86
ARMANDO M. DA SILVA
defected to the government cause, and that as a reward, their land holdings were recognized officially by the government. This is a very Chinese approach to the problem of pacification. The Cheng 鄭 family of Fan Lau claims to have ancestral connections with Cheng Lin Fuk 鄭連福 and his son, Cheng Yat 鄭一, both notorious pirates from Tai Yu Shan, who terrorized the Chu Kong estuary during the latter half of the 18th century. The Cheng family still owns the land nearest to the old fort, which may suggest that this family had ancestors who were also on the government side (plate 10). The garrison could not have existed for long without food and it is reasonable to suppose that the padi fields of Fan Lau supported the soldiers from the fort (plate 11).
There are reasons for believing that the Kai Yik Kok fort may have pre-dated the Coastal Withdrawal of 1662, and that it may have been a Ming rather than a Ch'ing fort. Some confirmation of this is afforded by a series of nautical charts in the Mo Pei Chi (A). The preface to this work is dated 1621, but it was not presented to the throne until 1628. However, it has been shown that the charts almost certainly date from the first half of the fifteenth century.
Many of the place-names in that section of the charts pertaining to the Chu Kong estuary are identifiable when checked against similar or equivalent place-names found in the maps of the 19th century editions of the Kwong Tung Tung Chi, San On Yuen Chi, Heung Shan Yuen Chi and O Mun Kei Leuk, but the reader must be warned on two points. First, place-names may differ in both pronunciation and orthography in different sources. Yung Hai is written as 容海 on the Mo Pei Chi charts, but as 雍海 on the maps of the Kwong Tung Tung Chi. A second point to remember is that adjoining districts on one island are not infrequently depicted as separate islands. The Kwong Tung T'ung Chi carries a map of the San On district, for instance, which marks Tai Yu Shan, Tung Chung and Kai Yik Kok fort as separate islands, whereas the last two places are in fact both located on Tai Yu Shan. It is obvious that the place-names on these maps serve not so much to pin-point localities as to mark well-known landmarks and stopping places. Navigation in these waters depended not on nautical instruments, but on the experience of pilots familiar with key channels and navigational landmarks, such as headlands and mountain peaks.
*Plates 12 and 13 also relate to this article.