74 



Journal of Agricultural Research 



Vol. XVII, No. 3 



been based on observation rather than on experimental data. The 

 question is an important one from a practical standpoint. Will the 

 selection of soils low in organic matter or avoidance of the use of green 

 or barnyard manures materially aid in reducing the disease? 



This is one of the most difficult problems to subject to experimental 

 test in such a way that wholly reliable conclusions can be drawn. It 

 illustrates equally well the fallacy of drawing far-reaching conclusions 

 from mere field observation. It is evident that the organic matter of 

 the soil has profound influence upon a large number of other factors 

 such as water-holding capacity, food supply, temperature, reaction, 

 texture, aeration, and saprophytic growth of microorganisms in the 

 soil. To eliminate all these factors, even in the most carefully con- 

 trolled experiments, is impossible. To judge of their relative impor- 

 tance in the results obtained in an experiment, however, on the basis of 

 the behavior of such factors from other experimental evidence, is quite 

 likely to yield fairly reliable results. 



T.\BLE VIII. — Influence of amount of organic matter in soil en rootrof of tobacco 



o Leaf mold heated to 110° C. before infestation. 



An attempt was therefore made to arrive at such conclusions by study- 

 ing the behavior of the disease in the purest ground quartz sand avail- 

 able and also in pure leaf mold, together with mixtures of the two in 

 various proportions (Table VIII). The chief difficulty met with at once 

 in such a combination is to obtain an approximately uniform supply 

 of plant food in these various media. The leaf mold was found to con- 

 tain sufficient plant food to support normal growth, though after the 

 third crop the plants showed potash hunger. To the pure quartz cul- 

 tures a nutrient solution sufficient for plant growiih was added, and 

 decreasing amounts were added to the various mixtures of sand and leaf 

 mold with a rough estimate of the amount of nutrient salts required. 

 Two pots of each series were then inoculated with T. hasicola and two 



