191-1] Bmndegee: Variation in Ocnoflicra ovata 45 



The ovaries spring from the axils of the densely crowded leaves 

 that protect and conceal them as wel as the almost membranous 

 capsules. The slender, long-produced empty portion of the ovary 

 which has been confounded with the calyx-tube of other species is 

 not tubular "with the st>'le united with the tube" as is often said, 

 l)iit is a solid 1)()(1\-. through the center of which the placentae and 

 even the ovules can be traced upward for some distance above the 

 fruiting portion. The length of this sterile portion of the ovary is 

 often very great, reaching in the outer early flower 18 cm. (figs. 17, 18). 

 The later "tubes" are successively shorter, coming down to 3 cm. The 

 structure is of nearly uniform diameter between the throat above and 

 the seed-bearing part below. This long sterile portion was described 

 twenty-five years ago, but the note appears to have been overlooked.^ 



The capsules, the fruiting parts, are of very irregular shape, some- 

 times broad at base, sessile, conical (fig. 22), sometimes narrow or 

 linear and occasionally more or less pedicellate (fig. 23). The walls 

 are thin, pale and conformed to the turgid, rather large (2 mm.) seeds 

 (fig. 29) which are purplish at first, becoming yellowish- white by the 

 development of the shaggy-pubescent outer coating. Later the 

 seeds become of a dark color, if remaining long in the ground. The 

 inner coat is purple, becoming nearly black. The embryo (fig. 30) is 

 oval, with no visible distinction in outline between cotyledons and 

 radicle. 



The seeds are ripened usually by the middle of June, the leaves 

 and sterile capsules withering and breaking away so that scarcely a 

 trace of the plant is to be seen, a circumstance which may account in 

 part for the general ignorance concerning its fruiting habits. 



The capsules are not regularly dehiscent, but split more or less 

 along the sutures, especially if removed from the ground. The seeds 

 are disseminated by cultivation, by rooting or gnawing animals or by 

 torrential rains. Consiciering their unusual shagginess it is difficult 

 to understand why, when mentioned, they have always been described 

 as "smooth." 



Sex-variation shows itself not in dimorphism but in great difference 

 in size of stigmas and anthers, connected by copious intergrades. The 

 stigma varies from linear-clavate to globular of greatly increased 

 diameter (figs. 0, 17, 24, 25, 26), either borne on a style equaling the 

 stamens in length, or nearly sessile (fig. 26), often m(n*e or le.ss 

 unilateral. The anthers are frequently effete, sometimes atrophied. t)ut 



2 Proc. Cal. Acad. Sci., ser. 2, i, p. 25."). 



