330 SECTIONAL TRANSACTIONS.—E. 
archeological and historical importance. Growing tendency for men of 
history and science to show distribution and evidence upon maps : increase 
of knowledge and growing interest of general public in scientific and historical 
subjects demands more room and clearer illustration ; hence gradual develop- 
ment from using very small scale maps or sketches in periodicals and 
professional papers to the use of larger scale national maps for the purpose. 
The national cartographic institutes well equipped for such work. All 
necessary material available in many forms and cartographic experience 
available to ensure best use of material. ‘These institutions have advantage 
also of getting in touch with colleagues of other nations if the theme be 
international. How the Ordnance Survey organises research into and 
collection of old plans, maps and other such material. Research into the 
past through the medium of air photography. 
Examples of national maps as backgrounds: Physical Map of Great 
Britain (1/M); Land Utilisation (1-in.); Population (1/M); Neolithic 
Wessex (4-in.) ; Roman Britain (1/M); Celtic Fields (1/25,000). 
AFTERNOON. 
Mr. W. SmitH.—A preliminary study of the rainfall of China (2.0). 
The paper is an analysis of the rainfall data, in the form of 35-year averages 
and corrected to a 30-day month, for some eighteen stations. These are 
distributed along the China Coast and the Yang-tze Valley, and they provide 
a framework of ‘ normals’ to which the short period data in the interstices 
will later be related. The 35-year period taken, 1895 to 1929, is shown to 
be one complete rainfall cycle. 
Certain rainfall provinces are distinguished and their seasonal distribution 
of rainfall analysed. The rainfall seasons distinguished are three : 
1. A winter monsoon from October to February or March. Rains are 
scanty in North China (% in. per month) but not inconsiderable elsewhere, 
and greatest in the lower Yang-tze Valley (2 in. per month). 
2. A pre-summer monsoon season from March or April until the beginning 
of the summer monsoon (June in the Yang-tze Valley and South China, 
July in North China). Rains are still low in North China (1 in. per month), 
but elsewhere considerable (4 in. per month and over), increasing rapidly 
in April and May. 
3. A summer monsoon from June or July to September. The front of 
the monsoon is usually the period of maximum precipitation, and the rains 
fall away month by month from this peak. Along the South China Coast, 
however, they increase again in August and September. 
In conclusion an attempt is made to relate the results of this analysis to 
the atmospheric circulation of the Far East, and to construct a rationale of the 
rainfall seasons distinguished. 
Mr. P. R. Crowe.—Rainfall probability, with special reference to the High 
Plains region of U.S.A. (2.30). 
The traditional method of summarising monthly rainfall data has been 
by means of the arithmetic average, and the recent history of geography 
points to increasing dependence upon data arrived at in this fashion. The 
fundamental faults of this system are: (1) that the average is influenced 
too greatly by occasional very high records, (2) that the inherent variability 
of rainfall is obscured, and (3) that hence we have no indication of the range 
within which differences between averages must be regarded as insignificant 
or fortuitous. 
