56 



STRUCTURAL BOTANY. 



[46, 47. 



fronts; as in St. Johnswort (129) and Lily (171). In 

 this case, he may prove the following propositions. 

 1st. The compound ovary will have as many cells as 

 carpels. 2d. The partitions between the cells will be 

 double, and alternate with the stigmas. 3d. A parti- 

 tion dividing the cell of a single carpel must be a 

 false one; as occurs in Flax (136). 4th. The PLA- 

 CENTA, as well as the ventral suture, will be axial. 



133. Again: the carpels may each be opened and 

 conjoined by their edges, as are the petals of a gamo- 

 petalous corolla. So it is in the ovary of Violet (137) 

 and Rock-rose (139). In this case, 1st. There will be 

 no partition (unless a false one, as in the Crucifers), 

 and but one cell ; 2d. The Placentce will be parietal, 

 i. e., on the wall of the cell (paries, a wall). 



140 



135 



133, Samolus Valernndi, section of flower showing the free axial placenta. 134, Ovary of Scrophula- 

 riaceae. 135, Ovary of Tulip. 136, Cross-section of ovary of Flax, 5-celled,' falsely 10-celled. 137, Ovary 

 of Violet, 1-celled. 138, Ovary of Fuchsia, 4-celled. 139, Ovary of Rock-rose, 1-celled, 5-carpelled. 

 140, Gentianaceae, 2-valved, 1-celled. 



134. Between the two conditions of axial (or central) and parietal placenta, 

 we find all degrees of transition, as illustrated in the different species of St. 

 Johnswort and in Poppy, where the inflected margins of the carpels carry the 

 placentae inward, well-nigh to the axis. Moreover, the placentae are not al- 

 ways mere marginal lines, but often wide spaces covering large portions of 

 the walls of the cell, as in Poppy and Water-lily ; in other cases, as Datura 

 (168), they become large and fleshy, nearly filling the cell. 



135. A free axial placenta, without partitions, occurs 

 in some compound one-celled ovaries, as in the Pink 



