352 AUTOGAMY. 



must also be noted that the inflection of the style only takes place should the stigma 

 have not previously been supplied with extraneous pollen. If there has been cross- 

 pollination at the commencement of flowering the inflection either does not occur at 

 all or is so trifling as to be of no significance. 



The transference of pollen to the stigma by means of a bending down of the 

 style is observed in various species of Tricyrtes, Marina, CEnothera, and EpiLobiwrn, 

 in several Rhinanthaceae, Caryophyllaceae, and Ranunculaceae, and in most of the 

 Malvaceae. The flowers of Tricyrtes pilosa, of which figures are given above, are 

 protogynous. Each of the three styles is bent down, and has a bifurcated extremity, 

 so that it looks not unlike a claw. The stigmatic tissue is situated at the end of 

 the claw, and is brushed by insects on their alighting to suck honey from the saccate 

 bases of the perianth-segments. The free extremities of the filaments curve down 

 in semicircles, and the anthers are suspended under the claws of the style (fig. 299 l ). 

 At the time when the anthers have their pollen exposed they are so situated as to 

 stand in the way of insects coming in quest of honey. Without the assistance of 

 insects there could be no transference of pollen to the stigma so long as stigmas and 

 anthers remained in the same relative positions, and the flower remained upright. 

 But the chance of insects not visiting the flower is provided for by a downward 

 bending of the claws of the style, which continues until the stigmatic tissue at their 

 extremities comes into direct contact with the pollen-coated anthers (see fig. 299 2 ). 



The process above described takes place in the course of a week in Tricyrtes 

 pilosa, but in Marina Persica, one of the Dipsaceae (see figs. 299 3> 4 > 6 ), it is all 

 accomplished within a few hours. The difference between the times at which 

 stigmas and anthers respectively attain maturity in Morina is scarcely half an 

 hour, but even this short interval suffices to render cross -fertilization possible 

 during the first stage of flowering, whilst in the second stage autogamy obtains. All 

 the species of the genus Morina including Morina Persica, the type here selected 

 for illustration unfold their flowers at dusk. As soon as the corolla-limb expands 

 the thick pulvinate stigma becomes visible in the middle of the flower just above 

 the entrance to the honey. The receptive tissue is on the upper surface of the 

 stigma. The two anthers are stationed behind the stigma, and when insects insert 

 their probosces into the long honey-filled tube of the corolla they are certain to 

 effect cross-fertilization, provided they have previously visited flowers at a some- 

 what later stage of development. In the case of other plants whose flowers open 

 in the morning it would be scarcely likely that insects should alight immediately 

 after the opening of the passage to the honey, but the flowers of Morina are 

 adapted to crepuscular and nocturnal moths, which only have two or three hours of 

 darkness in which to get the honey, and must, therefore, make great haste and 

 employ the whole of the time if they are not to fail in their quest. As a matter of 

 fact the moths in question leave their haunts within a quarter of an hour of the 

 time when the flowers of Morina open, and one may be sure that wherever Sphin- 

 gidae and Noctuae with probosces 3 or 4 centimetres in length abound, one or more 

 will come flying to suck the honey as soon as the floral receptacle becomes acces- 



