64 
To determine the influence of wet plants on the amount of gas, a 
half-bushel of leaves were sprinkled, shaken, and placed in the frame. 
_After five minutes only 54.3 per cent of the calculated amount of gas 
remained, or 75 per cent of what would have been secured without the 
leaves. In other words, the moisture absorbed one-fourth of the gas. 
This influence of moisture must needs be considered in the fumigation 
of nursery plants packed in wet moss, or sprinkled, or in field work 
after a rain or heavy dew. The bottom of the frame was then removed 
and the gas generated over soil containing an average degree of mois- 
ture. After five minutes 45 per cent of the calculated amount was 
found remaining, or 62.7 per cent of what would have been secured 
were the box closed. Thus the soil and possible leakage, which was 
slight, as the soil was well banked‘and packed around the bottom, took 
up three-eighths of the gas generated. This influence of the soil upon 
the amount of gas is very important in the use of these frames with a 
large surface exposed to the soil, which, of course, decreases the ratio 
of the soil surface to the volume of the gas. 
The diffusion of the gas in a wooden box which had been made for 
fumigation of nursery stock was then tested. This box is 8 by 3 by 23 
feet, and was made of two thicknesses of matched flooring, with build- 
ing paper between. The lid fitted tightly and the joint was sealed with 
putty. With the generator and intake at the same end, at the end of 
two minutes 33.7 per cent of the calculated amount was found; in ten 
minutes, 62.7 per cent; in twenty minutes, 56 per cent. With the 
generator and intake at opposite ends, however, 265 per cent was found 
in two minutes, 160 per cent in five minutes, 92 per cent in ten min- 
utes, and 9.5 per cent in sixteen hours. Thus the diffusion of the gas 
in this shaped space is just the opposite of that in a long frame—that 
is, in the cubical space the gas is puffed to the opposite end at the end 
of two minutes, but about one-third being at the end of generation and 
two-thirds at the opposite end, and requiring about ten minutes for 
the complete diffusion, while in the long frame the gas became *‘ banked” 
at the point of generation, and is never completely diffused without a 
mechanical agitation. 
These in brief are the facts drawn from the laboratory study of the 
diffusion. How were they supported by experiments in the field? 
It must be remembered that in the field tests a larger amount of gas 
would be required to compensate the influence of the soil and foliage. 
The triangular frame 15 feet long by 8 inches high by 18 inches wide, 
with a door at the center to admit the generator, was first used. It 
will be noticed that this frame has but about one-fourth the area in 
cross section of the square frame used at the laboratory, and therefore 
the diffusion would be at least inversely that much more difficult. 
With two-tenths gram KCn per cubic foot air space all lice were 
killed at the middle of the frame in ten minutes, the generator being 
