FEBTILIZATION. 639 



ii. Anthers near the stigma. 



* No fruit formed without impregnation by insects : Corydalis 



cava. 



(It would scarcely have been deemed credible that self-fertilization was 

 impossible in such an instance as Corydalis cava, where the anthers are 

 closely appressed to the stigma, and in which self-impregnation appears 

 inevitable. In his experiments, however, Hildebrand discovered that 

 when this plant was secured from the visits of insects, and also when the 

 pollen was artificially applied to the stigma of the same flower, no fruit 

 was set. To obtain perfect fruit he found it necessary to impregnate the 

 stigmas of one plant with pollen from another.) 



* * Fruit formed as a result of self -impregnation, but impregnation 



by insects not excluded. 



(Instances of undoubted self-fertilization of individual flowers are known 

 in the genus Fumaria, in Salvia hirsuta, Linum usitatissimum, Cephalanthera 

 (jrandiflora, Ophrys apifera, &c.; but the number and quality of seeds 

 borne is less than where cross-impregnation is effected.) 



/3. Flowers not opening (cleistogamous]. 



Self-impregnation alone results, every foreign impregnation being ex- 

 cluded ; but these plants all have other flowers which open and are thus 

 exposed to the possibility of extraneous impregnation. 



The general conclusions that may be drawn from the above facts are 

 thus given by Hildebrand: 1. The arrangements in the majority of 

 flowers are such that no self-impregnation takes ' place, but a trans- 

 port of the pollen from flower to flower is accomplished instead. 2. Insects 

 are necessary in most cases for the conveyance of the pollen. 3. When 

 the access of a flower's own pollen is prevented, it necessarily follows that 

 self -impregnation is impossible. 4. In those cases where self -impregna- 

 tion is possible, or even unavoidable, the possibility of foreign impregna- 

 tion is for the most part not excluded. 5. In these cases insects are 

 active, and accomplish the impregnation of the flowers. 6. There is 

 probably no flowering plant to which access of foreign pollen, at least 

 to a portion of its flowers, is possible, and continued self -impregnation 

 alone possible ; therefore no flowering plant which furnishes a proof 

 against the general law which negatives self-fertilization. 7. By experi- 

 ments it has been found that where, by accident or design, the pollen of 

 a flower falls on the stigma of the same flower, fertilization either does 

 not follow, or, when it does occur, the quantity of seed is less than where 

 foreign pollen is employed. 8. A gradual transition may be traced, start- 

 ing from those cases where self-impregnation is utterly impossible, to 

 those where it is possible and evident, but not to the exclusion of the 

 possibility of a foreign impregnation of the flowers. 9. The sexual 

 relations and mode of fructification do not invariably tend towards 

 the morphological affinities of the flowers. In some isolated families 

 the sexual conditions of all members are alike ; in other families, again, 

 and even genera, they are essentially different. The sexual relations, 

 therefore, have not developed with equal pace and in the same way as the 



