20 INSECTS INJURIOUS TO FORESTS. 
BIOLOGICAL EXPERIMENTS. 
THE TERMITARIUM. 
Early in March, 1912, work was begun on a large outdoor termi- 
tarium, at Falls Church, Va. The projects outlined for study in 
connection with the termitarium were (1) to observe the progressive 
development of the termite from the larva to the adult, especially 
in the case of the colonizing form; (2, to watch the insects when they 
swarmed, to determine whether this swarming is a nuptial flight, 
i. e., whether or not the sexes leave the nests in separate swarms 
and at what time copulation occurs; (3) to observe how the new col- 
ony is established, (4) in what proportion the pairs survive, and (5) 
the conditions in the parent colony after the swarm is over. 
The termitarium was merely a large cage in which the termites 
could swarm under conditions as nearly similar to those in nature 
as possible. The cage consisted of a bottom of galvanized iron in 
the form of a rectilinear box sunk in the ground to the depth of a 
foot, a framework of wood to support the wire netting, and a wooden 
roof covered with tarred paper. The bottom of the cage consisted 
of a galvanized-iron box 10 by 6 feet and 1 foot deep, of 27-gauge 
galvanized sheet iron with an inch flange turned inwards. (Fig. 5, a.) 
This box was raised off the ground on a wooden floor of seven-eighths 
inch material, 4 inches wide, which was laid on 2 by 4 inch studding. 
There was an air space of 1 inch between the sides of the box and the 
earth, as well as a similar space between the sheet-iron bottom and 
the earth. The top of the sheet-iron box was nailed to a framework 
of 2 by 4 inch studding, which was supported on cedar posts, the top 
of the box being several inches above the surface of the ground. 
Loose earth was placed in the box to about half its depth, over which 
was spread a shallow layer of black earth (leaf mold) from the forest 
floor, for the purpose of retaining moisture. The wooden framework 
consisted of 2 by 4 inch studding, which was covered on the inside 
with galvanized- wire netting (14 squares mesh per inch) and rein- 
forced at the side and ends by boards 4 inches wide by seven-eighths 
inch thick. The cage is about 7 feet in height, and the roof, slanting 
away from a shed against which it is built, acquires sufficient pitch 
to shed rainfall. (Fig. 5, b.) 
Two chestnut logs which had been cut and allowed to season 
about one month before the experiment was begun were placed in 
the cage. One was placed on end in the dirt, the top of the log 
approximating the height of a stump, while the other, a tangential 
section, was partially buried in the ground. The bark was loosened, 
but was left intact on both logs. The logs were both sound, but 
were kept moist. A number of small decaying branches and strips of 
decaying chestnut boards were also laid flat upon the earth in the 
