SELF-FERTILIZATION OF PLANTS. 363 
lateness of the season, just as the Sloe is similarly, but normally, self-fertilizing because 
it blossoms so early in the summer; conversely, that conspicuously flowered plants 
with much honey are correlated with a high temperature and brilliant sunlight. 
It may not, perhaps, be too wide a generalization to conclude that all flowers with 
very numerous stamens, as in Ranunculacee, Rosaceæ, Myrtaceae, &c., are thus adapted 
for both intercrossing and self-fertilization, the former process being effected by the 
outer and first dehiscing anthers, and before the stigma is mature, the latter by the 
inner and later ones, which dehisce at the same time that the stigma is ready to receive 
their pollen. 
Lyraraceæ.—The genus Lythrum gives us another instance of a transition from a 
high state of differentiation in its adaptation to insect-fertilization in its trimorphic 
species to the degraded state seen in L. hyssopifolia, which is undoubtedly self-fertilizing. 
The intermediate condition is seen in L. thymifolia, which is only dimorphic. The 
stamens are reduced to six, which are homologous with the longest stamens, namely, 
those opposite the sepals. As observed in the Caryophyllacee, it is the rule that the 
whorl of stamens opposite the petals should disappear first. With L. hyssopifolia a still 
further point is reached. It is no longer heterostyled. As in the preceding, the six 
shorter stamens are often suppressed. Mr. Darwin says (‘ Forms of Flowers,’ p. 166) :— 
* The stigma is included within the calyx, and stands in the midst of the anthers, and 
would generally be fertilised by them ; but as the stigma and anthers are upturned, 
and as, according to Vaucher, there is a passage left in the upper side of the flower to the 
nectary, there can hardly be a doubt that the flowers are visited by insects." Whether 
this be the case or no, it seems to prove that this form is a degraded one, which has 
retained the relative positions of the stigmas and anthers necessary in L. Salicaria, but 
has acquired self-fertilizing powers as well. Its habit, too, agrees with homogamy, in 
being an annual and solitary and not like the other mentioned species, which are social. 
5 Nesea verticillata. This, Mr. Darwin states, is trimorphic, while, according to Fritz 
Müller, a species of this genus in St: Catharina, in Southern Brazil, is homostyled. 
Cuphea purpurea “ was highly fertile with its own pollen when artificially aided, but 
sterile when insects were excluded” (* Forms of Flowers, p.168). It is therefore a case 
of morphological, and not physiological, sterility. 
Peplis Portula. I have not the slightest doubt but that this plant i is invariably self- 
fertilizing ; for all the features one is familiar with in Lythrum are entirely gone. The 
petals are reduced to a minute size, or else are wanting. The six stamens opposite the 
sepals are alone present. They incurve over the pistil, as is so thoroughly characteristic 
of self-fertilizing plants, and the capsule sets an abundance of seed, as is usually the case 
with such. Lastly, it is a small inconspicuous weed-like annual. 
ONAGRACE &. —JAEpilobium is an instructive genus from the point of view under con- 
sideration. E. angustifolium was first observed by Sprengel in 17 90* to be proteran- 
* This is quoted: on the authority of Sir John Lubbock (7. c. p. 27), and is certainly the case. Hence Mr. M. C. 
Cooke’s description in his ‘Manual of Structural Botany,’ as quoted by Prof. Dyer in ‘Journal of Botany,’ vol. ix. 
p. 22, must be wrong ; for it is described as being specially adapted to self-fertilization, ! 
*. 
