AUTOGAliy. 331 



the close of flowering. At first the plant seems to offer facilities for hj-bridization, 

 later for a crossing of different plants of the same species, and not until both these 

 proce.sses have failed, owing to lack of insect-visits, does it fall back on cross- 

 pollination between neighbouring flowers of the same plant. The processes and 

 contrivances in the flowers of Glandestina rectiflora, Bartsia aJ/pina and of some 

 other Khinanthaceae agree in the main with those of the Toothwort, so that there 

 is no need to describe them in detail. In conclusion, the Persian Crucvxnella. 

 dylosa, one of the Stellatae group of the Rubiace^, must be considered especially 

 with regard to the geitonogamy of its hermaphrodite flowers. This plant has 

 already been mentioned (pp. 265, 267), and it was stated that the tips of the corolla- 

 lobes formed a hollow cone which at a touch suddenly burst open, scattering the 

 pollen hidden beneath (fig. 272). If an insect is the cause of the disturbance it 

 becomes covered with the liberated pollen, and should it then visit other flowers of 

 this same CrvxianeUa it is not too much to say that the pollen it carries on its body 

 will be deposited on the stigmatic tissue at the thickened end of the slender styles 

 projecting far above the coroUa (see fig. 272 *). If the flower is not visited, however, 

 it will burst open and scatter its pollen spontaneously. The dust-hke pollen is then 

 diffused into the surrounding air, and so easily reaches the ripe stigmas of the 

 neighbouring flowers. 



AUTOGAMY. 



By the term Autogamy is understood the transference of pollen from the 

 stamens to the stigmas of the same flower, in other words — aelf-jpollination. Auto- 

 gamy can, therefore, only occur in hermaphrodite flowers, but it must not be 

 inferred that the only method of reproduction in hermaphrodite flowers is by auto- 

 gamy. In the last chapter we showed the error into which Linnaeus fell by making 

 this assumption, and referred to the almost interminable list of contrivances to 

 bring about cross-pollination in hermaphrodite flowers between different species, 

 between different individuals of the same species, and between flowers on the same 

 individual. This important result of modem investigation has led not only to a 

 correction of the views of the celebrated Swedish Botanist, but to the belief that 

 autogamy is avoided in the vegetable world. Darwin even held the opinion that 

 there must be something injurious in autogamy to account for the number of 

 obstacles placed in its way. This statement, though commonly made in text-books 

 as embodying a law of nature, is, however, not a correct expression of the facts 

 observed. It is true that cross-pollination appears to be the primary object aimed 

 at, but it is not true that autogamy is avoided. If cross-pollination takes place 

 there is naturally no necessity for subsequent autogamy, but if cross-pollination 

 fails autogamy assumes an importance of its own, and the contrivances which have 

 been observed to bring about autogamy are no less numerous than those which 

 favour cross-pollination. That flowers should be adapted at diflerent times to two 

 such diverse purposes as cross- and self-pollination is one of the marvels of floral 

 construction. 



