296 
JOURNAL OP ECONOMIC ENTOMOLOGY 
[Vol. 11 
low tents were charged at the same time, thus insuring a similarity of 
climatic conditions. The tents were of the same material and in the 
same condition, but as an additional precaution, they were inter¬ 
changed. 
The locations of the highest percentage killed are in the following 
order: Top of low tent, center of high tent, top of high tent, bottom of 
low tent, and bottom of high tent. With the exception of one set of 
experiments, more were killed in the top than in the center of the high 
wean average 
Mean average 
killed 
52.2# 
Fig. 11. Represents the summary of 121 experiments in which 2,873 insects were 
used showing the percentage of insects killed in different parts of the high and low 
tent and the mean average killed in each tent. The mean average of the high tent 
excludes the results in the center of that tent because no insects were placed in the 
center of the low tent, hence a more equable comparison. 
tent, while in the totals there is a difference of only one-tenth of 1 per 
cent. While the difference in killing efficiency of the gas is not great 
between the center and top, there is a very great discrepancy between 
these locations and the bottom of the tent, a difference as shown in 
these experiments of 26.9 per cent. Likewise between the top and 
bottom of the low tent there is a difference of 18.9 per cent, and between 
the two tents there is a difference of efficiency, in favor of the low tent, 
of 19.4 per cent. 
The fact that hydrocyanic acid gas is lighter than air, must account 
for a greater concentration accumulating in the top of the tents. The 
fact that the percentage killed is greater in the low than in the high 
tent, must be accounted for through the dosage schedules now in use, 
which do not adequately allow for the variation in the shape of the 
trees. We are not at this time, however, proposing any change in the 
schedules in practical use. Further extensive field experiments rather 
than laboratory calculations, must be the final test for any system of 
scheduling dosage. From the experiments here recorded which are 
field tests, the statement appears to be justified that our present 
