48 BULLETIN 879, U. S. DEPARTMENT OF AGRICULTURE. 
subsequent development is most common, but in some instances the 
disease has suddenly developed in epiphytotic form, a striking ex- 
ample of which occurred at Madison, Wis., in 1916, followed by 
a second and more severe case in 1917. During 1916 the writer was 
not at Madison, but Dr. M. W. Gardner reported that the disease 
appeared early in July and in a short time in many fields had affected * 
25 to 50 per cent of the plants. 
During 1917, frequent observations were made on the experimental 
fields at Madison from the beginning of the season. The fields under 
observation consisted of six plats within a radius of 1J miles and 
varied in size from one-half to 1 J acres. These were practically the 
only large cucurbit fields in the locality, with the exception of a few 
plantings of muskmelon and squash. Up to July 19 no mosaic had 
appeared on any cultivated cucurbits in the locality so far as known. 
On this date, however, two of the six plats showed a number of cases 
of mosaic in its earliest stages. On July 21 definite symptoms of >* 
mosaic were noted on 25 per cent of the plants in these two plats and 
on a few plants in each of the remaining four plats. On July 26 at 
least 50 per cent of the plants in the first two plats and from 3 to 18 
per cent of the plants in the other plats were affected. A survey of 
gardens in the vicinity also showed a few cases of the disease on 
July 26. The plants noted between July 19 and 26 were all in the 
early stages of the disease, and so far as could be determined all the 
infection must have occurred within seven or eight days. Such a 
sudden development of the disease would seem to indicate that 
some agent of dissemination was present which produced rapid and 
widespread infection at a definite period. 
The most probable factors in overwintering would seem to be: 
(1) Soil containing refuse from mosaic vines of previous seasons; <| 
(2) the use of seed from diseased plants; (3) possible wild cucurbit 
hosts; (4) the presence of an identical disease on plants of another 
family; or (5) some insect agency. Most of the work has therefore 
been conducted with these possibilities in mind. 
SOIL AS A SOURCE OF INFECTION. 
It has been shown that the expressed juices of mosaic plants 
remain infectious for only a short period and that dried tissues of 
such plants have never shown the presence of the virus. This indi- 
cates that soil containing the remains of mosaic plants is not likely 
to prove a source of infection to further crops. I 
GREENHOUSE EXPERIMENTS. 
During the winter of 1916, 40 plants were grown in the greenhouse 
in soil from a field at Hamilton, Mich., where mosaic had been seri- 
ous for three successive years. The plants were transplanted to 
