36 BULLETIN 1727, U. S. DEPARTMENT OF AGRICULTURE. 
To further test the possibility of leaf infection from soil, a large 
sample of soil was collected from anthracnose centers in field 2 on 
August 14, 1916, four days after a rain. Under healthy runners 
in each of eight locations in four plats in field 1, 4 tablespoonfuls of 
this soil were sprinkled. In an examination of these runners on 
August 31 infection was found in one ¢ase consisting of lesions on six 
leaves, abundant on three. In this instance the soil inoculum had 
been sprinkled upon the leaves and then shaken off, while in all 
the other tests an effort was made not to sprinkle the soil on the 
leaves. 
These tests prove that the spores of this fungus are abundant 
in the soil under diseased plants immediately after a rain and are 
viable even four days later. The common occurrence of anthracnose 
lesions on the lower side of a watermelon fruit in the field indicates 
infection from spores in the soil or in drainage water. 
With regard to the spattering of spores upon healthy leaves, two 
tests were made. Within two hours after a heavy rain, September 5, 
1916, an undiseased leaf was taken from field 2 near diseased leaves 
and the blade washed in 200 c. c. of sterile water. From the five 
dilution plates poured from this wash water, one colony of anthrac- 
nose was isolated. This indicated, according to the quantity of 
wash water represented, that there were 96 spores on the leaf lamina. 
A similar test made with an undiseased seedling within 1 foot of 
diseased leaves yielded negative results. The isolation of the fungus 
from an undiseased leaf proves that the spores were present on the 
leaf surface after a rain. 
During both seasons of 1915 and 1916 anthracnose occurred 
on plants under cheesecloth cages which eliminated both insects 
and pickers as agents of dissemination. Since these plants were 
very evidently not original centers of infection, the entrance of the 
fungus can be attributed only to water splashing throws! the cloth 
or, more likely, washing under the cages. 
Among watermelons in the field, cases have been noted in which 
the fruit lesions were arranged in simileal rows in such a way as to 
indicate beyond: a doubt drip infection from overhanging diseased 
leaves. 
ARTIFICIAL WATERING. 
Another line of evidence relative to the local spread of the disease 
by the spattering of water is furnished by the severity of anthrac- 
nose as noted among certain of the fields of cucumbers in coldframes 
at Norfolk, Va.,in 1917. Here, daily watering by an overhead sys- 
tem was practiced, and this may explain in part the advanced 
development of the disease in these fields as compared with its 
relative obscurity in other fields of the region. 
