Jones—Relation of Soil Temperature to Disease in Plants. 451 
on old infested fields, no matter how well fertilized and tilled, will 
remain stunted by the root rot during the earlier cool weather. With 
advent in midsummer of the higher soil temperatures, the root rot 
is suddenly checked, and as by magic the plants spring forward 
and produce a surprisingly good crop in September. The correct¬ 
ness of this interpretation and the details of how the temperature 
change operates were clearly evidenced by the following experi¬ 
ments, which Johnson has repeated with a number of plants. 
Plants have been grown in the temperature tanks for a period at 
the lower temperature, say 18°C., where the root rot develops so 
badly that the plants barely live, then the same pot is transferred 
to a higher temperature tank, say at 28°, when the plants quickly 
recover. Again Johnson has taken up stunted, sickly plants from 
a badly infested field in late summer and placed them in pots so 
that the soil temperature might be experimentally controlled. One 
half of these have been held at relatively low soil temperature, 
about 20 °C., and the other half at a high temperature, about 30 °C. 
The plants had practically no roots when set into the temperature 
tanks, only the blackened base of the stem remaining. At the 
higher temperature new roots promptly developed which remained 
white and healthy at the end of the experiment one month later, 
whereas at the lower temperature, although new roots continued to 
be formed, they survived barely long enough to keep the plant alive 
without growth. When at the end of the experiment the root sys¬ 
tems were washed out for examination, the vigorous healthy roots 
which had grown at the higher temperature, in contrast with the 
blackened diseased root which had developed at the lower tempera¬ 
ture, showed strikingly the dominant influence upon tobacco root 
rot of a change of only a few degrees in soil temperature (PI. 
XXXVI, fig. 1). 
Equally conclusive evidence of these temperature relations to 
crop production has been secured in a more direct manner. By 
means of soil temperature readings taken daily at Madison during 
the growing period, it was shown, for instance, that the soil tem¬ 
peratures in 1915 were relatively very low (average for June, July, 
and August 20.3°C.) as compared with 1916 when the average for 
the same period was 27.7° C., or 7.4° C. higher, a difference which 
it may be noted almost bridges the gap between the temperature 
most favorable for the Thielavia disease and one at which the para¬ 
site could not function. Correlated field records show that in 1915 
