274 PLANT-BREEDING 



include three subdivisions. These have to deal respectively 

 with the morphological marks, with the agricultural qualities, 

 and with the correlations between these two groups. The 

 agricultural qualities may be assumed to be manifest, but 

 the botanical peculiarities have hitherto been neglected, or 

 studied only in a superficial way, so far as their appreciation 

 was needed for the ordinary botanical systems. 



These older systems were based on a very imperfect 

 knowledge of the real nature of the variability. No sowings 

 had been made in order to study this phenomenon, and it 

 was simply assumed that all minor peculiarities, in which the 

 single individuals of a species differed from one another, 

 were changeable and did change from one generation to the 

 next. On this assumption their intricate nature could hardly 

 be of any significance for the systematist, and was therefore 

 almost always neglected. 



The discovery of the numerous elementary species, of 

 which the common varieties consist, has at once changed 

 the whole aspect of this question. Numerous apparently 

 insignificant marks proved constant in pure sowings. They 

 offered an easy and reliable means of distinguishing strains 

 o'f widely divergent practical value. Thence the conclusion 

 that they must be subjected to an exact study, which would 

 enable the breeder to discover any new and useful type that 

 might occur in his fields. 



In order to do this, a thorough study had to be carried 

 on for a number of years. The single characters had to be 

 classified, so as to give distinct systems. In these, each 

 botanical mark may be indicated in combination with its 

 correlative significance. By this means, whenever a distinct 

 practical quality is wanted, the correlated botanical marks 

 may be looked up in the system. 



Such elaborate systematic surveys have been published 

 from time to time by the experimenters of the staff of Svalof. 



