HONG KONG BEFORE THE CHINESE
45
is one important point to be cleared up. The Chinese are highly skilled farmers. Their techniques of land-winning and of irrigation change landscapes. So, alas, does their age-long war against trees. But since A.D. 900 the topography of this territory has been changed not only by human technique. There has also been a gradual, small, but identifiable and, I believe, measurable tilt of the surface of the earth along the axis of the four high peaks (the two on Lantao,37 Tai Mo Shan and Ng Tung Shan104) which has altered and is still altering the coast line. I leave it to geologists to say whether this is a necessary effect of what happens when the subsidence of a long straight shore meets a range of hills parallel to the shore (in which case it will be reproduced at many points of the Chinese coast), or whether it is a local peculiarity. It would also be interesting to fill in some of the chronological gaps and find out whether the two clear cases of recent river capture13 took place before or after the Chinese settlement. Until these gaps are filled up, I do not claim that the details of the shore line indicated on the map are authoritative, but they are not far wrong for the northwestern part of the territory, which was the part first settled by the ancestors of the Man94 and Tang.44
You will observe that the present Castle Peak and the mountain attached to it on the north42 were at that time an island, separated from the mainland of the New Territories by a sea channel which in A.D. 900 was probably very shallow but navigable. The traditions of the oldest villages leave no room for doubt that there has been a general uplift in excess of 5 metres in this area. The red line approximately follows the present 5 metres contour. The ground on both sides of the navigable channel was swamp, probably mangrove swamp, dotted about with small islands and intersected by creeks and streams. The first fort of which there is written record was known as Tuen Mun Chan141 and was almost certainly located at a point I have marked on the map,138 about three miles north of the present location called Tuen Mun.141 It would be an advantage if all doubts could be settled by excavation on the site, which can be seen even from the ground (and more clearly still from the air) to have contained old earth-works and possibly buildings.
It will be noticed that the present Sham Chun120 River had a much shorter course at that date, and the northern half of what