KECENT OBSEKYATIONS ON THE EEHTILISATION OF PLANTS. 345 
rotundifolia), a flower of which is represented by fig. 6 , a 
representing its ordinary utmost state of expansion, while at 
b are the organs of reproduction enclosed in it. The pistil 
ends in five club-shaped stigmas covered with receptive papillse; 
while the 8-12 stamens, one only of which is represented in 
our drawing, curve in and out among them by their flexible 
filaments, so as to bring the pollen into actual contact with the 
stigmas. 
The most singular, however, of these special contrivances 
for self-fertilisation are the peculiarly-shaped “ cleistogamous ” 
flowers, as they have been termed — which occur in many plants 
belonging to widely-separated natural orders, either intermixed 
with the ordinary conspicuous flowers or appearing at a different 
time of the year — with respect to which very little has been 
written in English botanical works. Among the natural orders 
in which these flowers have been found are Violacese, Cistacese, 
Oxalidege, Balsamineae, Polygalacese, Caryophyllacese, Malpi- 
ghiacese, Leguminosse, Campanulacese, Convolvulacese, Acan- 
thacege, Labiatse, and one order of Endogens, Commelynaceaa.* 
The two species of Impatiens or Touch-me-not, which grow wild 
in this country — 1 . Noli-me-tangere, native in Westmoreland and 
some other rocky and woody parts, and I. fulva., a North Ame- 
rican plant fully naturalised by the banks of the Wey and other 
parts of Surrey, as well as the smaller /. parviflora, now also 
rapidly becoming completely naturalised in the neighbourhood 
of London — have closed, imperfect, self-fertilised flowers inter- 
mixed with the showy yellow ones. They are far more 
numerous than the conspicuous flowers, much smaller, and 
easily recognised even in the bud.f The calyx is quite regular, 
not presenting the “ spur ” of the open flowers, always remains 
perfectly closed, and is pushed off at the extremity of the seed- 
vessel in the form of a little brown cap. The petals are 
entirely absent. The stamens are of an altogether different 
shape to those of the larger flowers, and contain but a very 
small quantity of pollen, which, however, is amply sufficient 
for the fertilisation of the ovules, the full number of seeds 
appearing to be always produced. 
The most easily observed instances are, however, in the case 
of our common wild violets, the sweet violet ( Viola odorata), 
or the various forms of the dog violet ( V. canina). The 
existence of these flowers in Viola was known as long ago as 
the time of Linnaeus, who, in his “ Praelectiones Botanicae,” 
says that the flowers of Viola mirahilis produced in the spring 
* See Von Mobl in Botanische Zeitung,” 1863, pp. 311 et seq. 
t See a full description and plate in the “ Journal of the Linnean Society,’^ 
Yol. xiii. p. 147, May 29, 1872. 
