OASES. 303 
The five types of oases behave in different ways toward the questions as to 
their early distribution and what antiquity may be discovered. It is obvious 
that all streams, and even valleys now dry and their possible prolongations on the 
plains, should be explored for remains of types I, II, and III. The possible dis- 
tribution of type IV is more indefinite. They may have existed on almost any 
steppe of clay or loess, even where now invaded by dunes. Obviously, type V 
should be explored for on all lake-shores, even where no lake exists to-day. 
RELATIONS BETWEEN EROSION AND BURIAL IN THE OBLITERATION OF KURGANS 
(OASIS CULTURE DEPOSITS). 
The chances of finding culture mounds more than a few thousand years old 
seem relatively slight. Under the arid conditions of most of Central Asia, mounds 
are bare of vegetation and stand windswept with no protection from the wear of 
sandstorms. They are no more than naked heaps of unresistant clay, of which 
the round-worn forms of those more ancient stand evidence of fast withering by 
wind. During our work at Ghiaur Kala there was a sandstorm of such strength 
that stones 2 inches across were bounded along and smaller ones driven through 
the air with a blast of fine stuff scoured off the mound. Fortunately such storms 
are rare, for if they came often all our older mounds would long since have dis- 
appeared. Every time a desert rain falls it wears the mound away to best advan- 
tage; and lastly but not least, anything upon aggrading ground must in time be 
buried out of sight. Obviously most of the important oases were on delta plains 
and flood-plains that were always aggrading through year after year of muddy 
floods and irrigation. 
Any discussion, therefore, about ancient oases, abandoned or occupied, must 
balance the effect of erosion and natural burial upon their remains. Sometimes 
we can be sure that no growth of the plain around has taken place, and in some 
more rare instances we know there has been protection against erosion. The most 
important remains, however, have been subjected to both. The problem then 
arises, how to deal with a time-equation into which there enter three rates— 
growth of plain, growth of culture, and erosion. We believe that in a given region 
these have fairly constant values when expressed in terms of centuries. Without 
pretending to get ultimately accurate results, we can give at least a logical 
expression of the relations between these different rates through time. Indeed, 
without keeping the problem in logical proportions, we could never hope to under- 
stand either the relative rarity or distribution of really ancient remains. 
Let ¢=number of centuries since a city was founded. 
/=number of centuries it was occupied. 
G=culture growth per century. 
A =aggradation of plains per century. 
FE =erosion per century. 
h=height above plain at any time. 
If the mound is still occupied and has been from the beginning, its height 
will obviously be h=l1G—At; and since in that case /=t, h=t(G—A), or the 
