304 ANNUAL REPORTS OF DEPARTMENT OF AGRICULTURE. 



know accurately the conditions prevailin^^ at such times. These 

 records show definitely in many instances the cause of decreased 

 yields, and taken in connection with the plat tests bein^ carried on by 

 the Oflice of Dry-Land Agriculture provide definite answers to many 

 of the questions relating to the most advantageous use of dry-farming 

 lands. 



One of the most important factors governing crop production in 

 drj'-farming regions is, of course, the rainfall. Another factor, how- 

 ever, of almost equal importance is the amoimt of evaporation. The 

 latter has been measured at each of the e:vi)oriment farms since they 

 were first established, and the data thus secured has been of the great- 

 est value in the interpretation of the crop yields at the stations. A 

 bulletin on this subject has been published during the past year in 

 wdiich the distribution of the rainfall and evaporation in the (Jreat 

 Plains and Intermountain States is given in detail and the relation 

 of the two factors to dry farming in these regions fully discussed. 



The most important line of work developed in the Office of Physi- 

 cal Investigations during the past year has been the determination 

 of the extent to which different crop plants vary in their ability to 

 reduce the moisture content of the soil before w^ilting takes place. 

 It has been believed that drought resistance is largely dependent 

 upon the ability of a crop to reduce the soil-water content lower than 

 is possible in the case of nondrought-resistant crops. A long series 

 of measurements made in cooperation with Dr. H. L. Shantz, of the 

 Office of Alkali and Drought Resistant Plant Breeding Investiga- 

 tions, has shown that all plants are practically alike in this respect. 

 In other words, if the water in a given soil is nonavailable to one 

 plant it is also nonavailable to all other plants. Drought resistance, 

 then, does not depend upon the ability of one plant to reduce the 

 moisture content of the soil to a lower point than is reached by 

 another plant. The establishment of this fact is of fundamental 

 importance in the study of drought resistance, since this supposedly 

 important factor can be eliminated and attention focused upon other 

 matters, 



AVhile plants do not differ materially in the percentage of water 

 which they leave in a given soil at the time of wilting, this percentage 

 varies to a remarkable extent according to the kind of soil used. 

 Thus, plants wdll reduce the water content of a coarse sand to 1 per 

 cent before wilting, while they will wilt in the heaviest types of clay 

 with a water content of 30 per cent. It is therefore essential in con- 

 nection with any critical study of field crops in regions of limited 

 rainfall to be able to determine accurately the available water supply 

 at any time. This is represented by the water content in excess of 

 the wilting coefficient, the latter term representing the Avater content 

 at the time the plants wilt. The extension of the above investigation 

 has shown that the wilting coefficient can be quickly determined by 

 several indirect physical methods without the necessity of growing 

 and wilting a plant in a particular soil. Of these indirect methods 

 the most satisfactory is the moisture-equivalent method, in which the 

 amount of moisture which the soil can hold in opposition to a power- 

 ful centrifugal force (1,000 times gravity) is determined. This 

 tears a fixed ratio to the wilting coefficient, independent of the kind 

 of soil used, so that the wilting coefficient can at once be calculated. 



