AUTOGAMY BY BENDING OF STAM1.\>. 337 



the approach to the floral receptacle open. Dehisc* m-r has already taken place in 

 the case of the anthers of the stamens inserted in front of the sepals, ami th.-ir 

 pollen is available for cross-fertilization by means of insects, but not for aut.- 

 gamy, owing to the fact that the receptive tissue of the style in the same ilower 

 is still inaccessible. (2) The anthers of the stamens inserted in front of the 

 sepals drop off, or else their filaments become reflexed and are exserted beyond 

 the periphery of the flower. The styles move asunder and arrange themselves 

 like the spokes of a wheel in the middle of the flower, where they are liable to 

 be covered with pollen brought by insects from other blossoms. The anthers 

 of the stamens inserted in front of the petals are still closed. (3) Owing to a 

 growth of the erect filaments of the last-mentioned stamens, their anthers are 

 brought to the same height as and in direct contact with the spreading stigmas. 

 Dehiscence ensues, and the liberated pollen is deposited on the receptive stigmatic 

 tissue. In annual caryophyllaceous plants e.g. in Silene conica the whole pro- 

 cess takes place in the course of a single day, whereas in the perennial Dianthv* 

 glacialis it occupies five or six days, or, if the weather is bad, from seven to nine 

 days. 



One of the commonest contrivances for effecting autogamy is the following. 

 Anthers and stigmas stand at the same height, though, owing to the position and 

 direction of the filaments, the anthers are so far from the stigma that no transfer- 

 ence of the adhesive pollen to it can take place. At the proper moment, however, 

 the straight and rigid filiform filaments perform certain special movements with the 

 object of conveying pollen from the anthers to the stigma in the same flower. The 

 filaments incline themselves towards the centre of the flower, bringing the anthers 

 into contact with the stigma there situated and pressing the pollen issuing from 

 their loculi on to the receptive tissue. In some plants belonging to this cate 

 the displacement of the stamens, which is like the motion of the hands of a ck 

 preceded by an elongation of the filaments, and in this respect the plants in quest 

 form a transition from those previously described, in the flowers of which auto 

 is due to the growth of the filaments. As instances of these transitional forms i 

 be mentioned Azalea procumbens, Draba aizoides, Hapl&phyllum Bvberst*,, tl 

 numerous Saxifrages comprised in the groups of Aizoonia and TridactylitoB, 

 more particularly many Alsine. and other Caryophyllacea, The Saxrf rag . exhibit 

 a number of individual peculiarities into which we cannot enter m any detad 

 must content ourselves with describing two spec.es aa represen^ve, 



VOL. II. 



