in.] THE FERTILISATION OF FLOWERS. 45 



flowers, rotundifolia has small ones. In the former 

 the stamens and stigmas are so arranged that self- 

 fertilisation is impossible ; in the latter they are so 

 placed that self-fertilisation cannot be avoided. A 

 similar instance occurs among the species of Willow- 

 herbs (Epilobiiun}. 



Among the various species of Wild Geranium we 

 have some with large flowers and some with small. 

 Of the large-flowered species we 

 will take Geranium pratense as the 

 type. Its flowers are erect and 

 opened wide by day ; at night 

 they hang down partially closed. 

 When the flower first opens the 

 pistil is immature. It is divided FlG 5I> 



into five stigmatic lobes, and when 

 immature these lobes have the stigmatic surfaces in 

 contact, as shown in fig. $\a. When the flower opens 

 the ten stamens also are immature and lie flat on the 

 petals. Five of them become matured and raise 

 themselves parallel with the pistil, shed their pollen 

 and retire to their former position. They are succeeded 

 by the other five, and when these have retired the 

 stigmas unfold as in fig. 51^. From this it will be 

 seen the flower cannot fertilise itself. But in the 

 smaller species, of which the well-known Herb Robert 

 (G. Robertianuni) is an example, the stigmas are 

 mature before all the pollen is shed, so that if cross- 

 fertilisation does not take place self-fertilisation does. 

 Here the size of the flowers is evidently due to the 

 visits of insects, and no doubt Sir John Lubbock is 

 right when he remarks: "It would seem that, as a 



