266 Flowers and their Pedigrees. 



alterations by which these structural modifications 

 were brought about must have been verj- slow ; and 

 they must have been produced by the constant fertili- 

 sation of such arums as best retained the visiting 

 flies, and the dying out of such as did not well retain 

 them. Last of all, the berries grew large and red 

 under the influence of animal selection, those berries 

 which attracted birds succeeding in producing new 

 plants, while those which did not so attract them died 

 out unsuccessful!}-. And at the same time the ovarj' 

 came to contain only one seed, instead of three cells 

 with many seeds, because one seed under the new and 

 improved method of dispersion went as far as five or 

 ten would have gone under the old and wasteful casual 

 method. Thus at last what had been a bunch of 

 distinct coloured lilies grew to be a cuckoo-pint with 

 an inclosing hood and a spike of minute central in- 

 conspicuous flowers. 



