A—MATHEMATICAL AND PHYSICAL SCIENCES 41 
have been made to provide for the general demand. In an article in 
1927, by C. F. Brooks, we read that in the absence of forecasts ‘ western 
farmers have paid a “ rainmaker ” thousands of dollars at a time ’ actually 
to produce rain ; that during the previous ten years ‘ well over fifty long- 
rangers of greater or lesser repute have been publishing and, in a great 
many cases, accepting money for worthless or damaging forecasts.’ As 
in Europe, they have predictions based on occurrences on critical days, 
such as Candlemas or St. Swithin’s, as well as on the doings of animals 
and birds. Thus Brooks quotes from an almanac of 1870: ‘ When 
you see 13 geese walking injun file and toeing in you can deliberately 
bet yure last surviving dollar on a hard winter, and grate fluktuousness 
during the next seazon in the price of cowhide boots.’ 
Undeterred by the difficulties, G. F. McEwen, of the Scripps Institution 
A 
NEW F OUN DLAND }CEBERGS. MARCH —-TULY. |, 
PUM ALIV TEV UELIVTTETHAQVTU 
N 
Fic. 13.—Atlantic Icebergs and the previous oscillation. 
of Oceanography in California, has for some time been forecasting rainfall 
by empirical methods, and at first attained considerable success, largely 
on the basis of a short series of ocean temperatures. These, however, 
as he has recognised, have not of late made good their early promise ; 
and he is driven to using sunspot numbers, a cycle of five or six 
years, and a complex method of smoothing in the hope of attaining 
reliability. 
A less difficult task confronts the International Ice Patrol Service of 
the United States in their desire to obtain advance information of the 
amount of Arctic ice drifting into the western North Atlantic. I do not 
know what progress has been made, but the dependence on the previous 
North Atlantic oscillation, with which there is a coefficient of 0:60, would 
appear to suggest a useful starting-point (Fig. 13). 
In Europe the only seasonal forecasts known to me that have a scientific 
foundation, and have been made for anumber of years, are those of Sweden 
and Russia. In Sweden Wallén has for eighteen years made predictions 
for rainfall and for the height of water. Regarding rainfall, he smoothes 
by taking the sums of consecutive twelve months ; and then, assuming 
C2 
