DISTRIBUTION OF THE ‘ANCIENT POPULATION OF MERV OASIS. 22:0) 
extended watercourses, while strips of desert lay between. In the case of a seden- 
tary agricultural population like that which is now coming to occupy the oasis 
on the other hand, especially where, as at Merv, the water-supply is strictly limited 
and all parts of the level plain are equally fertile, the invariable tendency is to 
utilize the land lying close below the center where the distribution of the water 
begins to take place. From there outward the land will be used in a nearly con- 
tinuous body, although, of course, along the main canals detached tongues will 
stretch out into the untilled wastes beyond. 
This is what is taking place at Merv—the people are moving into the upper 
portion of the delta of the Murg-ab, and the Imperial Estate at Bairam Ali, with 
its new dams and reservoirs, is hastening the process and at the same time making 
it possible for a greater number of people to be supported by the same amount of 
water. In the course of another generation it is probable that the change from 
semi-nomadism to agriculture will have been completed, a larger population will be 
supported by the oasis, and a larger number of acres of land will be under culti- 
vation, but the outposts of civilization will have been drawn inward, and the size 
of the cultivated delta, as measured by a mere inspection of the map, will have 
decreased. It is conditions such as these, rather than those of the immediate 
transition period of to-day, which should be looked at in comparing the distribu- 
tion and magnitude of the ancient population with that of the present; for the 
ancient times were days of intensive agriculture, when, as the ruins show, the people 
lived close together, and most of the land between the villages must have been 
cultivated. 
Any attempt to compare the area of the land under cultivation at the time 
of the kurgans with that under cultivation to-day is liable to error because of the 
limited study that has been given to the question. The best that can be done 
is to indicate the general direction in which the truth seems to lie. For the sake 
of conservatism we will disregard the important differences between the transition 
state of to-day with its large remnants of semi-nomadism and the state of intensive 
agriculture of the past, and will compare the limits of irrigation in 1904 with those 
at the time of the kurgans. Such a comparison indicates that the amount of irri- 
gated land has decidedly decreased since the kurgans were abandoned. This is 
roughly shown on the accompanying sketch-map, where the limit of the present 
irrigated area is drawn through the ends of the canals without regard to the very 
extensive unoccupied lacune that lie between. The lacune are so large that 
much of the region presents the aspect of a desert in which there are mere patches 
of cultivation. The limits of the formerly irrigated areas are drawn just beyond 
the outmost ruins, but it is hardly probable that the ancient inhabitants built large 
and important villages at the extreme limit of cultivation. Outside of these places 
there must have been a strip of cultivated land where the villages were so small 
that they have left no noticeable traces. 
In view of all the facts we should probably be justified in diminishing the area 
represented as now under irrigation, and, perhaps, in extending that represented 
as under irrigation during the earlier period. Yet even as the map stands, it is 
