NILSSON'S DISCOVERY 71 



to the same type. But, of course, the number of ears had 

 been different in the different groups, some being common 

 and represented by numerous individuals and others being 

 rare. Among the rarest, some types had only been met 

 with in one single head, and since the number of the ears 

 had been entered for each, it could be made out at the time 

 of the testing of the cultures, which among them were in 

 this exceptional condition. 



To this accidental circumstance, combined with the ex- 

 act scientific method of keeping extensive records, the dis- 

 covery of the cause of the diversity of the cultures was due. 

 For precisely those cultures which were derived from one 

 ear only were found to be pure and uniform, all others 

 offering to the eye a more or less motley assemblage of forms. 

 Hence the conclusion that cultures, in order to be pure, 

 must be started from single ears. Two or more ears may 

 seem similar enough to be considered as representatives of 

 the same type, while in reality they do not afford suffi- 

 ciently reliable marks. 



On the basis of this discovery, a distinction had to be 

 made between the selection of samples and the selection of 

 individual ears or panicles. The older experiments had 

 always been started from multiple samples, according to the 

 prevaiHng views, and it became at once clear that at least 

 one cause of their usual miscarrying must be sought in this 

 course of procedure. They must be designated as selections 

 of groups or of families, and could even appropriately be 

 denominated selections of crowds, an expression which would 

 at once convey the idea that the terms of selecting simulta- 

 neously more than one individual are intrinsically contra- 

 dictory. 



Cc(ntrasting with this old principle, the new one was 

 designated as that of the separate selection or separate cul- 

 tures. It has also received the name of pedigree- cultures, 



