EXPERIMENT ON SOIL IMPROVEMENT. 151 



In conducting the experiment on soil improvement according 

 to the plan outlined there has been some difficult}^ in selecting 

 crops that would give satisfactory returns in green manure be- 

 cause of the conditions under which they may be grown in the 

 experiment. Good crops of rye have been obtained each j^ear 

 on plot N. There was a fair growth of alsike clover on plots 

 M and P to plow under in the spring of 1900, but this had 

 been sown in the latter part of July, 1899, amongst the corn 

 after it was cultivated for the last time. Since the potatoes 

 grown in the experiment of 1900 were rather late in maturing, 

 and as clover could not be sown to advantage before digging 

 the tubers, it was thought best to try winter vetch as a catch 

 crop on these plots between the seasons of 1900 and 1901, but 

 as already explained, this crop made but little growth and the 

 very little quantity that could be plowed under somewhat mod- 

 ified the yields of oats and peas on these plots during the 

 experiments of 1901. Following the oats and peas, common 

 red clover was sown early in August, 1901. On the whole, 

 this crop gives the best promise of an}' 3'et tried for green man- 

 uring on this field, but in order to grow a crop of clover or 

 other legume following the regular crop of the experiment and 

 get a sufficient amount to plow under it is necessary either to 

 sow the clover in the growing crop late in July or early in 

 August, or to plant in the regular rotation only those crops 

 which ma}' be harvested early in the season and thus allow the 

 clover to be sown on the cleared field not later than August. 



Yields of oat and pea hay in igoi . — The 3'ields of oat and pea 

 hay from the different plots in the experiment of 1901 are given 

 in the following table. 



There was not a very heavy growth on any of the plots, the 

 heaviest yield being that from plot K with the complete fertil- 

 izer, as has also been the case each year of the experiment thus 

 far. As the field slopes considerably from plot P to plot K 

 there may be more moisture in the soil of the latter plot and 

 possibly some of the fertilizers from the soil of the other plots 

 may be washed upon it, which would probably benefit the 

 growth of the crop on plot K. However, from a comparison of 

 the costs of the fertilizers used on the two plots as well as of 

 the 3aelds produced there does not appear to have been any 



