4 BULLETIN 1461, U. S. DEPARTMENT OF AGRICULTURE 
land, however, had shown that at least one perennial wild host, the 
milkweed, existed outside the cucurbits and indicated that the work 
must be continued before successful control methods could be 
established. 
EXPERIMENTS OF 1921 
ROCKLAND, WIS. 
During 1921 the experiments at Eockland, Wis., were continued 
over approximately the same area as in 1920, but an effort was made 
to remove all mosaic milkweeds as well as wild cucumbers. No effort 
was made to remove ail healthy milkweeds, but all the plants found 
were examined and any groups of plants showing mosaiclike symp- 
toms were removed. Since the roots of the milkweed usually lie 8 
to 15 inches below ground and extend laterally for some distance, it 
was not found practicable to dig the plants. The shoots were pulled 
out and frequent inspections made in order to remove any new shoots 
which might appear later in the season. An effort was also made 
to enlist the cooperation of the growers on whose land mosaic plants 
were found. Where the plants were pulled up, the mosaic milk- 
weeds did not appear to send up new shoots in less than 10 to 
15 days, and in dry weather no new shoots developed for some weeks, 
if at all. In other cases, however, mosaic milkweeds appeared late 
in the season on cultivated land on which no plants had been found 
during the early summer. A portion of the 1921 infection in fields 
at both Eockland and Marengo was traceable to such plants. 
The eradication work of 1920 was found to have greatly reduced 
the number of wild cucumbers at Rockland and only a few mosaic 
plants were found. Several groups of mosaic milkweeds were dis- 
covered, all of which were on land which had previously grown 
cucumbers. Subsequent observations in various localities have con- 
firmed this fact, both with respect to the milkweed and with other 
perennial hosts as well. Examination of thousands of milkweeds 
growing on land which has never been planted to cucurbits has 
shown only three cases of mosaic. It is probable that the milkweeds 
are first infected from adjacent mosaic cucumbers through the 
agency of the cucumber aphis and during the following year act as 
sources of infection to the cultivated cucurbits. It is, therefore, 
highly important that both healthy and mosaic milkweeds in the 
vicinity of cucurbit plantings be destroyed. 
The 1921 experiments at Eockland included IT cucumber fields, 
all of which remained free from mosaic until August 1. On the 
final inspection on August 24 the average infection per field was 
approximately 25 per cent, as compared with 80 per cent in 1920. 
The effect of the removal of mosaic milkweeds was apparent in the 
case of 3 fields in which such milkweeds occurred in 1920. In that 
year the cucumber plantings in all these fields were practically 
destroyed by the disease early in August. In 1921 the same fields 
were planted to cucumbers, but the milkweeds were removed. On 
August 24 none of the fields showed more than 10 per cent of mosaic. 
These results at Eockland showed an apparent reduction in the 
amount of mosaic, but as there appeared to be a marked decrease 
in the mosaic infection in most sections throughout the State, there 
was a possibility that more than one factor was involved. At the 
