THE CARPEL 



237 



The major differences in interpretation of the morphology of free 

 central placentation relate to the nature of the placenta. Are the ovules 

 borne on a prolongation of the receptacle, that is, are they cauline in 

 nature— borne on the stem, not on the carpels, which merely enclose 

 them? Does the placenta represent the ventral strips of all the carpels, 

 freed from the dorsal section and connate, with the ovules borne as in 

 typical free carpels? Is the central placenta a compound structure, the 



Fig. 89. Cross-sectional diagrams of ovary structure showing free-central placenta- 

 tion. A to D, Lychnis, showing ontogeny: A, B, early stages, Literal walls present; 

 C, D, later stages, lateral walls degenerated. E, Lysimachia, and F, Steironema, 

 showing lateral walls absent, placentae suspended from central column of car- 

 pel margins. Dotted lines indicate limit of carpels. (A, B, after Lister; C, D, after 

 Van Tieghem; E, F, after Douglas; F, modified. ) 



tip of tlie receptacle coated by the fused ventral margins? Ontogeny 

 and comparative and anatomical studies all demonstrate clearly that 

 this kind of placentation is derived from axile placentation, that the 

 central placenta consists typically of the fused placentae and ventral 

 margins of the united carpels, freed from the dorsal section by loss of 

 the lateral walls which connected them. This loss is ontogenetic in sev- 

 eral famihes (Fig. 89C, D)— Caryophyllaceae, Primulaceae— and readily 

 recognized, but phylogenetic, in others. In some genera, the receptacle 

 extends upward into the placenta, as it does between carpel margins in 

 some types of axile placentation (Fig. 88). If this receptacle tip is 



