93 



each sort usually occupies two plots which are located some distance from 

 each other, with a view to equalizing as far as possible, the effects of soil 

 variations, etc. The value of these preliminary yielding tests has been re- 

 peatedly proven, showing as they do how easily one may be deceived by 

 appearances. A good example is given in the crossing between Cone and 

 Grenadier wheat. Here two lines in the fifth generation, judged from appear- 

 ances, seemed equally promising, but the yielding tests, even on these small 

 plots, revealed a decided difference and on this evidence the choice was 

 made. The best yielding line came into the large comparative trial plots in 

 1909, since when it has shown the same good yielding power which dis- 

 tinguished it in the smaller plot. 



The large comparative plots determine the fate of a given sort, in so Lar y e com ~ 

 far as such can be determined at the station, as here it must enter into com- , tg 

 petition with the very best old and new sorts. 



All sorts which have been produced by the Association and which are 

 now on the market must be represented in these trials in order that later 

 productions may be compared with them. Only in this way is it possible to 

 know whether a new sort is better than the old and thus deserves to come 

 into general cultivation. 



When a sort of good repute is first imported from another country it is 

 placed either directly in these trials or tested for a year or two in the pre- 

 liminary trials. If it proves promising and is not already a pedigree sort, a 

 number of plants may, under certain circumstances, be taken for pedigree 

 culture. If, on the other hand, it is known to be a pedigree sort no further 

 selection is given it, its future depending entirely upon its performance in 

 comparison with other sorts. Each of these large trial plots is 2 metres wide 

 (about 6 feet) by 25 metres (82') long, therefore occupying about .012 

 acres. Heretofore each sort occupied three such plots but in 1910, owing to 

 the great importance of equalizing soil variations, arrangements were made 

 to increase the number to four, although the size of each plot was reduced 

 so that practically the same total area will still be occupied. While even a 

 greater number of plots should properly be devoted to each sort, yet owing 

 to the fact that a large number of sorts must always be under trial, it is not 

 found practicable to handle more than four. In arranging the different plots 

 .care is taken that no two plots of the same sort adjoin each other. By this 

 arrangement variations in the quality and character of the soil may be 

 overcome materially, and thus a fairer expression of the yielding power of 

 each sort obtained. 



These plots are sown with a small one-horse machine of 7 drills, each 

 plot having 14 drills (See Fig. 20) . This machine is extremely simple and 

 can be thoroughly cleaned of all seed within the shortest possible time. A 

 new machine has recently been devised by Mr. W. Kinberg, field foreman at 

 Svalof, as a result of the past 20 years' experience which marks certain 

 improvements over that formerly used. This machine is based on the Sachs 

 System, but includes a number of changes which render it more suitable for 

 sowing small experimental plots. 



