192 THE STRUCTURE OF FLOWERS. 



flowers as these, energy is especially directed into the 

 development of the corolla and androecium ; the former 

 being large, and the latter supplied with much and often 

 highly differentiated pollen. All this means the consumption 

 of so much more nutriment ; and, as the chief amount of 

 floral energy is thus directed first into the androecium, then 

 into the corolla — which often attains a far greater size than the 

 other organs, — consequently these tw,o whorls tend to draw 

 a large amount of nourishment to themselves. In conse- 

 quence of this, the pistil has, temporarily at least, to 

 suffer ; so that its growth is for a time delayed, and it 

 does not mature as early as the stamens, which had, moreover, 

 a considerable start in the race to maturity. Hence the 

 result is that the stamens are often mature and even shed all 

 their pollen long before the stigoaas are prepared to receive it. 



This, then, accounts for protandry being almost invariably 

 the rule in the case of relatively conspicuous flowers.* 



If flowers have two or more whorls or many series of 

 stamens, as have many genera of Canjophyllece, GeraniacecB^ 

 BananculacecB, and RosacecB, then the pistil may arrive at 

 maturity between the periods of different series, or con- 

 temporaneously with some of them ; so that, while the flower 

 is protandrous with regard to the first stamens which mature, 

 it is homogamous with others, and thus self-fertilisation can 

 be readily secured if the flower fail to be crossed. 



It may be here observed, though the fact will be dwelt 

 upon again, that by far the greater majority of flowers, 

 conspicuous or not, retain this provision for self-fertilisation ; 

 and that those flowers which normally cannot possibly 

 fertilise themselves are in a very small minority. 



* There are a few protogynons flowers, it is true, which are more 

 or less conspicuous, but these exceptional cases have their own inter- 

 pretations, which will be considered later on (see Chap. XXII.). 



