6 BULLETIN 14G1, U. S. DEPARTMENT OF AGRICULTURE 
in 1921, and in most cases the number of plants seemed considerably 
reduced. Practically no wild cucumber plants were found during 
1922. After the first eradication no mosaic milkweeds appeared 
except in one case, but 8 of the 16 fields in the area showed traces of 
mosaic on July 12. The first appearance of the disease was in a field 
which contained a large number of mosaic milkweeds in 1920 and 
which the grower had left practically uncultivated, making it diffi- 
cult to prevent the appearance of some mosaic plants. The in- 
fection of the other fields, however, was not traceable to this source. 
The disease progressed gradually throughout the season, and on 
August 28 the average infection per field was approximately 50 per 
cent, or more than twice that of the previous year. 
Since practically no mosaic milkweeds or wild cucumbers had 
appeared since June 15, a search was made for other possible wild 
host plants, and evidence was obtained which indicated that the 
cultivated ground cherry, Phy sails pubescent L., was susceptible to 
cucumber mosaic (4. 11). Further observation showed that two 
species of wild perennial ground cherry, P. heterophylla Nees and 
P. subglabrata Mackenzie and Bush, occurred in considerable num- 
bers, and later experiments proved that both species were susceptible 
to cucumber mosaic (11). It seemed probable, therefore, that some 
of the infection might be traced to this source. 
MARENGO, ILL. 
The size of the experimental area at Marengo, 111., was reduced 
about one-half in 1922. It contained eight fields, and the work was 
carried on in the same manner as at Rockland, the eradication being 
confined to the immediate vicinity of the fields. A number of mosaic 
milkweeds were found at one or two- points, but they were much 
fewer in number than at Rockland. Practically no mosaic occurred 
in the area in 1922, but, since little had occurred in the previous two 
years and other fields in the vicinity showed little infection, no con- 
clusions could be drawn from the work. 
The experiments of 1921 and 1922 showed that mosaic was not 
sufficiently prevalent to make this section suitable for conclusive 
experiments on the control of the disease, and the work was discon- 
tinued after 1922. The experiments at Marengo, however, indicated 
a direct correlation between the presence of mosaic on wild host 
plants and its occurrence in the field. The comparatively few mosaic 
hosts found at Marengo were accompanied by a correspondingly 
slight occurrence of the disease in the field, whereas at Rockland the 
greater amount of mosaic on the wild hosts was coupled with severe 
losses each season. 
EXPERIMENTS OF 1923 AT ROCKLAND, WIS. 
As a result of experiments based on observations made in 1922, it 
was found that cucurbit mosaic was transmissible to the two species 
of perennial Physalis mentioned above, P. heterophylla and P. sub- 
glabrata, by means of the cucumber aphis (11). (PI. 3.) It was 
also found that the mosaic occurring on these hosts in the field was 
transmissible to the cucumber by this insect, and it was evident, 
therefore, that they were factors in the overwintering of the disease. 
