14 BULLETIN 796, U. S. DEPARTMENT OF AGRICULTURE. 
3 ounces each and over were entirely free from larvae; 20 trees re- 
ceiving 2 ounces each on June 24 contained 7 larvae, an average 
infestation of 0.35 per tree; 5 trees receiving 2 ounces each August 
22 contained none; 20 trees receiving 1 ounce each on June 24 con- 
tained 55 larvae, an average infestation of 2.7 per tree; 5 trees re- 
ceiving 1 ounce each _ on August 22 contained none; 15 trees receiv- 
ing one-half ounce each on June 24 contained 54 larvae, or an average 
of 3.6 per tree ; and 5 trees receiving the same dose on August 22 con- 
tained none. It is especially significant that the single application 
of a moderate dose the last of August gave as good control, in the 
end, as a June application of a much larger dose. 
INFLUENCE OF SOIL TYPE. 
So far as the present experiments go the influence of soil type on 
the effectiveness of the application of p-dichlorobenzene has im- 
posed no limitations on its use. While the fumigation probably 
may be expected to prove somewhat less effective in some cases where 
extreme variation in soil types exists, it has been used so far with 
almost uniform results on a variety of soils. In this respect it pos- 
sesses a distinct advantage over gases like carbon disulphid and 
carbon tetrachlorid, which volatilize so rapidly at ordinary tempera- 
ture that the fumigation is over in a few days. The relative dif- 
fusion of gases at various times in a given soil type depends mainly 
on the variation of moisture content. A highly volatile gas for that 
reason possesses a distinct disadvantage, since it is governed entirely 
by the particular condition of the soil, with respect to moisture, 
which exists within a period of a few days. This condition may be 
one of extreme dryness or extreme wetness, or, in case of a heavy 
rain, both conditions may be present. It is apparent, therefore, 
that a material of low volatility at ordinary soil temperature, which 
gives a comparatively mild fumigation, and requires a long period 
for its action, not only takes advantage of the equalization of soil 
moisture conditions that usually occurs over a period of several 
weeks, but at the same time is less seriously affected by sudden 
violent fluctuations that may occur at any time. 
Table VI gives the results obtained on several types of soil vary- 
ing in character from sandy loam to comparatively heavy clay loam. 
At Pinto, Md., where the soil type is a variation of the Franks- 
town series, known as stony silt loam, locally called chert land, the 
greatest difficulty is encountered in its use. When cultivation of 
this soil is followed for some time, it appears often almost entirely 
free of fine material on the surface. Below the immediate surface, 
however, it is usually well filled with fine soil. The chief difficulty 
in making application to this type of soil was due to the lack of 
fine material. As will be seen in Table VI the results were quite 
