ANIMAL SURVIVAL IN HOT DESERTS—EDNEY 409 
evolved the highly efficient tracheal system. This consists of numer- 
ous branching tubules leading air directly to all parts of the body 
from a small number of external openings, the spiracles, which are 
capable of being closed. Occlusion of the spiracles reduces loss of 
water from inside the tracheae when the insect is at rest and oxygen 
is required only in small quantities. Because of this respiratory 
system it has been possible to reduce cutaneous absorption of oxygen 
to a minimum, so that the integument need not be permeable to 
water and is, in fact, efficiently waterproofed. Although the total 
surface area of the animal is of the order of 5,000 times greater than 
that of the spiracular openings, as much as 60 to 70 percent of the 
total water lost by transpiration passes through the spiracles. 
Spiders, another very successful group of arthropods, have not 
developed the tracheal system very effectively, and they rely upon 
“book lungs” (the term is self-explanatory) situated in pits whose 
openings can also be closed. The system is good enough to provide 
for normal oxygen requirements, and the integument can be water- 
proofed; but spiders cannot indulge in long bursts of great activity 
because they rapidly run out of oxygen. 
Another class of arthropods with pretensions to land life are the 
Crustacea, and these animals present a very different picture. They 
have never evolved an effective respiratory system for use on land. 
Woodlice absorb oxygen through external, leaflike organs which are 
but little modified from the gills of their aquatic relatives. Tran- 
spiration from the gills constitutes some 40 percent of the total tran- 
spiration, and the absolute rate of transpiration is much higher than 
it is in insects. Some woodlice, indeed, have developed short bunches 
of internal tubules in the gills, and these are the most “terrestrial” 
of the group. But in all of them absorption of oxygen also occurs 
through the general body surface, and the integument is much more 
permeable to water than is that of insects. This is true only in humid 
air, however; otherwise the outer layers of the skin become too dry, 
and measurements show that the effect of exposure to very dry air 
may be to cause death by asphyxiation rather than by desiccation 
(Edney and Spencer, 1955). 
NITROGEN EXCRETION 
As regards nitrogen excretion, an interesting relation between the 
availability of water in the habitat (particularly during embryonic 
development) and the nature of the end product has been pointed out 
by several authors (Needham, 1929; Delaunay, 1931). 
We have seen that in land animals the waste products of nitrogen 
metabolism cannot be excreted as ammonia because this is a highly 
toxic and highly soluble substance. More wasteful methods have to 
be adopted, and ammonia is converted into urea, uric acid, or other 
