62 BULLETIN 879, U. S. DEPARTMENT OF AGRICULTURE. 
niary, 1918, in connection with the other test of seed from mosaic 
micrampelis mentioned below. No mosaic symptoms appeared on 
any of the plants. No mosaic was present in the greenhouses in the 
vicinity of Madison, and while striped beetles attacked the vines, 
their importance in overwintering the disease is unknown. 
Wild micrampelis plants are common in all parts of Michigan and 
Wisconsin, and in addition to their occurrence in nature the vine is 
often planted where a quick-growing ornamental is desired. After 
being once planted it usually continues to appear each year from 
self-sown seed. 
The presence of mosaic on Micrampelis lobata in advance of its 
appearance on any cultivated cucurbit host, coupled with the fact 
that the striped beetles feed upon micrampelis as soon as they emerge 
from hibernation and go directly from the wild plant to the cultivated 
cucumber when the latter appears, offers a most promising explana- 
tion for the early infections of the cucumber. There is little doubt, 
apparently, that Micrampelis lobata may be a source of such early 
infection to the cucumber. 
The source of primary infection on the wild plant is still uncertain, 
but much of the work on the overwintering of the disease on the cul- 
tivated cucumber seems equally applicable to the wild species. Soil 
is probably not a factor and the question of seed as a means of trans- 
mission is still unsettled, although the appearance of the disease in 
the same spot each year would indicate that the seed may be the 
means of overwintering. During the winter of 1917-18, 1,100 plants 
were grown in the greenhouse from seed collected from mosaic 
micrampelis plants in Michigan, Indiana, Wisconsin, and Colorado. 
.An additional lot of 1,000 plants was also grown from seed taken from 
plants supposedly free from mosaic. The conditions under which the 
plants were grown were unfavorable, however, and caused the leaves 
to develop various abnormal colorings, which made it difficult to 
determine whether any of the plants were infected with mosaic. 
Inoculations from suspected plants gave negative results in all cases, 
but further tests of seed from mosaic plants are in progress under 
more favorable conditions for observation. 1 
WILD NONCUCURBITACEOUS HOSTS. 
The work of Jagger (19), already mentioned, has shown that the 
mosaic disease of cucurbits may be transferred to some plants outside 
the Cucurbitacese. Although there is no definite evidence that the 
disease overwinters on plants of other families, it is possible that such 
wild hosts, particularly if they are perennial, might serve as sources 
of early infection to the cucurbits, especially as cucumber insects are 
often found feeding on plants of other families. All cross inocula- 
i Subsequent tests have proved that the mosaic disease may be carried over winter in the seed of 
Micrampelis lobata. See Phytopathology, vol. 9, p. 326, 1919, 
