;SMay. 1908.] Improvement of Cereals by Selectifl)! and Crossing. 285 



The essential point in these pedigree-cultures is the repeated selection, 

 for in advertising his wheats he distinctly announced " ' Bred' upon the 

 .■same principle of repeated selection which has produced our pure races 

 of animals." But if we look more narrowly into the question it will be 

 found that from the very nature of his method he started in each case 

 from a single plant and thereby made a selection from the types in his 

 field just like Le Couteur and Shirreff, and it was this initial choice, not 

 his subsequent selections, which gave the hereditary qualities to his pro- 

 ductions. This produced the desired strain and it had sunplv to be kept 

 pure and reproduced. This is shown, firstly, in the fact that he produced 

 independent varieties, such as Golden Drop Wheat and Chevalier Barley, 

 which ha^'e retained their distinctive characters e\en although their 

 pedigree-culture has long since been discontinued. And secondly that 

 some of his cultures turned out failures, as in the acknowledged case of 

 .his " Original red wheat," showing that no amount of subsequent selec- 

 tion could make up for the initial .selection if faulty. There is no doubt, 

 however, as to the improvement effected by his methods of culture in the 

 productiveness of his selections, but being acquired during the life-time 

 •of the individual he rightly assumed that it had to be continued in order 

 that his pedigree-races might be kept up to their highest point of develop- 

 ment, although he confounded this with his continued selection. For 

 general practice, the selected plants should be reared on the same soil 

 •and under the same conditions as the ordinary field crops ; but any culti- 

 vator who will make the right initial selection from a single individual 

 and attend to careful cultivation as Hallett did. should succeed in in- 

 creasing the yield trom this cause alone, so long as the practice is kept np 

 year after year. 



Improvement of Cereal.s on a Large Scale. 



We will now glance at a modern establishment where the process of 

 selection is carried out on a large scale. 



Of all the present da\ institutions devoted to the improvement of 

 ■cereals, perhaps that of Svalof in Sweden is the most instructive example 

 for us and the one from which the most useful and up-to-date information 

 may be obtained. It has been so successful in attaining the objects for 

 which it was established, that a short account of its origin and progress 

 will be given and this account is entirely ba.sed on that given by Professor 

 De Vries in his valuable work on Plant-breeding. The institution, kno-wn 

 as the Swedish Agricultural Experiment Station, exists for the benefit of 

 •the farmers of Sweden and the results of the work are accordingly pub- 

 lished in the Swedish language. 



It took its rise in the little Swedish village of Svalof, opposite Copen- 

 hagen in Denmark, where a company was formed in 1886 for the produc- 

 tion and improvement of seed-grain's. It was onlv too evident that the 

 Swedish sorts were deteriorating- and required an infusion of new blood, 

 .and the aim of this new companv was simply to test new and foreign 

 varieties, some of which might be found suitable to replace the inferior 

 kinds. It had a definite object in view, to introduce new and better 

 seed-grains and to see that they were pure before placing them in the 

 hands of the Swedish farmer. 



The Companv was founded bv pri\ ate agriculturalists and its objects 

 were immediately practical. It did not attemf^t to educate the farmer 

 nor to carrv out purelv scientific researches, but in order to succeed in 



