16 DR. WM. SAUNDERS. 



west where so much depends on the conditions of moisture, the results are very con- 

 tradictory. The crops had at the experimental farm for the maritime provinces 

 follow ours very closely except that it does not seem quite so important to get the 

 grain in very early there; but the sowing of grain in these provinces should be 

 finished within fourteen days after the season opens, to give the best results. 



By Mr. Featherston: 



Q. You mean the season opens earlier there ? 



A. Yes; a little earlier, and the spring season is longer. Peas have been tested 

 for five years, and the following results have been had. The second sowing has given 

 the best results, beyond that, a delay of one week has resulted in a loss of 4 per cent, 

 two weeks 12 per cent, three weeks 22 per cent, and a delay of four weeks has caused 

 a loss of 30 per cent. The lesson taught by these experiments is that in Ontario and 

 Quebec all cereal crops should be in the ground within ten days after the season 

 opens. To accomplish this, the land intended for the spring crops should be ploughed 

 the previous autumn, eo that it may be got ready for seeding when spring opens 

 with the least possible delay. 



By Mr. Rogers: 



Q. And if any crop must be late, peas will bear late sowing the best? 



A. Yes; but still the loss is from 22 to 30 per cent if sowing is delayed for three 

 or four weeks. 



Q. But something must be the last? 



A. Yes; that is necessarily so. Spring wheat should be sown first, for the 

 reason that the loss from delay is greater there, and after that barley and oats. The 

 loss, however, from delay in sowing these is pretty nearly the same in each case. 

 Peas hav6 less loss than any other of the four crops named from delay. 



By the Chairman : 



Q. There is another lesson to be taught from this, and that is the importance 

 of having the land underdrained, so as to be able to get on it very early in the 

 spring ? 



A. Yes, Mr. Chairman, that is a very important point, and one which cannot be 

 too strongly urged, as the draining of land often enables a farmer to get in his crop 

 from ten days to a fortnight earlier. 



EARLY AND LATE SOWINGS OF KOOT8. 



Along this same line, I brought before the committee two years ago the results 

 we had from the sowing of root oiops at different periods. We have now teried for 

 five years two periods of sowing, one being two weeks later than the other. The 

 date of the sowings has not been the same each year, for the reason that the seasons 

 vary considerably. In 1895, when we began this experiment, the first sowing of" 

 turnips was made on May 11 and the second on June 12. Twelve varieties were 

 sown, and the average gain from early sowing of these twelve varieties was 1 ton 

 642 pounds per acre that year. 



In 1896 the season was a little earlier, and we began on May 8, when the first 

 sowing was made, and the second on May 22. With fourteen varieties which were 

 sown that year the results wore in favour of early sowing by an average of 4 tons 

 1,424 pounds per acre. 



In 1897 the earliest sowing was on May 8 and the second sowing on May 21. 

 The gain on the plots early sown for the eighteen varieties tested that year was an 

 average of 3 tons 1,870 pounds per acre over those later sown. 



In 1898 the first sowing was on April 28 and the second on May 6, and the gain 

 from early sowing on the nineteen varieties sown that year was 1 ton 488 pounds 

 per acre. 



