HEREDITY AND EVOLUTION. 527 



proof. Those who find satisfaction in demonstrations of the obvious 

 may amply indulge themselves by starting various sorts of some an- 

 nual, say French poppy, in a garden, letting them run to seed, and 

 noticing in a few years how many of the finer sorts are represented; 

 or by sowing an equal number of seeds taken from several varieties of 

 carnation, lettuce or auricula, and seeing in what proportions the fine 

 kinds survive in competition with the common. 



Selection is a true phenomenon; but its function is to select, not 

 to create. Many a white-edged poppy may have germinated and per- 

 ished before Mr. Wilks saved the individual which in a few generations 

 gave rise to the shirleys. Many a black Amphidasys betularia may 

 have emerged before, some sixty years ago, in the urban conditions of 

 Manchester the black var. doubledayaria found its chance, soon prac- 

 tically superseding the type in its place of origin, extending itself over 

 England, and reappearing even in Belgium and Germany. 



Darwin gave us sound teaching when he compared man's selective 

 operations with those of nature. Yet how many who are ready to 

 expound nature's methods have been at the pains to see how man really 

 proceeds? To the domesticated form our fashions are what environ- 

 mental exigency is to the wild. For years the conventional Chinese 

 primrose threw sporadic plants of the loose-growing stellata variety, 

 promptly extirpated because repugnant to mid- Victorian primness. 

 But when taste, as we say, revived, the graceful star primula was saved 

 by Messrs. Sutton, and a stock raised which is now of the highest fash- 

 ion. I dare assert that few botanists meeting P. stellata in nature 

 would hesitate to declare it a good species. This and the shirleys pre- 

 cisely illustrate the procedure of the raiser of novelties. His opera- 

 tions start from a definite beginning. As in the case of P. stellata, he 

 may notice a mutational form thrown off perfect from the start, or, 

 as in the shirleys, what catches his attention may be the first indication 

 of that flaw which if allowed to extend will split the type into a host 

 of new varieties each with its own peculiarities and physiological con- 

 stitution. 



Let any one who doubts this try what he can do by selection with- 

 out such a definite beginning. Let him try from a pure strain of 

 black and white rats to raise a white one by breeding from the whitest, 

 or a black one by choosing the blackest. Let him try to raise a dwarf 

 (' Cupid ') sweet pea from a tall race by choosing the shortest, or a 

 crested fowl by choosing the birds with most feathers on their heads. 

 To formulate such suggestions is to expose their foolishness. 



The creature is beheld to be very good after, not before its creation. 

 Our domesticated races are sometimes represented as so many incar- 

 nations of the breeder's prophetic fancy. But except in recombina- 

 tions of preexisting characters — now a comprehensible process — and in 

 such intensifications and such finishing touches as involve variations 



