820 
frequency of severe winters in central Europe, as well 
as in central and eastern North America, is increased. 
Figure 3 shows that in central Hurope, during the 
period 1755-1949, winters with a negative temperature 
Fie. 2—The average course of some meteorological ele- 
ments during a sunspot cycle. (The dashed portion of some 
curves indicates that these cases, because of the relative brev- 
ity of the interval, are based on only few years of observa- 
tions.) (a) Departure of precipitation (im mm) in central 
Europe during the midsummers, 1803-1943; (6) pressure de- 
parture (in mm Hq) in Berlin for the midsummers, 1875-1944; 
(c) departure of the pressure difference (in mm Hq) between 
the Azores and Iceland for the winters, 1866-1940; (d) departure 
of the acne between winter and summer pressures (in mm 
Hq) for 44 (Barnaul + Irkutsk) + 4 (Lahore + Karachi) for 
1881 to *1940; (e) departure of the mean annual pres- 
sure for 44 (Capetown + Adelaide) (in 1/1000 in.) for 
1865 to 1940; (f) yearly number of severe winter months 
( | departure ‘ >1.6 c) in central Europe, 1752-1947; (g) yearly 
number of cold winter months (| departure | >4. 5E) for Chi- 
cago + St. Louis, 1847-1946, (h) departure of the mean annual 
temperature in WoC for 4 (Apia + Colombo), 1890-1945. 
WEATHER FORECASTING 
departure of more than 1.8C occurred particularly 
around the time of sunspot maxima and minima. In 
central Europe, during the intervals between 0.5 yr 
before and 1.0 yr after a sunspot maximum, nine out 
BEFORE AFTER BEFORE AFTER 
MAXIMUM MAXIMUM MINIMUM MINIMUM 
) o 7) ) 
ia « Cs c 
— oe Oe oo olee OG ha See Be AeGsb eae ese Slice Fs segces 
Wi u W Ww 
> > > > 
CHA 
Ea See eo | 
ig ane ae ESN 
Gis BS NSBR 
0.0% 29.4% 
Bee Slime 
29.0% 
DEPARTURE OF WINTER TEMPERATURES 
fA 0.0 OR ABOVE NORMAL ffJ-1.9 To -2.9°c 
fe] -0.1 To -1.8 °G -3.0 T0 -6.0°C 
Fie. 3.—The position of the winter (January) with respect 
to the nearest sunspot extreme and the temperature character- 
istics of the winter in central Europe for the period 1755-1949 
(for lack of space the individual years have not been identi- 
fied). The frequencies below the abscissa refer to winters with 
negative temperature departures of more than 1.8C. 
of thirty-one winters (29 per cent) were too cold by 
more than 1.8C, whereas only one such winter (2.5 
per cent) occurred in the mtervals from 1.1 yr after a 
sunspot maximum to 3.6 yr before a minimum. The 
probability that this difference in the relative frequen- 
cies is due to chance (assuming a constant fundamental 
probability) amounts to 0.0018 (computed according 
to R. A. Fisher’s H-method). This probability is be- 
low the residual probability of the usual limits of the 
chance range. Thus a physical relationship must be 
assumed to exist here. 
It has been shown in a similar manner that the rela- 
tive frequencies of various Grosswetter phenomena in 
different seasons and in different parts of the world ex- 
hibit fluctuations within the solar cycle, a fact incom- 
patible with the assumption of a constant fundamental 
probability. For this reason, there can no longer be any 
doubt that the Grosswetter is dependent on the solar 
cycle [18; 19; 20, pp. 962-966]. It must be emphasized, 
however, that this dependency does not appear in the 
primitive form of a linear relationship with sunspots. 
Relationships between Grosswetter and Variations in 
Solar Radiation. It seems surprising, at first, that pre- 
cipitation and temperature in the tropics show mainly 
a single oscillation within a solar cycle, whereas in the 
temperate zones, a double or higher multiple oscilla- 
tion can be observed. This can, however, be easily ex- 
plained if we consider that, on the one hand, the sun 
can exert an influence upon the weather only by way of 
radiation and that, on the other hand, the sun’s radia- 
EE 
