482 DOVE ON THE PERIODICAL VARIATIONS 
year in this series ; since the monthly means of single years differ 
less hygrometrically than barometrically. This in fact justifies 
the course adopted. There being no barometrical observations 
for Lugan, the vapour curve of that place is combined with the 
barometrical for Taganrog. From the proximity of the places of 
observation, the error committed can be but inconsiderable. The 
exponent affixed to the names of the localities indicates the 
number of years from which the means have been calculated. 
If we determine the differences of temperature of the coldest 
and warmest month in degrees of Reaumur, we find for Nert- 
chinsk 38°:3, Barnaul 34°:3, Bogoslowsk 33°1, Slatoust 35°'8, 
Catharinenburg 28°2, Kasan 27°3, Archangel 25°, Moscow 
22°-9, St. Petersburg 21°8, Wilna 18°-2, Cracow 195, Prague 
18°0, Berlin 17°°5, Regensburg 17°3, Stuttgard 17°9, Carls- 
ruhe 15°'8, Maestricht 15°°3, London 12%5, The continually in- 
creasing bend of the curve for the pressure of the dry air in ad- 
vancing from the west coasts of Europe into the interior of North 
Asia thus receives its explanation. Butif we observe the num- 
bers of Table IX. under the head oscillation, we find for the 
Asiatic stations nearly the same numbers as for the Europzan, 
by oscillation being understood the difference between the month 
with the highest and that with the lowest pressure of vapour. 
With the slight concave bend of the Europzan curves for the 
pressure of the dry air the convex bend of the curve of the vapour 
consequently preponderates ; while for the Asiatic stations of 
observation the reverse happens. The complicated problem in 
Europe is solved in a very simple manner if we refer the pressure 
of the dry air, as well as that of the vapour, to a straight axis 
of the abscissa. 
Von Humboldt first showed that near the equator the mean 
barometrical pressure on the surface of the sea is less than in the 
latitude of Paris, but Von Buch has called our attention to the 
high mean barometrical pressure near the Canary Islands. That 
between those two extremes a gradual transition occurs, that the 
pressure consequently continually decreases in approaching the 
zone of the trade-winds from the tropics to the region of calms, 
has been found by direct observations by A. Erman and Her- 
schel, and is likewise confirmed by the ships’ journals of Ryan 
and MacHardy, Quevedo, Beechey, Horner, Spencer and Lund. 
The phenomenon of the trade-winds however is not invariably 
fixed throughout the year to the same geographical latitude, but 
