THE WATER-COST OF DRY MATTER 137 



conditions, obtained similar results. The transpiration 

 ratio of wheat varied from season to season from 507 

 to 883, and of corn from 332 to 477. Both the Utah and 

 the Indian experiments showed, for all crops investiga- 

 ted, a similar seasonal variation in the water-cost of dry 

 matter. Whenever the season is favorable for the produc- 

 tion of much dry matter, the water-cost is reduced; that 

 is, a good season produces not only a large yield, but pro- 

 duces it with a relatively small quantity of water. A 

 poor, backward season not only produces a small quantity 

 of dry matter, but produces it at a high water-cost. The 

 variation in the water-cost of dry matter, with the sea- 

 sons, is much less than that of the total yield, as determined 

 by the seasons. The water-cost is, however, influenced 

 materially by the general seasonal conditions. 



98. The soil. The vital relation of the soil to crops 

 would naturally suggest that the quantity of water 

 required to produce one pound of dry matter would be 

 partly determined by the nature of the soil. This has been 

 conclusively demonstrated by many elaborate investiga- 

 tions. Pagnoul, working in France, found that the trans- 

 piration ratio of fescue grass on a fertile soil was 555; on 

 an infertile soil, 1,190. In the Utah experiments, the 

 transpiration ratio for corn was 386 on College loam, 408 

 on Sanpete clay, 561 on sand, and 601 on clay. Similar 

 variations, some much larger, were observed with other 

 crops on similar soils. For corn, the transpiration ratio 

 varied, according to the soil used, from 432 to 579; for 

 wheat, from 466 to 849. Similar results could be quoted 

 in great abundance to substantiate the statement that 

 the nature of the soil is a determining factor in the 

 relative quantity of water used by the plant for the 

 production of a given quantity of dry matter. 



