50 Journal of Agricultural Research voi. xvii. No. 2 



In order to get more accurate evidence on the influence of soil moisture, 

 a series of pot experiments, with the moisture supply controlled as 

 closely as possible by weight, were carried out. 



Several difficulties, not readily overcome, exist in such an experiment, 

 the principal one being that it is practically impossible to maintain a 

 uniform moisture content throughout the soil. No doubt the use of 

 Livingston auto-irrigators would have made possible more uniform results, 

 but these were not available at the time. Two-gallon crocks, perforated 

 at the base for drainage and holding about 10 kgm. of soil, were used. 

 The naturally infested soil from the old tobacco field on the Station 

 Farm was used. After a large quantity of this soil had been dried, 

 thoroughly mixed, and screened, its moisture content and water-holding 

 capacity were determined in the ordinary manner. Ten kgm. of the soil 

 were then placed in each of twenty 2-gallon crocks. The soil in 8 of 

 these crocks was sterilized by steam at about 100° C. for the purpose of 

 destroying all the T. hasicola present in order to provide disease-free 

 controls in the experiments. The water relations, as well as the food 

 relations, were, of course, changed in some degree by the sterilization, 

 and an absolute comparison between the sterilized and infested series 

 was therefore not permissible, although it is believed that the results 

 are not altered appreciably by this fact. 



The crocks of soil were then divided into four series, each containing 

 three infested and two uninf ested crocks of soil. Two glass tubes, X inch 

 in diameter, one being inserted to a depth of 2 inches and the other to a 

 depth of 6 inches, were placed in each crock for the purpose of permitting 

 a more uniform distribution of water in the soil. Of the four series, one 

 was now made up to one-fourth its full water-holding capacity, and the 

 others to one-half, three-fourths, and full water-holding capacity. After 

 the water had been allowed to distribute itself fairly evenly, one plant 

 of the White Burley variety grown in sterilized soil was transplanted to 

 each of 20 crocks. The loss of moisture from the crocks was very slow 

 when the plants were small, especially during the winter in the green- 

 house. Usually it was not necessary to make the pots up to the required 

 weights oftener than once every three days, but later in the tests daily 

 attention was usually necessary. In an experiment begun on February 

 13, 1917, with the White Burley variety, it was noted at the end of one 

 week that in the infested series the plants at one-fourth saturation wilted 

 during days of high transpiration and showed the poorest growth. The 

 plants at three-fourths saturation got the best start, while those at full 

 saturation were already yellowing and apparently diseased, since no 

 such condition was observed in the sterile controls. On March 5 the 

 conditions were about the same in relative growth except that the diseased 

 condition of the plants at full saturation in infested soil was greatly 

 increased, and the controls in sterilized soils were now beginning to forge 

 rapidly ahead of those in infested soil. On March 14 it seemed quite 



