68 PLANT-BREEDING 



the year 1890, the present Director, Dr. Hjalmar Nilsson 

 was appointed as such. With a broad conception of the 

 practical interests of Swedish agriculture, he combined the 

 conviction that only really scientific studies are adequate to 

 the solution of difficult practical problems. A thorough 

 knowledge of the laws of variability and inheritance seemed 

 to him to be the principal requirement for the solution of 

 [The prevailing problems. As we shall presently see, it has 

 'afforded him the answer to the main questions and after- 

 l ward led him to the establishment of his great principle of 

 Lcorrelated variability as one of the principal foundations of 

 agricultural plant-breeding. But the interest of this sub- 

 ject is great enough to justify a separate treatment, and so 

 we shall at present defer it to another occasion. 



Before going into details, I will give a short survey of 

 the work, in order to facilitate its appreciation. Progress 

 has chiefly been achieved by the discovery of the numerous 

 elementary units of which the ordinary varieties of cereals are 

 built up. That varieties were, as a rule, neither pure nor 

 uniform, was a fact that could no longer escape observation, 

 and selection as a means of purifying the races and of keep- 

 ing them up to their main standard had already received 

 general recognition. But no agriculturist had even the re- 

 motest idea of the real state of their compound nature, and 

 not even the work of Le Couteur and Shirreff had been 

 sufficient to afford such a conception. A protean group of 

 types was found to constitute each so-called variety. These 

 types were seen to be different from one another in a pre- 

 viously unsuspected degree, covering a range of variability 

 adequate to comply with almost all the needs of practice. 

 Moreover, these types proved to be constant ; they had only 

 to be isolated and multiplied to yield new and uniform races, 

 directly suitable for the farmers' purposes. 



Confronted with these new facts, the current conception 



