion or infolding at all (Fig. 504), or at least 

 Avithout their infolded edges having reached the cen- 

 tre and united there (Fig. 505, 506). The combina- 

 tion is accordingly much like that by which petals 

 unite to form a monopetalous corolla, only the edges 

 of the pistil-leaves are always turned in, where they 

 bear the ovules. Such an ovary may 

 Avell be compared with the valvate un- 

 opened calyx of Clematis, the margins 

 of the sepals more or less turned in- 

 wards (Fig. 445). Every gi-adation is 

 found between axile and parietal pla- 

 centation, especially in the St. Johns- 

 wort family (Fig. 508, 509) and in the Gourd family. 



555. An ovary with parietal placentte is necessarily one-celled; 

 except it be divided by an anomalous partition, such as is found in 



Cruciferous plants, and in the Trum- 

 pet Creeper. 



556. It will be seen that parietal 

 placentae are necessarily double, like 

 the placenta of a simple ovary, or of 

 each carpel of a compound several- 

 celled ovary ; but with this difference, 

 that in these the two portions belong to the two margins of the 

 s^me carpel ; while in parietal placentae they are formed from the 

 coalescent margins of two adjacent carpels. 



557. The number of carpels of which a compound ovary consists 

 is indicated by the number of true dissepiments when these exist ; 

 or by the number of placenta, when these are parietal ; or by the 

 number of styles or stigmas, when these are not wholly united into 

 one body. Thus a simple pistil has a single cell, a single placenta. 



503 



509 



FIG. 506. Plan of a one-celled ovary with three parietal placentje, cut across below ; the 

 upper part showing the top of the three leaves it is com|)ose(l of, approaching, but not united. 



FIG. 507. Ovary of Uelianthemum Canadense, cut across, showing the ovules on three 

 parietal placentae. 



FIG. 508. Transverse section of the ovary of Hypericum graveolens ; the tliree large 

 placenta? meeting in the centre, but not cohering. 509. Similar section of a ripe pod of the 

 game ; the placentae now evidently parietal. 



25* 



