Cereals. 1 09 



seeds which it produces only the very best are kept for 

 sowing for the continuation of the stock. The com- 

 mercial race is not in the strict sense of the term a race, 

 for the successive generations which compose it are not 

 genetically connected. Each generation begins as a lat- 

 eral branch of the main stem; the first harvest, after the 

 seeds for the pedigree stock have been set apart (and 

 after inferior seeds have been rejected) is grown on 

 special fields for at most 2 or 3 generations to produce 

 the quantity of seed required for the market. In every 

 successive year, therefore, the stock for the market starts 

 as a new branch of the main stem ; it is not for two or 

 three years after improvements appear in the latter that 

 they are obtainable in the former. 



It obviously follows from this that the race never 

 becomes independent of selection, as a true species or 

 subspecies does. There are, of course, beautiful ex- 

 amples of subspecies among cereals, but they have not 

 arisen by selection. Patrick Shirreff's older varieties 

 are examples of these; they are independent of selection 

 and often so good that they cannot be improved by it — 

 e. g., the Talavcra wheat. A sharp distinction must 

 therefore always be made between species and subspecies 

 on the one hand, and races on the other. 



The greater the improvement of a race has been, the 

 more does it deteriorate in ordinary cultivation ; its seed 

 can only quite exceptionally be used for further crops. ^ 



IMethodical selection is of tw^o essentially dift"erent 

 kinds. Their most distinguished exponents were Hal- 

 LETT of Brighton (England) and Rimpau of Schlanstedt 

 (Saxony). I shall now endeavor to give an accottnt of 



^ I shall deal with this question in more detail in § 14 of this 

 part. 



