187 



teriorating and so lessening the value of harvests. It soon 

 became manifest that the German method of amehorating a given 

 variety of grain by multiple selection was a failure and that any 

 success was very exceptional, in fact a mere matter of chance. 

 Nilsson brought to the solution of the problem a full knowledge 

 of the laws of variability and inheritance and early in his work 

 made the remarkable discovery that each of the so-called 

 varieties of agricultural crops was in reality made up of a 

 great series of distinct strains or elementary species. The number 

 of these elementary species is surprising ; for example, in the 

 common peas 500 distinct forms have been separated and each 

 variety of the cereals is composed of several hundred forms. 

 More remarkable still is the degree of variation among these ele- 

 mentary species, which far exceeded all the expectations based 

 upon their divergencies when first selected. In fact it was found 

 that when these elementary forms were isolated and bred that they 

 produced offspring so divergent in morphological characters and 

 qualities as to meet practically any demand that the farmers re- 

 quired. Some were suitable for light soils, others for heavy ; some 

 were early and others late in ripening. They differed in stiffness 

 of culms, length of ears, size and number of grains, etc. This 

 discovery revolutionizes the common method of plant breeding. 

 All success depends upon the initial selection of a pure strain. 

 Manifestly the old multiple selection must result in the association 

 of several strains — some being poor or indifferent — with the re- 

 sult that the offspring will be a mixture and fall short of a high 

 staLndard. These early experiments demonstrated that the plant 

 develops according to its inherent nature and that it can not be 

 made to conform to a desired pattern. So the old idea of amelior- 

 ating a crop by repeated selections must be abandoned. A single 

 initial selection is the important point. 



Not less important from a practical standpoint was the work ot 

 Nilsson in studying the relation or correlation between the botan- 

 ical characters of the elementary species and their agricultural 

 qualities. No sooner was it discovered that the elementary spe- 

 cies are so divergent as to meet almost any demand of the 

 breeder than the need of a system was felt whereby the desired 



