LEAKAGE OF GAS IN SM \i.l. 1 BEES. 
43 
destroy the purple scale in all its stages on the leaves and wood. If 
the tree contain fruit infested with this scale it will be necessary to 
slightly increase the dosage. The exact amount of this increase can 
not be stated with accuracy at tins time, owing to the fact that in the 
single experiment performed very little infested fruit from winch 
data might be secured was available. 
Table VI. Fumigation for (hi purpU scale, experiment No 
[Length of exposure, one and one-hall hours; size of trees mostly 7 to 10 feet; oe< aslooall; on 11 oi I 
» >n ii'.i\ es 
and branches 
< »n fruit. 
Number 
of trees 
treated. 
( yanld 
per 100 
cubic feel 
of space. 
Insects all \ e, 
i normal, 
I n ects ;iin e, 
approximate^ . 
approximate^ . 
ipproxlmatelj 
i g i normal 
Ounce*. 
4 
. 
3 5 per cent. 
10 50 per cent. 
10 15 per cent. 
Vbove 75 per cent. 
l 
l 
l live female. 
20 25 i«t cent. 
i 2 per cent. 
i i 75 pel cent 
4 
! 1 

i 5 per cent . 
4 
1* 
i) 
l 2 percent. 

rwoor inges examined; 
many norma] i 
present . 
1 
i; 

3 Instances ol doi • 
in ;i l 
" 
A I'm normal 
4 
2 

o 
(") 
(") 
4 
• >i 


(") 
4 
u 


(«) 
'"> 
a \o material for examination. 
In this experiment live insects were found on the branches and 
leaves in the cases where three-fourths ounce and 1-ounce dosages were 
employed. Normal eggs were found up to and including the 1 |-ounce 
rate, but were destroyed by dosages exceeding this. As in the case of 
experiment Xo. 2, so little scaly fruit was available at that time that 
we are inclined to consider the results in this part of the test as yet 
incomplete. 
THE LEAKAGE OF GAS IN FUMIGATING SMALL TREES 
When the results of experiment No. 3 are compared with those of 
experiment No. 2 we are at first led to believe that an error has been 
made. Jn experiment No. 2 it was found that the 1 \-ounce dosage 
rate destroyed all insects and eggs on the leaves and branches, whereas 
in this experiment it required one-half ounce more cyanid per 100 
cubic feet, or a 2-ounce dosage rate, to accomplish the same result. 
Since the period of exposure was thirty minutes longer than that of 
experiment No. 2, we would naturally expect that tin' results accom- 
plished would be as good or better, all other conditions being the 
same. The apparatus and chemicals employed were identical in both 
cases; and the conditions under which the fumigation w as conducted 
were practically the same. There was, however, one difference: The 
trees involved in the one and one-half hour fumigation were much 
smaller than those of the one-hour test. This fact accounts for the 
