42 
Journal of Agricultural Research 
Vol. XXXI, No. L 
large that they must have been present for at least three seasons 
prior to the observations. Under these conditions, the presence of 
aphids would be very likely to result in infection of the young cucum¬ 
ber plants and the disease would in this way be introduced early in 
the season. The field evidence, therefore, indicates that the poke- 
weed should be classed with the wild cucumber and milkweed as a 
source of primary infection in some localities. 
PRESENT STATUS OF THE PROBLEM OF OVERWINTERING 
From the work described above, it is evident that the importance 
of the various wild host plants varies with the locality. Up to the 
present, however, all of our observations in Wisconsin, Indiana, and 
Illinois have shown that mosaic infection is common on at least one 
of the above wild host plants in all districts where * the disease is 
prevalent on the cultivated cucurbits. The overwintering of mosaic 
on the wild cucumber, milkweed, and pokeweed seems to be estab¬ 
lished, but it is possible that the list of host plants may be increased 
by further investigations. 
Recent experiments have confirmed the work of Muncie'(./£), 
which indicates that the catnip, Neveta cataria L., may also be a 
perennial host of cucurbit mosaic. The importance of this host in 
the field in Wisconsin has not yet been determined. Studies by 
one of the authors (Walker) have also indicated that certain perennial 
species of Physalis are susceptible to cucurbit mosaic and the possi¬ 
bility of other sources of infection is indicated by the results of cross 
inoculations with annual plants of other families. These experi¬ 
ments have shown that cucurbit mosaic may be transmitted to a 
wide variety of host plants and have furnished evidence which indi¬ 
cates that certain supposedly distinct mosaic diseases may be 
identical. If this is the case, the possible sources of infection for 
the cucurbits might perhaps be considerably increased in number. 
The results of these experiments with annual host plants of the cucur¬ 
bits and of other families are given in the following section of this 
paper. 
CROSS-INOCULATION EXPERIMENTS WITH CUCURBIT MOSAIC 
CROSS INOCULATIONS FROM THE CUCUMBER TO OTHER SPECIES OF THE 
CUCURBITACEAE 
It has been shown in earlier papers ( 8 , 11) that many species of 
the Cucurbitaceae are susceptible to cucurbit mosaic. In these 
earlier experiments all of the cucurbits tested proved to be sus¬ 
ceptible to the disease, with the exception of those in the genus 
Citrullus. In this genus, which includes the watermelon and citron, 
infection was produced on the green-seeded citron only. The 
results with this latter host were also confirmed by the work of 
Jagger {If). These cross-inoculation experiments have since been 
continued and a considerable number of additional species and vari¬ 
eties, including many European, Asiatic, and African cucurbits, have 
been found to be susceptible to the disease. The earlier results have 
been confirmed in that all varieties except the genus Citrullus have 
proved to be susceptible to the disease. Inoculation of many vari¬ 
eties of watermelon and citron, however, have never resulted in 
infection, except in the case of the green-seeded citron just mentioned. 
