Honckenya peploides affords another illustration of the sexual arrangements in the flower being altered as it would seem by climatal conditions. Thus, in the United States, according to Professor Asa Gray, the flowers are frequently hermaphrodite, while in this country they are usually sub-diœcious.[197]
Treviranus[198] says that the flowers of Hippuris and Callitriche are apt to be hermaphrodite in summer, but female only at a later period.
For further remarks on this subject, see sections relating to suppression of stamens and pistils.
Change from unisexuality to hermaphroditism.—This occurrence depends on one of two causes, either organs are developed (stamens or pistils as the case may be), which are habitually absent in the particular flower; or some of the stamens may be more or less completely converted into or replaced by pistils, or vice versâ.
The first condition is the opposite of suppression; it is, as it were, a restoration of symmetry, and might be included under the head of regular peloria, inasmuch as certain organs which habitually undergo suppression at a certain stage in their development, by exception, go on growing, and produce a perfect, instead of an imperfect flower. In teratological records it is not always stated clearly to which of the two above-named causes the unusual hermaphroditism belongs, though it is generally easy to ascertain this point. Very many, perhaps all, diclinous flowers may, under certain conditions, become perfect, at least structurally. I have myself seen hermaphrodite flowers in Cucurbita,[199] Mercurialis, Cannabis, Zea Mays, and Aucuba japonica, as well as in many Restiaceæ, notably Cannamois virgata and Lepyrodia hermaphrodita. Spinacia oleracea, Rhodiola rosea, Cachrys taurica, and Empetrum nigrum are also occasionally hermaphrodite.
Gubler[200] alludes to a similar occurrence in Pistacia Lentiscus, wherein, however, he adds that there was a deficiency of pollen in the flowers.
Schnizlein[201] observed hermaphrodite flowers in the beech, Fagus sylvatica, the ovaries being smaller than usual, and the stamens epigynous.
Baillon[202] enumerates the following Euphorbiaceæ as having exceptionally produced hermaphrodite flowers, Crozophora tinctoria, Suregada sp., Phyllanthus longifolius, Breynia sp., Philyra brasiliensis, Ricinus communis, Conceveiba macrophylla, Cluytia semperflorens, Wall, non Roxb. Mercurialis annua and Cleistanthus polystachyus.
In some of these cases the hermaphroditism is due to the development of anthers on the usually barren staminodes, though, in other cases, the stamens would seem to be separate, independent formations, as they do not occupy the same relative position that the ordinary stamens would do if developed.[203]