SECTION A.—_MATHEMATICAL AND PHYSICAL SCIENCES, 
SEASONAL WEATHER AND ITS 
PREDICTION 
ADDRESS BY 
PROF, SIR GILBERT -T. WALKER, C.S.I., FR.S., 
PRESIDENT OF THE SECTION. 
I nave chosen the subject of seasonal weather for my address, because 
its economic importance is obvious to most men who have lived in the 
Tropics, and its scientific problems are full of interest. Unfortunately 
there is an additional motive, the need of warning against dangers ahead. 
For the difficulties of long-range forecasting are not in general adequately 
recognised, so that some of the most progressive countries in the world 
are inclined to make predictions on an insecure basis ; their technical 
staff does not realise that though the prestige of meteorology may be 
raised for a few years by the issue of seasonal forecasts, the harm done 
to the science will inevitably outweigh the good if the prophecies are 
found unreliable. We only learn from experience that while the fore- 
casting efforts of a charlatan are judged by their occasional successes, it 
is the occasional failures of a government department which are remem- 
bered against it. 
In a country where conditions are as changeable from day to day as 
they are here, it is natural that we should think in terms of wet or fine 
days rather than of wet or dry periods ; but in the greater part of our 
empire the different seasons are much more sharply defined, and so 
their dominant features stand out more clearly. Also the variability of 
their seasons is in general materially greater than here. Thus in the 
annual rainfall measurements of the last half-century the smallest rainfall 
of Great Britain has been 23 per cent. below normal ; but that of large 
areas in South Africa has been in defect by 40 per cent., in north-east 
Australia by 50 per cent., and in the Punjab by as much as 58 per cent., 
or two and a half times that of this country. 
Now a season that is unusual seems to have some abnormal factor 
permanently at work diverting the weather from its ordinary course ; 
in India I found, when issuing the daily forecast in a dry winter, that I had 
at times to predict no rain, when with identical conditions as shown by 
the weather map I should in a wet winter have predicted a widespread 
fall. Even in England, in winter, there is an appreciable persistence 
in the characteristics: during the last sixty years the fifteen wettest 
Januaries were followed by Februaries of more than average rainfall in 
