18 BULLETIN 833, U. S. DEPARTMENT OF AGRICULTURE. 
application all the cuttings but one in row (a) were dead. In row 
(b) all looked very poorly. In (c), the check row, one cutting was 
dead while all the remainder looked fairly healthy. 
Conclusions.—It appears that the variety may have played an 
important part in this case. In this lot, as was so evident in the 
previous treatment, (a) causes injury, while (b) gives a double bene- 
fit—first, it does not cause injury; and second, it protects the plant. 
The Sets: spread to the check row (c), at two galls had adults 
emerge from them. 
LOT 3. 
Three rows of seven infested cuttings each. 
Row (a). Nicotine sulphate 40 per cent (1-1,000) plus fish-oil soap 1 ounce 
_ to 1 gallon of solution. 
Row (b). Volatile nicotine sulphate 40 per cent (1-1,000) plus fish-oil soap 1 
ounce to | gallon of solution. 
Row (c). Check; not treated. 
Results.—This variety stood up much better than did that in lot 2. 
Row (a). Slight burning evident. In fair condition 10 days after treatment. 
Row (b). Only two poor plants; no burning. All recovered 10 days after 
treatment. 
Row (c). All plants in good condition, slightly better than row (b). No 
galls developed. 
Either treatment (a) or (b) might be used successfully. 
FINAL CONCLUSIONS. 
Comparing the three lots with one another the following conclu- 
sions may be drawn: The practices described injure the cuttings 
only slightly and afford reasonable protection from the midges. 
EXPERIMENTS ON CONTROL OF GALL STAGE. 
The final consideration concerns the control of the gall stage, 
which is the hardest to combat. Owing to the habit of the larve 
of burrowing into the plant tissues, the chief difficulty is that most 
spraying mixtures when applied to foliage fail to exert their toxic 
properties against the insects contained therein. This may be 
accounted for by the resistance of the leaf structures to penetration 
by insecticides. 
EXPERIMENT 1. 
A preliminary experiment, in which was used 1 part 40 per cent nico- 
tine sulphate to 250, 500, and 1,000 parts water, respectively, to which 
fish-oil soap or laundry soap at rate of 1 ounce to each gallon of 
solution was added, showed conclusively that one application was 
entirely ineffective in controlling the gall stage. The data which 
were taken at various intervals showed that the adults emerged 
