REPORT OF THE DIRECTOR 



SESSIONAL PAPER No. 16 



and all tlic plots were similarly prepared. Six sowings were made in each case, the 

 first at the earliest time practicable, the second at the end of a week and others at the 

 end of each subsequent week until six successive sowings had been made. These plots 

 were all harvested and threshed separately, and the results recorded. The best crops 

 were had from the second sowings, made just one week after it was possible to get on 

 the land; beyond this delay has resulted in loss, which has become more serious as the 

 delay has been greater. The average of the ten years experiments shows that with 

 spring wheat a delay of one week after the period named has entailed a loss of over 

 thirty per cent, two weeks forty per cent, three weeks nearly fifty per cent and four 

 weeks fifty-six per cent of the crop. 



With oats a delay of one week has caused an average loss of over fifteen per cent, 

 two weeks, twenty-seven per cent, three weeks thirty-two per cent and four weeks forty- 

 eight per cent. 



In the case of barley a delay of one week has resulted in a loss of twenty-three per 

 cent, two weeks twenty-seven per cent, three weeks forty per cent, and four weeks 

 forty-six per cent. 



With peas a delay of one week caused an average loss of four per cent, two weeks 

 twelve per cent, three weeks, twenty- two per cent, and four weeks thirty per cent. 



The results of these experiments have been widely published and farmers in the 

 eastern provinces of Canada now pay general attention to early sowing. The early 

 sowing of grain in the Northwest provinces of Canada has also been shown to be 

 highly profitable to the farmer. 



THE SELECTION OF PLUMP SEED. 



The selection of plump and well ripened seed for sowing is also a great advantage. 

 In each seed is laid up a store of food to be used by the young plant in the early stages 

 of its growth. In a shrivelled seed the store which can be drawn on is very meagre and 

 the growth under such circumstances is slow, but in a well-developed and plump kernel 

 the supply is abundant and the plant starts out with a degree of vigour which is usually 

 maintained and the resulting crop, all other conditions being equal, is usually satis- 

 factory. 



SELECTION OP THE BEST AND MOST PRODUCTIVE VARIETIES. 



Another important consideration in connection with successful farming is the selec- 

 tion of the best varieties of seed for sowing, taking into consideration productiveness, 

 quality and earliness of maturing. That there are varieti3s more productive, of higher 

 quality or earlier in ripening than others has been abundantly proven, and the object 

 in view in experimental work along this line has been to introduce or to produce varie- 

 ties which combine these good qualities in the highest degree. 



As to productiveness, a quality of the highest practical importance, in the tests con- 

 ducted at the Experimental Farms, the variation in different varieties has been very 

 great. In plots of oats adjoining each other and all sown on the same day, the yield has 

 ranged from 89 to 42 bushels per acre. In spring wheat under similar conditions from 

 31 to 16 bushels, and in barley from 58 to 33 bushels per acre. The experiments car- 

 ried on have shown also that this productiveness is in a large degree persistent. During • 

 a five years' trial 41 varieties of oats were sown every season at the same time and on 

 adjoining plots. Each year a select list was published of the twelve heaviest yielding 

 sorts. During the whole of the five years only 15 of the 41 varieties found their way into 

 the select list, and 9 of these appeared among the best 12 sorts every year. 



Of spring wheat, 31 varieties were under trial for a like period. In this case 

 16 only of the 31 sorts have appeared among the twelve best yielding sorts during the 

 five years' test. The evidence obtained as to the persistent productiveness of certain 

 varieties of barley is also very striking. 



