322 ANIMAL COMMUNITIES 
are fundamental in ecology. We can sketch out here only such methods 
as are modifications of the usual method of these branches of biological 
science in such a way as to be intelligible to those somewhat familiar 
with such laboratory methods. 
(a) Experiments in the field are of prime importance in ecological 
work. Here smaller animals can be secured in numbers and subjected to 
experimental conditions before their physiological state has been modi- 
fied by bad treatment. Any student competent to undertake ecological 
investigation will find no difficulty in devising apparatus which can 
be carried into the field and which will enable him to do work of a 
high degree of scientific accuracy. Each experiment should be accom- 
panied by a control. That is, the same number of animals should be 
put under the same conditions as in the experiment, except for the one 
factor which is to be varied. For example, in an experiment designed 
to determine the reaction of animals to light, the control should be 
either equally lighted or entirely dark (more easily accomplished), and 
the experiment which is exactly the same except that the light ranges 
from darkness to bright sunlight. 
The apparatus which we have just begun to develop for this purpose 
is still in need of much perfecting. Thus far it consists of granite-iron 
and galvanized-iron containers about 13 in. long, 3 in. deep, and 4 in. 
wide. These are provided with galvanized-iron covers, somewhat larger, 
and a little deeper. One of these is provided at one end, with an adjust- 
able slide which may be used to open a slit to admit light when desired. 
In connection with this slit a mirror is provided with which the sunlight 
may be projected into the pan as nearly vertically as possible. The rays 
are allowed to pass through a water screen to cut out the heat. For work 
with temperature the same receptacles have been used and temperature 
differences secured by placing one end of the experimental tank in contact 
with hot soil and the other with cold soil. Land animals are confined in 
tubes 11 in. long by r# in. in diameter with round bottom and close- 
fitting cap, shaped like the bottom. Reactions to gravitation have been 
tested with the use of wire cylinders for land animals, and glass cylinders 
lined with screen for aquatic animals. Black covers are used to exclude 
light in various ways as a check. For the study of reactions to current 
two long galvanized boxes (24X54 in.) have been used, one having 
screen ends and the other tight ends. They are placed in the stream 
side by side, one serving as an experiment, the other as a control. Large 
tin pans have been used in connection with the long boxes, the water in 
the experiment being stirred so as to produce a circular current, while the 
