Sec. v] TRANSPORT PROBLEMS OF THE ANCIENT LOU-LAN ROUTE 337
SECTION V.—TRANSPORT PROBLEMS OF THE ANCIENT LOU-LAN ROUTE
Having now completed our survey of the whole of the desert ground which the ancient Han Traffic
route crossed between the Tun-huang Limes and Lou-lan, we may now in conclusion briefly review conditionsonHan
the conditions under which traffic is likely to have been maintained along it. We may conveniently route.
take the east as point of departure ; for it was undoubtedly from the side of China that came most
of the organization and of the resources that were needed for the maintenance of regular commu-
nication on a route beset with such great physical difficulties.
The problem of assuring these resources for the troops and convoys moving along the route Start from
to the ` Western Countries ' must always have demanded much care on the part of the Chinese Limest
administration. Of this we have direct archaeological evidence in the imposing ruins, situated Magazine,
on the line of the Limes to the east of the ancient Yü-mên and marked by me T. xviII, which my T. xvIlr.
explorations and finds of 1907 have definitely shown to be those of an ancient Chinese magazine.'
I have fully explained in Serindia the advantages which such an advanced base of supplies must
have offered for the victualling of military expeditions, political missions, convoys, &c., whether
going to or coming from Lou-lan.2 The base itself could easily be reached by three marches
from the town of Tun-huang.
On the two first marches beyond it, which led along the Limes line to its extreme western Difficulties
point, T. iv. a, near Toghrak-bulak, two essential items of supply, viz. water and such forage as plenti- beyond
ful reed-beds and scrub can offer, were, no doubt, as conveniently obtainable in ancient times as Toghrak-
they are at present. In addition, the ground, mostly hard gravel ` Sai ', offered easy going for bulak.
laden animals and carts. Conditions were not so favourable on the two following marches, which,
as we have seen, crossed the extremity of the ` Three Ridges Sands ' and the ancient lacustrine
basin, before bringing the traveller to the vicinity of the present Bésh-toghrak. The ridges of
drift-sand encountered before reaching the latter point would necessarily present difficulties,
especially for carts. But they are not likely to have ever been more formidable than the similar
difficulties which traffic still encounters and somehow overcomes on the present ` highways ' of
Chinese Turkestan and westernmost Kan-su.3 Water, too, was then probably more accessible
than now at the eastern extremity of the old lacustrine basin.
From the vicinity of Bésh-toghrak, where, as we have seen, the ` Chü-lu granary ' of the Wei Route along
ho itinerary may with some probability be located,4 the route would inevitably lead along the N.S side of
northern side of the valley. Drinkable water from wells and some grazing on reeds and scrub were tog
brings us to the vicinity of the point marked by our Camp cvi where vegetation at present ceases.
It is somewhere near this point that, as explained before, the position of the Sha-hsi well ' may
with good reason be looked fors Considering how closely the salt-encrusted bed of the ancient
sea beyond it approaches the foot of the cliffs marking the old shore-line, I think it very improbable
that any appreciable quantity of vegetation could have been found beyond this point any more in
Han times than now.
It seems safe to assume that, so far, the provision of water for men and beasts and of forage
for transport animals could not have offered more serious difficulties in ancient times than are to
' See Serindia, ii. pp. 712 sqq. ; Map No. 35. D. 4. formidable ridges of sand were surmounted by our carts at
2 See ibid., ii. p. 715. two points on the way from Kan-chou to Mao-mei ; see
3 Thus, e.g., belts of drift-sand are crossed by carts on Map No. 43. D. I.
the high road ' from Yangi-hissar to Kashgar, Map No. 5. 4 Cf. above, p. 308.
A. 2 ; from Yarkand to Maral-bashi, No. 5. c. 3 ; west of the 5 See above, p. 309.
Khotan oas s near Kum-rabat-padshahim, No. 9. c. 2. Very