2 1 Fruits and Seeds. 



to leave their seeds on the surface of the soil, but 

 actually sow them in the ground. 



Thus in the Subterranean Clover, one of our 

 rarer kinds, only a few of the florets become per- 

 fect flowers, the others form a rigid pointed head 

 which at first is turned upwards, and as their ends 

 are close together, constitute a sort of spike. At 

 first, I say, the flower-heads point upwards like 

 those of other clovers, but as soon as the florets 

 are fertilised, the flower-stalks bend over and grow- 

 downwards, forcing the flower-head into the ground, 

 an operation much facilitated by the peculiar con- 

 struction and arrangement of the imperfect florets. 

 The florets are, as Darwin has shown, no mere 

 passive instruments. So soon as the flower-head is 

 in the ground they begin, commencing from the 

 outside, to bend themselves towards the peduncle, 

 the result of which of course is to drag the flower- 

 head further and further into the ground. In most 

 clovers each floret produces a little pod. This 

 would in the present species be useless ; many 

 young plants growing in one place would jostle 

 and starve one another. Hence we see another 

 obvious advantage in the fact that only a few 

 florets perfect their seeds. 



2. I have already alluded to our Cardamines, 

 the pods of which open elastically and throw their 

 seeds some distance. A Brazilian species (fig. 81), 

 besides the usual long pods (fig. 81, a a), produces 



