1448 
Over 1915 the strength of the Trade during Summer was about 
3.27 Beaufort-units, the 17 years average amounts to 2.90; so that 
as far as the rather small number of years permits to make a 
prognostication, the following forecast was made: 
A mild winter in Holland, Germany, Austria-Hungary, Switzer- 
land, North-Franee, Belgium, Denmark, South-Scandinavia, EKast- 
England and West-Russia; in Iceland, Greenland, North-Seandinavia 
and North-Russia a co/d winter. 
Generally this prognostication was a success with rather large 
departures in temperature. 
5. It is a well-known fact that for a number of years more than 
once large, even larger, correlationfactors than the above mentioned 
were found, between deviations of various meteorological ele- 
ments, whereas during a following series the existence of a correlation 
could hardly be demonstrated. *) This proved to be the case when 
elements were considered between which not such a strong physical 
relation could be expected as between the Gulfstream and the 
European climate and this justifies the supposition that a following 
series of years will confirm the result given above. 
The best plan would-have been to repeat the same investigation 
before the year 1900 and indeed we tried to do so for the period 
1884—1899, but had to give it up. 
As far as temperature-observations were concerned there was no 
difficulty at all; the number of wind-observations also was sufficient 
and these observations were extracted and computed. They originated 
from sailing ships however and this was the reason why they 
proved to be unfit for our purpose. With feeble winds a ship 
remained for a longer time in the region in question, whereas with 
strong winds the region was quickly passed through. The result was 
a surplus of observations with lower figures, a deficit of higher 
ones; and not only the force, but also the direction of the Trade 
wind. was a factor not to be neglected. 
Till now we have not succeeded in applying a cerrection to the 
observations of sailingships so as to make them equivalent to those 
of the steamships, which practically always spent the same time to 
pass through the region and this was the reason why that part of 
the investigation, before 1900, was abandoned. 
1) Southern Hemisphere Seasonal Correlations by R. C. Mossman F‚ R. S. E, 
Reprinted from Symons’s Meteorological Magazine Vol. 48 1913. 
