10
-
extreme north of the island, are omitted. It seems likely that the populations of these villages most of which are rather small were combined with the populations of the nearest market, port, or major village. In most cases the market, port, or major village was where the police post was from which the census was being conducted. Thus, the populations of the missing villages are probably buried in the figures recorded for Tai O, Sheung Ling Pei, Shap Long, Cheung Chau, and Ma Wan.
This is certainly what happened at Tsuen Wan and Kowloon City. In Tsuen Wan, populations are recorded only for Tsing Yi, Tsuen Wan, Ma Wan, Chai Wan Kok, and Kwai Chung.1 Clearly, all the Tsing Yi villages are lumped together, as are all the Kwai Chung villages. Equally clearly, the Tsuen Wan villages - with the odd exception of Chai Wan Kok - are combined in a single entry with Tsuen Wan Market. In Kowloon City district, none of the central Kowloon villages (i.e. the very important villages of Nga Tsin Wai and Po Kong and the smaller villages such as Chuk Yuen) are entered separately - their populations are, clearly, subsumed under the entry for Kowloon City.1 In part, the lack of detail in the Kowloon City census may be due to the heavy rain which interfered with the first attempt to hold it.
Thus, when conducting detailed analyses of the tables of statistics in the 1911 Census, it is necessary to bear in mind that the populations recorded for the towns and major villages in the south of the New Territories are inflated to some degree, and their social characteristics are likely to be obscured, at least in part.
The villages still existing on Hong Kong Island and Old Kowloon in 1911 are separately recorded. Po Toi Island is included under the Hong Kong villages.1
The process of holding the house-to-house enumerator visits lasted “a few days” on Lamma, and three months in the bigger districts.3 Assuming Lamma was completed in five days, and the largest districts (Au Tau, Sha Tau Kok, Ping Shan, and Sai Kung) required 50-60 working days, the average population enumerated each day varied between 143 and 181, with between one and four villages being dealt with each day.1 This is clearly not excessive, and, again, suggests that the statistics produced should be treated as reasonably accurate.