PLANT-BREEDING METHODS 425 



The Svalof System. At the Swedish Institute for the Improvement of 

 Field Crops, Nilsson has worked out a very complete and efficient system 

 of plant breeding. Gradually, as increased appropriations of funds have 

 permitted expansion, a corps of experts has been employed, each in- 

 vestigator concentrating on one or two species, and thereby training 

 himself to distinguish all the different forms so as to judge of the relative 

 value of different combinations of characters. Furthermore a definite 

 course of procedure has been developed as a result of many years of 

 experience during which time marked success has been achieved in the 

 inprovement of Swedish field crops. To begin with the work consisted 

 mainly of variety testing and extensive effort at improvement through 

 mass selection. These methods still find a place in the routine work, but 

 they are of insignificant value as compared with the coordination of 

 intensive methods which makes the Institute's system a model which 

 institutions engaged in similar work may profitably follow. From 

 Nilsson's description we find that the Svalof system may be briefly out- 

 lined under three heads, viz., genotype selection, strain tests and hybridiza- 

 tion. In all this work the methods of pedigree culture are followed so 

 that the original source and performance record of each form grown at 

 the station or distributed for trial can be accurately stated. Genotype 

 selection in all plants except the self-sterile species is accomplished by 

 the pedigree method of testing the progeny of single individuals. In 

 wheat, barley, oats, peas and vetches, which were the first crops chosen 

 for improvement at Svalof, this means of course the isolation of pure 

 lines from the beginning, and the more recent work with rye, clover, 

 forage grasses, beets, etc., has determined that allogamous species are 

 composed of biotypes which are analogous to the pure lines of autogamous 

 species and which can be segregated from one another by continued 

 inbreeding, exactly as inbreeding in maize has been found to isolate 

 biotypes. Although inbreeding must be continued for several years 

 before these biotypes or strains acquire a satisfactory degree of purity 

 and stability, yet it has been shown already according to Nilsson that 

 this method can be used to bring about the same practical results as have 

 been secured in wheat and other autogamous plants. In connection 

 with this preliminary selection of promising forms the intensive study of 

 the specialists at Svalof has made each member skilful in detecting 

 different forms in the species with which he is working and in judging 

 the relative value of the characters displayed. Having separated from 

 the population some of the biotypes of which the "variety" is composed, 

 it next becomes necessary to subject all these strains to comparative 

 tests in order that the few superior forms may be discovered and pro- 

 pagated more extensively. This requires long years of careful work and 

 the overcoming of certain difficulties which will be discussed later. The 



