40 BULLETIN 727, U. S. DEPARTMENT OF AGRICULTURE. 
drainage water also passed through the rows of Cucurbita gourds, 
none of which became diseased (fig. 14). 
In field 2 all of the rows of interplanted seedlings showed recent 
and even-aged infection on September 13. Beaae leaf and coty- 
ledon lesions numerous enough to cause these organs to curl up, 
some seedlings bore girdling stem lesions as well. This infection 
is likewise attributed to surface drainage during the rains of the 
week previous. 
That the spores are washed from the leaves to the soil was proved 
by the tests previously cited. That the spores are carried consider- 
able distances by the surface run-off during rains and spattered from 
this drainage water upon all plants along the drainage channel is to 
be inferred from the observations presented above. 
To test experimentally the latter hypothesis, steps were taken to 
prove that the spores were present in surface drainage water. On 
August 14 four glass tumblers were sunk flush with the soil at the 
points indicated in figure 13, so as to intercept surface drainage 
water during the next rain. No tests were made until the rains of 
the first week in September. On the morning of September 5 the 
traps were emptied. A heavy shower occurred in the afternoon, 
after which the contents of the four traps were collected in sterile 
flasks. At the same time samples of the soil near two of the traps 
were collected. 
Each sample of drainage water was tested by poured plates as 
follows: Of the drainage water 1 c. c. was transferred to a flask con- 
taining 99 c. ¢. of sterile water, and from this 1 c. c. was transferred 
to a second 99 c. c. flask. Using recorded amounts of inoculum 
from these dilutions, three plates were poured from the first flask 
and two from the second in water agar plus 2 per cent dextrose. 
No colonies of the anthracnose fungus appeared in any of these 
plates. The soil samples were also tested by the method outlined 
earlier, and only negative results were secured. 
That spores of the anthracnose fungus were present in the samples 
of drainage water was proved in another way, however. The con- 
tents of each trap, after the removal of the sample for plating, were 
sprayed upon healthy potted cucumber plants.t The results of 
these tests are presented in Table IV. 
Spores of the anthracnose fungus were present, therefore, in three 
of the four traps. Trap 1, in which no spores were found, was located 
near the upper edge of the field, and hence there was a smaller reser- 
voir of disease contributory to it. 
After the additional rains of September 6 st a series of five 
plates each was poured from traps 1 and 2 in field 2, and, as in the 
1 These inoculations, as well as those described later, were made by Dr. E. Carsner in connection with 
his investigation of the angular leai-spot of cucumber. 
