22 BULLETIN 204, U. S. DEPARTMENT OF AGRICULTURE. 
some species has been favored in situations quite unsuited to their 
requirements. This is particularly true in the case of the oaks and 
gray birch. These species constitute a very large proportion of the 
deciduous growth of the region and are very susceptible to gipsy- 
moth attack. On a great deal of the area now covered by these 
species the white pine would grow to much better advantage, would 
yield a much more valuable product, and if pure, or nearly so, would 
prove immune to gipsy-moth attack. The white pine reproduces 
itself readily under favorable conditions, and is already fairly 
abundant in numerous localities. For these reasons the white pine 
recommends itself very strongly as a substitute for the existing moth- 
susceptible species, and this species has been considered to a very 
large extent by this department in its experiments to create a safe 
stand of timber. The deciduous species which are of value and im- 
mune to gipsy-moth attack require most favorable situations for 
their profitable development, and such situations are very few and of 
small area. An exception in the case of chestnut may be made in 
this connection. This is a valuable tree and one well suited to grow 
over a considerable area. Its growth is not recommended on account 
of its susceptibility to the widespread and fatal chestnut blight. But, 
like the red oak, it may be found possible to grow it satisfactorily in 
small numbers with other species. 
EXPERIMENTAL WORK. 
The experimental work has been conducted by means of small 
areas known as "sample plats. 7 ' These vary in size from one-half 
an acre to 6 acres and occur both scattered and grouped in different 
parts of the infested region. They are necessarily located on the 
lands of private owners who are willing to submit their lands to this 
use. 
An effort has been made to distribute this work as widely as pos- 
sible over the infested region. (PL IX.) In this way the greatest 
variety of conditions is encountered and the results are available to 
the greatest number of woodland owners. Each sample plat varies 
from another in one or more of the following points: Composition of 
stand, age of stand, degree of infestation, and method of treatment. 
Each sample plot generally consists of two parts. One of these is 
the portion upon which actual experimental work is done and the 
other serves as a control or check plat. Upon the latter nothing 
whatever is done, as its purpose is to provide a means of comparing 
results under natural and artificial conditions. The corners and 
boundaries of all plats have been plainly marked, and the areas 
surveyed and mapped. All trees 1 inch and over in diameter have 
been calipered on each plat and control, and the measurements 
recorded. Forest descriptions of each plat have been written. 
