908 PHENOMENA OF SEXUAL REPRODUCTION. 



In connection with the tendency so clearly evidenced even among Cryptogams, 

 and still more among Phanerogams, to prevent self-fertilisation within the same 

 hermaphrodite group of sexual organs, it is a very remarkable fact that there are a 

 number of plants among Angiosperms which form two kinds of hermaphrodite 

 flowers, VIZ. large flowers which can generally be fertilised by the pollen of other 

 flowers, and small, more or less depauperated flowers, sometimes underground, which 

 never open \C lei's togamous Flowers], the pollen emitting its tubes immediately from 

 the anthers and thus fertilising the ovules. There occur therefore in these cases 

 difl"erent kinds of flowers on the same individual, one kind being adapted for cross-, 

 the other kind exclusively for self-fertilisation ^ This occurs, for example, in Oxalis 

 Aceiosella, where the small flowers are formed close to the ground when the larger 

 flowers have already ripened their fruit ; in Impatiens Noli-me-iangere, Lamium 

 amplexicaule, Specularia per/olmta, many species of Viola, as V. odoraia, elaiior, 

 canina, mirabilis, &c., Ruellia clandesltna, many Papilionaceae, as AmphicarpcEa, and 

 Voandzeia, Commelyna bengalensis, &c. When in these cases the large typically de- 

 veloped flowers are fertile, cross-fertilisation with other flowers of the same species 

 must happen occasionally in the course of generations, and the small depauperated 

 self-fertilised flowers then seem to be a subsidiary contrivance whose purpose is 

 altogether unknown. It is however remarkable, and apparently in contradiction to 

 the general rule, that the large normal flowers sometimes exhibit a tendency to infer- 

 tility (as in species of Viola) or are altogether unfruitful (as in Voandzeid), so that 

 reproduction depends in such cases mainly or entirely on the cleistogamous self- 

 fertilised flowers. But since there are many questions in connection with this subject 

 which find their solution in the foregoing facts, these rare exceptions cannot over- 

 throw the general law^. 



In other cases, as in most Fumariaceae, Canna indica, Salvia hirla, Linum usita- 

 lissimum, Draba verna, Brassica Rapa, Oxalis micrantha and sensiliva, the pollen 

 must also, according to Hildebrand, owing to the position of the sexual organs, fall 

 on the stigma in the same flower, and is potent ; but in such cases, since the flowers 

 are visited by insects, an occasional crossing with other flowers is not impossible. 

 Even among Orchidese, where we find the most wonderful contrivances to prevent 

 self-fertilisation, Darwin found an instance in Cephalanthera grandiflora in which the 

 pollen-tubes are emitted from the pollen-grains on to the stigma while the former 



Rye and other cereals, see Hildebrand in Gardener's Chronicle, March 15 and 22, and May 24, 1873 ; 

 also A, S. Wilson, Trans. Bot. Soc. Edin. XI. 506 and XII. 84. Flowers the pollination of which 

 is effected by the wind are termed anetnophilous, in contradistinction to the entomophilous, or those 

 pollinated by the agency of insects.] 



^ H. V. Mohl, Einige Beobachtungen iiber dimorphe Bliithen, Bot. Zeit. 1863, Nos. 42, 43. 

 [See also A. W. Bennett on the closed self-fertilised flowers oi Impatiens in Journ. Linn. Soc. 1872, 

 p. 147; ditto, Pop. Sci. Rev. 1873, p. 337. In Junais hifoniiis the pollen-tubes are emitted while 

 the pollen-grains are still enclosed in the anther, perforating the wall of the latter. Henslow, On 

 the Self-fertilisation of Plants, Trans. Linn. Soc, Series II, vol. I, 1879.] 



^ [Herrmann Miiller (Nature, vol. VIII. p. 433 et scq.) has pointed out the existence of another 

 kind of dimorphism, in which a species presents two different forms of flowers, one adapted to self- 

 fertilisation, smaller and less brightly-coloured, growing in situations where there are but few 

 insects, the other adapted to cross-fertilisation, larger and more brightly-coloured, growing where 

 insects abound. These two forms have occasionally been described as distinct varieties or even 

 species.] 



