THE CARPEL 227 



supplies of the lost carpels. Cimicifuga and Delphinium show well the 

 origin of solitary carpels. 



The position of the solitary carpel on the receptacle has been dis- 

 cussed in relation to the natiue of the carpel. Evidence that the carpel 

 is a leaflike, lateral organ is seen in the more or less clearly lateral posi- 

 tion in many taxa, but, in other taxa, the carpel appears to be median 

 and terminal. A median position has been considered evidence that the 

 carpel is cauline, not appendicular, in nature, that it is a continuation 

 of the stem. But ontogeny shows that the solitary carpel in the Legumi- 

 nosae is a lateral structure; the carpel primordium arises on the side of 

 the floral apical meristem and is moved to a "terminal" position by 

 differential growth. Vascular traces to carpels are derived laterally from 

 the receptacular stele and give evidence of the morphologically lateral 

 position of carpels that appear terminal, especially in flowers in which 

 bundles of the receptacular stele continue beyond the levels of depar- 

 ture of the traces. In unisexual flowers, a single carpel may be the 

 only floral appendage and may seem to be a continuation of the 

 pedicel — Euphorbia, Heliconia (Fig. 36), Nojas — and, therefore, cauline, 

 but a typical carpellary vascular supply and dorsiventral form show its 

 appendicular nature. 



Ontogeny of the Carpel 



The primitive carpel arises on the floral meristem as a more or less 

 crescent-shaped primordium, which soon becomes broadened by lateral 

 extension. Apical and marginal meristems appear early and increase its 

 length and width (Fig. 87). The presence of marginal meristems in the 

 carpel was recognized only in the early twentieth century. The ap- 

 parent absence of marginal meristems had earlier been considered evi- 

 dence that the carpel was unlike the leaf. Differential growth brings 

 about an upturning and bringing together of the lateral "wings," with, 

 ultimately, their appression and more or less complete histological 

 union. Comparative studies of the general method of closure of the 

 carpel seem not to have been made. The bringing together of the 

 laminar wings has been loosely described both as an upfolding and as 

 an inrolling. The carpel closed by a simple upfolding is termed con- 

 duplicate; that by inrolling, involute. Union of the contacting margins 

 has been commonly described as "by the margins." But "margin" is a 

 poor descriptive term here, because it may mean either the edges of 

 the blade or marginal strips (of either surface) near the edges. 



Earlier descriptions of carpel closure as an upfolding or inrolling 

 appear to be generally accurate, though little attention has been given 

 to details of contact from the standpoint of position of the marginal 

 meristems. Histological studies of the development of some primitive 



