414 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1959 
piration is not reduced, surface cooling may occur and be taken 
advantage of, but the animals will be confined to the eryptozoic niche. 
Larger animals can afford surface transpiration, and in normal en- 
vironments this may be used as a temperature-regulating mechanism. 
It seems, then, that three interdependent variables have to be con- 
sidered: an animal’s size, the nature of its respiratory system, and 
the permeability of its integument. We should expect that while 
certain combinations of these factors would permit life in fully ter- 
restrial conditions, other combinations would forbid it. Thus small 
size combined with an occlusible respiratory system and an imper- 
meable integument is a satisfactory combination for land life; large 
size, a ventilated respiratory system, and a permeable integument is 
also satisfactory, but other combinations are not. To these inter- 
dependent factors must be added a number of others which are not 
interdependent—that is to say, their optimum value lies always at one 
end of the range. Thus it is always advantageous in hot dry sur- 
roundings to lose as little water with nitrogen excretion as possible, 
and it is always advantageous to have a high upper lethal tempera- 
ture. Optimum values of these factors are, of course, not always 
attainable. The evolutionary history of an animal may forbid it. 
Thus a mammal owes its success to homoeothermy—constant tem- 
perature—and to permit the body temperature to rise as high as it 
may in some insects would adversely affect the whole organization. 
When this happens, the delicate mechanism of the brain is the first to 
suffer. 
Before we examine the validity of these principles in real situations, 
something must be said of the other side of the picture, that is, gain 
of water. This may be achieved by straightforward drinking, when 
water is available, or by taking in water with the food. A few animals 
can absorb water by rectum (some woodlice do this) or through 
the skin. But these are comparatively unimportant for our purpose. 
A few arthropods can absorb water vapor from unsaturated air. 
This is a remarkable and rather rare phenomenon, but one which may 
have a wider significance in water conservation than is apparent. It 
has been reported in ticks and certain beetle larvae in air with a 
relative humidity of 80 percent or above, and in flea larvae in even 
drier air (references in Edney, 1957). Now the osmotic pressure 
of the body fluids of insects and ticks is in equilibrium with air at 
about 99 percent relative humidity. There can therefore be no ques- 
tion of water flowing down an osmotic gradient. The mechanism of 
this active transport has not been demonstrated, but a further piece of 
information may be relevant. There is some evidence, obtained by 
the use of heavy water, that water in the body fluids of an insect is in 
continuous interchange with water vapor in the air outside. If this 
