sunshine the plants may use half of the moisture which the soil 

 contains. The draft upon the soil that occurs during this period of 

 growth, both as to moisture and mineral-food constituents, is prob- 

 ably as great as that under field conditions by the removal of a large 

 mature crop, providing we assume that the removal takes place to 

 no greater depth than 1 foot, which, in the case of wheat and other 

 similar crops, would be approximately correct. For example, the 

 amount of green matter in the plants of the untreated poor Leonard- 

 town loam soil was at the rate of 12,000 pounds to the acre-foot, and 

 the water removed was equal to 20 per cent of the weight of the soil, 

 or the equivalent of 3.1 inches of rainfall. 



A large number of moisture determinations show that the wheat 

 plants at the time of cutting from the baskets contain about 85 per 

 cent of water. On this basis the air-dry matter produced in the tops 

 of the plants would be equal to 2,000 pounds per acre. In case of a 

 mature crop of wheat we know that the straw and the thrashed grain 

 are about equal in weight, and on this basis the above amount of mate- 

 rial removed from the soil is equivalent to a crop of about 16 bushels 

 of wheat per acre, or far more than would probably be obtained in 

 practice on the field from which this sample was taken, especially 

 when untreated. 



Another portion of the same sample of soil when given a liberal 

 application of barnyard manure, nitrate of soda, sulphate of potash, 

 and acid phosphate produced a yield of green matter equal to 32,300 

 pounds per acre-foot of soil, and the plants used an amount of water 

 equal to 44 per cent of the weight of the soil, or the equivalent of 

 about 7 inches of rainfall. By computation as above this would be 

 equal to 45 bushels of wheat per acre, or a crop which on Leonard- 

 town loam soil would be considered unusually large, although per- 

 haps not beyond what has occasionally been obtained under most 

 favorable conditions in field practice. 



In the case of the sample of good soil the yield from the untreated 

 was 50 per cent greater than that from the poor, or the equivalent of 

 24 bushels of wheat per acre. The difference is not so great as 

 might be expected, judging from the crop-yielding power of the two 

 soils as observed in the field. In case of the poor soil, however, the 

 above-estimated yield of 16 bushels per acre for the untreated portion 

 is probably far greater than would have been obtained under field 

 conditions, and has probably arisen from the fact that the mechan- 



il condition of the soil has been very greatly improved as used in 

 basket experiments, and nearly the optimum amount of water has 

 all times been maintained. From the above data it will be seen, 

 lerefore, that by this method of cropping the draft made upon the 

 >il within the short period of twenty -five daj's is fully as great, if 

 greater, than that made in field practice during the whole season. 



