ATMOSPHERE 161 
definite positions with reference to wind (anemotaxis) (128a), as for 
example some flies hover in the air in one position with the head toward 
the wind. Some animals, such as the land salamanders, frogs, toads, 
millipedes, spiders, and insects turn away from currents of air because 
of increased evaporation. 
7. ATMOSPHERIC ELECTRICITY (125) 
The effect of atmospheric electricity upon organisms is little known. 
It varies with variations in other conditions of the atmosphere. It will 
probably be found to be important in the life of animals. 
IV. COMBINATIONS OR COMPLEXES OF FACTORS 
As has already been pointed out (55), the animal environment is a 
combination of moisture, temperature, light, pressure, materials for abode 
and food, all of which factors taken together constitute a complex of 
interdependences. These various factors are so dependent upon one 
another that any change in one usually affects several others. This 
property of environmental complexes is what makes ecology one of the 
most complex of sciences, and experimentation in which the environment 
is kept normal except for one factor, an ideal rarely realized in practice, 
even under the best conditions. 
The efforts of ecologists, geographers, and climatologists have long 
been directed toward the finding of a method of measuring the environ- 
ment which shall include a number of the most important environ- 
mental factors. De Candolle undertook to base the efficiency of a 
climate, for supporting plants, upon the mean daily temperatures above 
6° C., this temperature being taken as the starting-point of plant activity. 
Merriam has followed this lead and calculated total temperatures for 
many places in North America and made maps and zones based upon 
such totals. This system, however, has been rejected by botanists and 
plant ecologists on account of much evidence, both experimental and 
observational, which is quite out of accord with this view. The scheme 
has not been generally accepted by zodlogists outside of the United States 
Biological Survey. There is practically no evidence of an experimental 
sort for the application of such a scheme to animals. Relative humidity 
has been suggested as an important index (128) but does not properly 
express the influence of atmospheric humidity upon the animal body 
(125, p..53). The saturation deficit has also been suggested but does 
not take temperature into account. 
