30 THE FORMING OF THE OASIS OF ANAU. 
the piedmont border, through which the stream has maintained its valley. It is 
therefore evident that the plains have been sinking relatively to the dissected 
piedmont on the border. Now, the position of the valley mouth and of the alluvial 
shore-line on the delta is dependent on two factors—the rate of sinking on the 
one hand, and the rate of aggrading, that is, the rate of accumulation of silts, on 
the other hand. 
It is also evident that as long as the rate of aggradation exceeds the rate 
of sinking, the alluvial shore-line will move towards the apex of the delta, carrying 
the valley mouth with it, and will tend to bury the dissected border; and that 
when the rate of aggrading falls short of the rate of sinking of the plains, the 
dissected border will tend to widen; that is, the piedmont valley will be prolonged 
outward, as the alluvial shore-line retreats to the foot-plain of the delta or to 
the dune-locked basins beyond. 
It so happens that the Anau kurgans, standing as they do on the side of a 
valley that has been repeatedly aggraded and dissected, were critically situated 
in their relation to the zone of sinking on the one hand, and to the dissected 
piedmont on the other. Thus, the area occupied by the kurgans belongs potentially 
to the zone of marginal deformation on the one hand, and to the zone of sinking 
and of secularly maintained alluviation on the other hand. 
Let us now apply these principles to an interpretation of the facts observed 
in the shafts at Anau and their bearing on the history of the successive ancient 
settlements. Obviously the valley on which the North Kurgan was founded 
was cut down when the zone of depression was sinking at a faster rate than the 
rate of alluviation. Conversely, the refilling of this valley is indicative of a period 
during which the surface of secular alluviation rose more rapidly by aggradation 
than the rate at which the accumulating strata beneath were sinking. In a later 
chapter, in treating of the cyclical character of these aggradings and dissectings 
of the valley and their relation to the changes of cultures, I have given reasons 
for assigning them to climatic changes causing secular fluctuations in the precipi- 
tation over the mountains of the highlands. In future we shall reason on the 
hypothesis that each cutting-down of the valleys represents a period of abnormally 
low precipitation, while the refilling represents a reversion to greater precipitation 
with more rapid alluviation. 
The information that was obtained in the shafts is brought together on plate 4, 
in which the essential facts are represented correctly in the vertical scale, while 
the positions of the valley walls are necessarily idealized. During a dry period 
preceding the founding of the North Kurgan, a valley had been cut in the delta- 
plain, the surface of which dated from loess-forming time. Then came a period of 
increased precipitation, during which the valley was refilling, throughout the life 
of the oldest culture and into that of the second. During part of the second 
culture—the latter part of the life of the North Kurgan—there occurred a dry 
period, during which the valley was reexcavated. When under renewed precipi- 
tation it began to refill again, the South Kurgan was started on the west side 
of the valley on the original loess-plain. This growth of sediments continued 
till it rose higher than the previous aggrading, overflowing not only the terrace 
