WATER RESOURCES IN ARID REGIONS 93 
cast and to evaluate these periods of drought, as well as the periods 
of abundance that present the same characteristics. Even if the 
arrival of a given period cannot be predicted, it is useful to know 
its statistical probability which has evident political and economic 
consequences: a completely unorganized country will suffer 
famines which could be avoided by an organization at the pro- 
vincial level. More and more serious famines could be avoided 
only by national, federal, or international organization. 
There exist, in fact, periods of famine, of long duration, cover- 
ing more or less extensive territories. They have meteorological 
causes that should be studied in two phases: (a) the meteorological 
situations which cause the periods of famine, and (4) the causes 
of such meteorological situations. 
The statements made regarding the historical permanence of 
the climate lead to the supposition that variations in climate are 
the result of causes which themselves fluctuate. Thus one pro- 
ceeds to a study of the fluctuations in climate and to a study of 
the fluctuations in all the geophysical or cosmic phenomena 
capable of varying in a concomitant manner, choosing those 
phenomena whose fluctuations can be forecast, such as cycles of 
sunspots. 
The simplest method has been to try to deduce periodicities 
from the analysis of past observations and to suppose that they 
will continue in the future. This analysis, to be convincing, should 
be made by dividing the time into two periods for which one 
possesses information. The first serves for analysis, the second for 
verification. The conclusions are of value only if they cover a 
group of stations for which there are available long-term observa- 
tions. 
To my knowledge these attempts have not yet yielded usable 
conclusions. It is hoped that conclusions may be reached in the 
future since documentation at hand is constantly increasing and 
also the field of study of geophysical and cosmic phenomena is 
widening. 
In addition, the study of natural fluctuations in climate has 
become more and more urgent in the past few years with the 
appearance of possible causes of artificial man-made variations: 
