ANTHRACNOSE OF GUCURBITS. 37 
SURFACE DRAINAGE WATER. 
But the mere spattering of rain and blowing of the droplets be 
wind can not account for the extensive epiphytotics of this disease 
which suddenly appear in certain fields. A considerable mass of 
observational evidence has been accumulated relative to this point. 
It has been generally noted that extensive spread of this disease 
does not follow every rain, but only the very heavy rains during 
which considerable surface run-off takes place. The general trend 
of the evidence seems to indicate that epiphytotics tend to occur 
more commonly where the topography of the field is sloping or 
rolling. 
As to the effect of heavy rains, it will be recalled that the Aa 
general spread of the disease in fhe Princeton fields in 1915 followed 
the rains of the first week in August. In the Madison fields in 1916, 
two periods of extensive spread of the disease were clearly recog- 
nized, one after the heavy rains of August 3 to 5 and the other, which 
resulted in a serious epiphytotic, after the heavy rains of September 
4to 7. Intervening lighter rains of 0.28 inch on August 10 and 0.31 
inch on August 26 were not followed by an extensive spread of the 
disease. It is realized that here there is a difference in duration as 
well as in amount, and the longer rainy periods no doubt afforded 
conditions more conducive to infection. Furthermore, reference has 
been made to the significance of the lower temperatures of the periods 
following the light rains. 
Additional observational evidence relative to the agency of surface 
drainage during heavy rains in spreading this disease is furnished 
by the direction of this spread from the old centers. This is revealed 
by the location of even-aged new infection following such rains. 
Field 2 at Madison in 1916 afforded a good opportunity to observe 
this phenomenon. This field occupied a decided south slope and 
the rows extended across the slope. The new infection resulting 
from the rains of the first week in August was very evident by 
August 12 and a careful inspection of the field showed the manner 
of spread from the old centers of infection to be somewhat as illus- 
trated in figure 13. There was a decided tendency for the greatest 
spread from the old centers to be across the rows and distinctly in 
the direction of the slope. At least nine of the areas of infection 
show elongations extending downhill; in fact, in four of these cases 
the new infection in the row below the old center was along small 
gullies or drainage channels leading directly from the old center in 
the row above. 
Field 3 occupied a gentle southeast slope, and here again the rows 
crossed the slope. One original center occurred in a row along the 
upper edge, and from this center a drainage channel led down across 
the entire field to the southeast corner, where it formed a delta 
