62 Journal of Agricultural Research voi. xvii, No. » 



water. Two jars in each compartxnent containing sterilized or uninfested 

 soil were used as controls^for plant growth alongside two jars containing 

 infested soil. Naturally infested soil from the old tobacco field on the 

 Station farm, previously referred to, was used in most of the experi- 

 ments. After being given a good application of well-rotted manure, 

 the soil was thoroughly mixed and screened before weighing equal 

 quantities into the jars. The sterilized soil used in the earlier experi- 

 ments was sterilized by steam to destroy the infestation by T. basicola. 

 Considerable difficulty was experienced, however, as a result of the toxic 

 action of the heated soils on plant growth at the lower temperatures, 

 which interfered to some extent with the reliability and uniformity of 

 the data obtained by leaf measurements. 



In later experiments the employment of soil steamed two or three 

 weeks previous to being used and allowed to stand in a moist condition 

 at a fairly high room temperature reduced this action to a minimum. 

 In still other tests formalin-sterilized soil was used with equal success, 

 and in the final experiment another uninfested soil was used and artificial 

 innoculation resorted to for the infested series. 



The data taken in the earlier experiments were mostly in the form of 

 measurements of leaf area in square inches as determined by a standard 

 chart of various leaf sizes whose areas had previously been determined by 

 the use of a solar planimeter. In later experiments air-dry weight de- 

 terminations of the stalks and leaves were made. 



The determination of the actual amount of disease on the roots is, of 

 course, the final criterion for judgment, and in the last experiments it was 

 found that with care the greater part of the roots could be washed out 

 from the soil, examined for disease, dried, and weighed; these weights are 

 closely correlated with growth aboveground, so that either the area of 

 the leaves, weight of the leaves and stalks, or weight of the roots alone 

 give a good index of the extent of the disease. A preliminary report of 

 the results obtained has been given and an abstract published (15). 



Eight separate experiments have now been made upon the influence of 

 soil temperature on the extent of the root disease, four determinations 

 being made in the winter and spring of 191 7, and four during the fall and 

 winter of 191 7-1 8. The first experiments were made over a range of 

 about 35° C, but, as these were found to be beyond the ranges of normal 

 growth and infection, the later experiments usually included a tempera- 

 ture range of about 15°. Three of the experiments failed more or less 

 to give uniform results; one due to nematode infection at the higher 

 temperatures, another to toxic action of the heated soil, and a third to the 

 accidental use of infected seedlings. 



The procedure in each experiment consisted merely in filling the re- 

 quired number of jars with soil; one-half with uninfested and one-half 

 with infested soil. Glass tubes 2)4 inches long, were inserted into each 

 jar to permit watering part of the soil at about half its total depth. After 



