308 



THE CROSSING OF FLOWERS. 



and hidden by the petals as soon as the neighbouring stigma begins to mature, 

 so that they are no longer able to shed their pollen. The consequence is that the 

 stigmas can only be pollinated with foreign pollen, which is of course the same thing 

 as saying that only cross-fertilization can occur in these hermaphrodite flowers. In 

 the hermaphrodite flowers of the Spiderworts (Tradescantia craskula, Virginica, 

 &c.), the anthers dehisce a considerable time before the stigmas mature. When 

 the flower first opens, therefore, pollen only can be removed. But as soon as the 

 stigmas become capable of fertilization the stamens roll up in a spiral, and soon 

 afterwards the perianth withers and forms a moist, pulpy mass, quite covering the 

 anthers on their rolled-up filaments. The style still pi'ojects stiSly from the flower 

 and the stigmas remain capable of fertilization the whole of the following day. 

 Small flies and other insects with short probosces now visit these flowers to suck 

 up the juice of the pulpy petals, and at the same time the stigma is pollinated with 



Fig. 292. — Dichogamy in Saxifraga rotundifolia. 



A portion of the inflorescence with flowers at different stages: that to the right still young, in the middle older. 2 Longi- 

 tudinal section through a single flower with folded stigmas and one stamen shedding its poUen. Another stamen (to left 

 of pistil) has lost its anther, and four others have anthers which have not yet dehisced, s The same flower at a later stage 

 of development, with mature stigmas. 1 nat. size; 2 and s x 4-5. 



pollen which they have brought from distant flowers, it being impossible to obtain 

 that of the neighbouring anthers. It is an odd fact that some of the flowers of a 

 Tradescantia plant, all of which opened simultaneously in the morning, will be 

 already closed the same evening, whilst others will remain open the whole of the 

 following day. It would seem that in those flowers which remain open the succu- 

 lent hairs of the staminal filaments are devoured by flies, thus is the pollen obtained 

 which is to be taken to the stigmas of the flowers whose anthers are hidden under 

 the pulpy perianth. A peculiar process is observed in the flowers of Telephium 

 Imperati, a native of Southern Europe, belonging to the Caryophyllaceee. Here 

 the anthers open first, but, as soon as the stigmas mature, the anthers — even if they 

 have not as yet discharged all their pollen — are covered over by the petals, so that 

 only pollen from other younger flowers can reach the ripe stigmas. 



By these contrivances the same result is obtained in hermaplirodite flowers as 

 by the separation of the two kinds of sexual organs on diflferent plants, or on 

 different flowers of the same plant. In all cases it seems to be the separation of 

 the two kinds of sexual organs within the limits of the same species which is aimed 

 at. The separation of the two kinds of sexual organs by the non-simultaneous 

 maturation of tlie pollen and of the stigmas in any one species is just as effective in 



