the factors introduced by sxirface relief, soil and geological structure, which cause 

 great variations from place to place in the amount of available water. Third we have 

 the probability of prolonged periods of cloud associated with the rainfall. Fourth and 

 perhaps most important of all we have the variability of rainfall from one year to an- 

 other. In some cases fog, frost -rime and dew may also be important. 



Exclusion of Permanently Irrigated Areas 



Life can luxuriate in the atmospherically driest spots on earth provided adequate 

 and reliable water is made available (and provided this water can also drain away). 

 The words adequate and reliable should here be emphasised. Such permanently irri- 

 gated spots have of course many specialised biological interests, but they are clearly 

 not themselves part of a desert, though they may be surrounded by desert. The clear- 

 est example is an area to which the supply comes direct by river or canal from else- 

 where beyond the desert. Another not so well recognised example is the oasis fed from 

 a large artesian reservoir beneath the desert. Here the supply comes not from else- 

 where but from elsewhen — from the fossil water of the rains of long ago. When geolo- 

 gical permeability limits the rate of supply and the sites are limited economically by 

 the pumping lift from the water table to the surface, the supply is virtually inexhaust- 

 able. Desert biology ought therefore to be confined to life that relies on the precarious 

 atmospheric moisture supplied from within the desert region itself. 



Variations in Available Water Supplies from Place to Place 



Effect of surface relief, soil and geology. 



Because of the high evaporation light showers and dew ought to be excluded from 

 any estimate of the effective mean annual rainfall, except for those forms of life which 

 are specially adapted to absorb and store moisture very quickly. To what extent this 

 is possible appears to need a good deal more investigation. 



Owing to surface dryness, lack of plant cover and to the fact that desert rainfall 

 in general is characterised by heavy and infrequent storms, run -off is high and local 

 sub- surface storage low. Hence we may get very strong contrasts in the available 

 moisture between the catchment grounds and the drainage lines which thread them. If 

 the geology is suitable we may get considerable storage in shallow underground pools 

 along the drainage lines, where water is preserved from evaporation and is near enough 

 to the surface to be directly available. In this case one finds narrow streaks of vege- 

 tation threading barren country. In other cases the run -off water flows too deep below 

 the drainage lines, or there are no impermeable rock barriers to hold it up. On the other 

 hand desert rainfall is markedly affected by changes in ground elevation- An isolated 

 group of hills a few hundred metres high may attract rainfall many times that over the 

 surrounding country. So if the drainage lines are highly permeable to some depth, we 

 may find life confined to the high ground and none below. 



It is these places, where a good water supply is concentrated along the drainage 

 lines, but where not much is now directly available, that offer the greatest scope for 

 artificial improvement. Mislead by unfounded theories of very recent climatic change, 



8 



