165 



in the soil for the use of growing plants during a drought, or to ger- 

 minate seed that might be planted. During the summer of 1897 this 

 experiment was repeated in a modified form. Four small plots were 

 laid off in a field where conditions were as nearly uniform as possible. 

 A mulch of green grass about four inches deep was placed upon No. 1; 

 No. 2 was left untreated; No. 3 was cultivated about 4 inches deep, with 

 a hoe; and No. 4 was raked about 1 inch deep with a garden rake. The 

 treatment given plots 3 and 4 was repeated after every rain as soon as 

 the crust became dry enough to work without puddling. The percent- 

 ages of moisture present at certain dates were as in the following table: 



Number 

 and Treatment. 



July 10 



CO 



bL 

 

 < 



Aug. 14 



Sept. 1 



Sept. 3 



Oct. 15 



No. 1- 



-Mulched 



27,7 



24.7 



29.1 



23.4 



20.9 



12.9 



No. 2- 



- Untreated 



27.1 



19.3 



23.3 



18.0 



17,0 



12.6 



No. 3- 



-Deep tilled 



27.9 



20.2 



23.9 



20.8 



20.9 



17.5 



No.4- 



-Shallow tilled 



27.4 



19.6 



23.7 



20.0 



20.0 



18.0 



It will be seen that at the beginning of the experiment, July 10, 

 the plots were of practically the same moisture content, but by August 

 13 there was a decided advantage in favour of the mulched plot. A 

 rainfall of 0.65 in. occurred on that date after the sampling, but from 

 that date until October 17 was a period ol almost unbroken drought. 

 The change in the moisture- content of the plots during this period is of 

 s pecial interest and value. 



On August 14 the mulched plot contained about 5£ per cent more 

 moisture than the others, and it continued to maintain about this ad- 

 vantage over the untreated plot up to September 1. The cultivated plots 

 lost water at a much slower rate than the mulched plot, so that by Sep- 

 tember 3, the three had practically the same amount. From that date 

 on, the tilled plots showed an increasing advantage over the other plots. 

 On October 15 the mulched plot was as dry as the untreated one, while 

 the tilled plots contained 5 per cent more than the others. The dry 

 soil-covering was thus very effective throughout the entire period in 

 conserving the moisture. The untreated plot was dry to the depth of 

 the sample, one foot, while the tilled plots were in excellent condition for 

 seeding below the soil mulch. The soil mulch was very dry,and as it was 

 a part of the sample in cutting a core to the depth of a foot, the part of 

 the sample below the dry covering was more moist than the figures 

 would suggest if compared with determinations of water in soil in the 

 ordinary state. The soil from the lower 8 inches was moist enough to 

 stick together when compressed with the hand, 



Experiments were also carried out on a large scale in the field, and 

 the effectiveness of simple ploughing as a means of conserving soil 

 moisture was clearly demonstrated. It is not necessary to go to large 

 additional expense in both tools and labour in order to preserve the 

 moisture, but it is necessary to do the ploughing promptly, while the 

 moisture is still in the soil. Another experiment was the following : 



