96 REPORT— 1848. 
causes render September, as has been shown in the memoirs on the “ Non- 
periodic Variations,” the month which shows the fewest anomalies in single 
years ; for when the temperature is equally distributed in the east and west 
direction, easterly and westerly winds or currents of air cease to exert any 
disturbing thermal effect. Hence we prefer September as a travelling month, 
and our after-summer, though less beautiful than that of America, is not with- 
outcharms. Nature falls gently asleep in autumn, and awakens with feverish 
starts in spring ; if the last-named season were not set off by winter as a foil, 
autumn would surely stand the higher in our estimation. 
Within the tropics the temperature begins already to sink, showing clearly 
that as the sun passes from the northern to the southern signs, a larger por- 
tion of the heat dispensed by his rays becomes latent. The West India Islands 
are now withdrawn from the space enclosed by an isothermal of 22° Reau- 
mur, or 812° Fahrenheit, which has now contracted to a narrow strip of coast 
from Vera Cruz to Cayenne; the space included by the same fSothermal in 
Africa has retreated from the west coast to the interior ; and the space en- 
closed by an isothermal of 24° Reaumur, or 86° Fahrenheit, now only in- 
cludes Kordofan, Nubia, and Arabia, and no longer embraces Hindostan. 
In October it begins to disappear. Thecold nowcomes in decidedly from the 
north ; at the mouth of the Yana the isothermal of — 22° Reaumur, or — 172° 
Fahrenheit, already touches the continent of Asia, and the temperature of 
Melville Island has sunk to — 16° Reaumur, or — 4°°8 Fahrenheit. The cold 
comes in the old continent from the north-east, and in the new continent 
from the north-west. But it is not until November that the isothermals be- 
come in both continents decidedly concave. At the same time the curves in 
the southern hemisphere become increasingly inflected as the increasing alti- 
tude of the sun, causing the ice to melt, renders the differenee between land 
and sea more marked. 
The isothermals of the torrid zone north of the equator, on the contrary, 
run almost completely in the direction of the parallels of latitude. In Europe 
meanwhile those extraordinary involutions have already begun which in 
December are still more decidedly formed, and which cause the isothermal of 
4° Reaumur, or 41° Fahrenheit, to run from the Feroe Islands to Rochelle, 
passing along the west coast of Great Britain. In a similar manner the 
south point of Nova Zembla and the Kirghis Steppe have now the same 
temperature. In the curves of December we recognise already almost the 
extreme forms of January. 
Such important variations in the distribution of temperature cannot but 
react in the strongest manner on the movements of the atmosphere, and 
consequently also on the distribution of the atmospheric pressure. Graphi- 
cal representations of a fresh and more detailed examination of the annual 
variations of the pressures of the gaseous and aqueous atmospheres which 
are now lying before me, show that the interchange between masses of air _ 
does not only take place between the northern and southern hemispheres, but 
that a lateral flowing off also takes place at certain times. Thus in thespring 
of the year the air accumulates over the part of America where the cold . 
still continues; while in Asia the increasing warmth already causes it to ex- _ 
pand so that its amount and pressure are diminished. Hence the countries - 
which have a cold spring (as the Arctic regions of North America), have the — 
maximum pressure in the spring, as the author showed fifteen years ago 
(Pogg. Ann. xxiv. p. 205); hence the west coasts of America have the 
maximum of pressure in summer ; and the interior of Asia, on the contrary, — 
has its minimum of pressure in summer, as the longitudinal axis of the isother- _ 
( et ae 
