This section is from the book "On British Wild Flowers Considered In Relation To Insects", by John Lubbock. Also available from Amazon: Nature Series On British Wild Flowers Considered In Relation To Insects.
This order contains four British genera; Geranium, Erodium, Oxalis, and Impatiens.
The genus Geranium possesses a peculiar interest in the history of the present subject, because, as Sprengel tells us, the hairs in the corolla of G. sylvaticum (See p. 1), attracted his attention, and led to the researches which are so well described in his interesting work.
The flowers of the species of Geranium differ considerably in size: the larger flowered species, such as G. sangiiineum, G.phceum, G.pratense (Figs. 40 and 42), and G. sylvaticum, being perennial, the smaller ones annual, or biennial. Geraniumpalustre, with which G. tratense, G. sylvaticum, and G. sanguineum, closely agree, is taken by Sprengel as a type of the large flowered species. The honey glands are five in number, situated near the base and at the outer side of the outer stamens; and are effectually protected by fringes of hairs arranged just above them, so as to prevent any rain from obtaining access to them.
The stamens are ten in number, of which one half are longer than the remainder: the pistil terminates in five lobes, the upper surfaces of which constitute the stigmas. The flower opens widely by day, hangs down, on the contrary, and partially closes at night. The petals are ornamented by purple lines, which serve as honey-guides, pointing to the honey glands. When the flower first opens (Fig. 41) the stigma is immature, and the five lobes are closely pressed together (b), so that the stigmatic surfaces are not exposed. Nor do they separate (Fig. 42), or become susceptible of fertilisation, until after the anthers have all shed their pollen. The flower, in fact, passes through three distinct stages: first, the five outer stamens open, and shed their pollen; then the five inner ones; and lastly, after the pollen is all shed, the stigmatic surfaces expand and attain maturity. The flower therefore cannot fertilise itself.
On the other hand, in the smaller species of Geranium the stigmas come to maturity before the stamens have shed all their pollen; hence the visits of insects are not so necessary, and hence, probably, the smaller size of their flowers. (See ante p. 48.)
There is also another difference, to which I will call attention. G. Robertianum does not possess the fringes of hairs by which the honey is in G. sylvaticnm protected against the access of rain; on the contrary, the petals are entirely glabrous. This difference is apparently connected with the form of the flower, which is less open than is the case with G. sylvaticnm On the contrary, it forms a distinct tube, the entrance to which is sufficiently protected against rain by the stamens and pistil.
The smaller flowered species moreover offer remarkable differences among themselves. Thus G. molle and G.pusillum are at first sight very similar, and, as Bentham observes, are no doubt sometimes confused; yet they differ remarkably. When G. molle first opens, the pistil is immature, and the stigmatic surfaces are closely appressed. The outer anthers then begin to open one after the other, so that the flower is for some time merely male. Before, however, the first five anthers have completely shed their pollen, the stigmatic surfaces arrive at maturity and expand; so that, during the second period, the flower is both male and female.
In G. pusilhim, on the contrary, when the flower first opens, the stigmatic surfaces are mature and expanded, but the anthers, are not yet ripe; the flower consequently is merely female, and can only be fertilised by pollen from an older flower. Soon, however, the anthers of the five outer stamens ripen, and open, so that the flower is both male and female. Another remarkable difference is, that in G. pusillum, as in the allied genus Erodium, the five inner stamens produce no pollen.
Fig. 63. - Young flower of Tropaeolum major. Neither the anthers nor the stigma are yet mature.
In the genus Erodium, as in G. pusillum, the five inner stamens are rudimentary, and produce no pollen. The stamens ripen, however, before the stigma, though if the visits of insects be deferred, the flower is capable of self-fertilisation.
To this family also belongs Tropaeolum, the common Nasturtium of our gardens. Here the honey is contained in a long spur. The flower passes through three well marked stages (Figs. 63 - 65). When it first opens, as shown in Fig. 63, the anthers (a) are unripe, the pistil (p) is short and immature. Soon, however, one of the anthers ripens, opens, and turns up, as shown in Fig. 64 a, a, so as to stand directly in front of the opening to the tube; a humble bee, therefore, or other insect of similar size, visiting the flower for the sake of its honey, could not fail to rub some of the pollen off on to her breast. Shortly afterwards a second stamen ripens, and assumes the same position, with the same result, and the rest gradually follow. In flowers which I have watched, this process occupies from three to seven days, by which time the stamens have all come to maturity, after which the anthers drop off, and the filaments turn down as shown in Fig. 65, so as to be well out of the way. It is now the turn of the pistil, which in the meantime has elongated, and assumes the position which the stamens had successively occupied; the result of which is, that a bee which had previously visited a younger flower and dusted her breast with pollen could not fail to deposit some of the pollen on the stigma. It will be observed that the lines on the flower as usual point to the honey. The three lower petals bear a number of lanceolate processes, which, as Sprengel has pointed out, serve to prevent rain from obtaining access to the tube, and also perhaps in guiding the insects, so as to bring their breasts truly against the anther or stigma. The calyx, which from the position of the flower is almost as much exposed as the corolla, is of the same colour as the petals.
Fig. 64. - Flower of Tropaeolum major in the second stage. Some of the anthers are now mature, and stand upright in front of the entrance to the spur.
Fig. 65. - Flower of Tropaeolum major in the third stage. The anthers have all shed their pollen, and the pistil now occupies the position in front of the entrance to the spur.
Oxalis acetosella is one of the species which produces "cleistogamous" flowers (see ante p. 36). This was first, I believe, observed by Michelet (Bull. Soc. Bot. de France, 1860, p. 465).
Hildebrand has shown that of the foreign species of Oxalis some are dimorphous, some trimorphous, This is, however, not the case with either of our English species, and I will therefore postpone any remarks on this curious arrangement until we come to some of the English species in which it occurs.
Impatiens noli me tangere is proterandrous, and the larger flowers cannot fertilise themselves. This species, however, also produces cleistogamous flowers. (Mohl. " Bot. Zeit," 1863. Bennett, "Linn. Jour.," vol. xiii.) The seed capsules, when ripe, burst open if touched.