38 
BULLETIN 124, U. S. DEPAETMENT OF AGRICULTURE. 
was again too high, and a large enough number of worms were not- 
killed to justify the expense incurred. 
ROLLING AND BRUSH DRAGGING. 
At the time a field is being damaged by the worms the hay that 
remains undestroyed can be cut and then either a brush drag or a 
roller run over the ground, by which a great many of the larvae will 
be destroyed. Some experiments tried along this line by Mr. T. 
Scott Wilson were quite successful. On August 15, in a 5-acre patch 
a brush drag was used and a great many larvae were killed. This 
field was overrun by Bermuda grass, which protected many larvse 
that would have been killed. A roller here would doubtless have 
(Original.) 
mashed all larva?. On the 26th of August another test was made, 
using the same drag. In this case the larvse were about full grown, 
and 55 per cent were killed by the operation. The latter experiment, 
however, was carried on in alfalfa of considerable height, and con- 
sequently the larvae were afforded much protection and as large a 
percentage was not killed as would have been the case had the drag- 
ging immediately followed cutting. 
A good brush drag and one that is well adapted to dragging 
alfalfa is shown in figure 20. The plan for constructing this, as 
given by Mr. E. S. G. Titus, in Bulletin No. 110 of the Utah Agricul- 
tural Experiment Station, is as follows : 
The drag is made by laying the butts of rather short brush, five or six feet 
long, in a row on a plank twelve or fourteen feet long, then another row should 
