THE COMPOUND PISTIL. 



291 



pistils) were pressed together as they grew and consohdated more or 

 less completely into one. And in tliis, the most normal case, we have 

 as the result compound pistils 



549. With two or more Cells and Axile Placentffi. For it is evident 



that, if the contiguous parts of a whorl of three or more closed car- 

 pels cohere, the resulting compound ovary will have as many cavi- 

 ties, or cells, as there are carpels in its composition, and the placentas 

 (one in the inner angle of each carpel) will all be l)rought together 

 in the axis of the compound pistil. And the partitions, or Dissep- 

 iments, which divide the compound ovary into cells, manifestly 

 consist of the united contiguous portions of the 

 walls of the carpels. These necessarily are 

 composed of two layers, one belonging to each 

 carpel ; and in ripe pods they often split into 

 the two layers. True dissepiments must always 

 be equal in number to the carpels of which the 

 compound pistil is composed. 



550. In certain cases, indeed, there are addi- 

 tional partitions, or false dissepiments. These ai'e commonly projec- 



FI6. 496. Pistil of a Saxifrage, composed of two carpels or simple pistils united below, but 

 distinct above ; cut across both above and below. 



FIG. 497. Pistil of common St. Johnswort, of three united ovaries ; their styles distinct. 



FIG. 498. The same of another species of St. Johnswort (Hypericum prolificumj, the styles 

 also united into one, which, however, split apart in the fruit. 



FIG. 499. Pistil of Tradescantia or Spiderwort, even the three stigmas united into one. 

 The ovary in all cut across to show the internal structure. 



FIO. 500. Cross-section of a flower of Flax ; each of the five cells of the ovary partly divided 

 by an imperfect false partition from the back. 



500 



