127 



also for such qualities as resistance to extremes of temperature, 

 to lack of moisture in dry or semi-arid regions, to resistance 

 towards specific diseases, and even for the more esthetic qualities 

 of flavor or color. The old hit or miss methods of the improve- 

 ment of strains by empirical rules of selection is passing away, 

 and more and more scientific methods, based on the latest results 

 of investigations of heredity and variation, are being employed. 

 Passing over the older methods I will take up two very different 

 types of plant breeding, both modern : one the strictly scientific, 

 the other the intuitive. 



The first method we owe largely to Nilsson, who introduced 

 it at an experiment station in Sweden in connection with the cul- 

 tivation of various cereal crops. It may be said that previous 

 to his advent the older methods had been tried and abandoned 

 as a failure. With his knowledge of what had been published 

 about heredity and variation, Nilsson, after some preliminary 

 experiments, arrived at the conclusion that no new, pure, or con- 

 stant strains of wheat could be obtained unless the fruit of a single 

 ear was bred separately, and thus he established what is known 

 as the principle of breeding from the single ear and not from 

 assorted lots of seed taken from many individuals. This breeding 

 he continued, picking out any chance favorable ear which he could 

 find, until he obtained many thousands of different forms owing 

 to this multiplicity of strains mixed in the ordinary wheat. Of 

 course some turned out to be mere bastard strains and only the 

 ones which continued to breed true to character were kept. These 

 constituted the new agricultural varieties — in reality elementary 

 species and mutants — which, after severe tests had proved them 

 suitable, were raised in marketable quantity for seed. The amount 

 of work involved was enormous, the mere bookkeeping of the 

 accurate pedigree record with notes on the life history of each 

 form and its progeny was in itself no small matter. Besides the 

 principle of single-ear breeding, Nilsson also established the fact 

 that but a single selection alone is necessary to fix a new strain, 

 provided the progeny of the chosen ear are carefully guarded from 

 admixture with other forms. All this seems absurdly simple, and 

 it is simple, so much so that it is quite possible of application by a 



