484 DOVE ON THE PERIODICAL VARIATIONS 
in it, the assumption, that the region of calms has here tempo- 
rally approached the Arctic region in a manner which does not 
occur in any other part of the globe, will no longer appear daring. 
Again, bearing in mind that Christiansborg, on the coast of 
Guinea, belongs to the southern hemisphere, in its barometrical 
relations, that according to Burnes (Cabool, App. 318) south 
winds prevail from May to September in the valley of the Indus, 
that in Chusan the northern monsoon only sets in in October, 
while during the summer months calms appear to predominate 
rather than a real south monsoon; when we see that the baro- 
metrical oscillation, decreasing in Seringapatam, Katmandu, 
Mussuree, Kotgurh, Simla, Darjiling, in the high land in Hin- 
dostan generally (Repertorium, iv. p. 236), again attains the same 
magnitude in the steppes of Barabinsk as in the low land of the 
Ganges ; when the rainy season in southern Siberia occurs at 
the same period as on the southern declivity of the Himalaya, 
without being interrupted by winter rains, as is the case at the ex- 
terior limits of the trade-wind,—the whole of these phenomena 
are more simply explained by admitting the occurrence of an 
ascending current of air over south and central Asia during the 
summer. If this region of diminished atmospheric pressure 
forms a point of attraction for neighbouring masses of air, it 
appears with all the characters of the region of calms, which 
consequently rotates during the annual period about a more or 
less stationary point, situated in America, in such a manner that 
at its greatest elongation to the north in the summer it reaches 
into Asia, but in the winter recedes to the south, followed im- 
mediately by the north-east monsoon, which re-establishes the 
normal relations. 
This change reacts very considerably on the relations of the 
weather in Europe. The axis of the thermal wind-rose, which in 
the winter connects the warmest point at south-west with the 
coldest at north-east, changes ’to such an extent, that in the 
summer, on the contrary, the coldest point falls in a western, 
the warmest in an eastern direction. While in the winter the 
barometrical extremes of the wind-rose coincide with its extremes 
of temperature, this no longer takes place in the summer; on 
this account the barometer is a far more certain guide for jud- 
ging of the weather in the winter than in the summer. More- 
over, the barometrical wind-roses of more easterly localities will 
necessarily be less regular than those of western places. It is 
