ON THE MONTHLY ISOTHERMAL LINES OF THE GLOBE. 93 
with increasing elevation than is shown by the observations at St. Helena, 
delago of the Low Islands. The reason of this great inflection of the isother- 
\ shall still find the temperature there much lower than that of the Archi- 
e 
a 
al lines having been hitherto overlooked, is probably that navigators usually 
‘eep nearer either the American or the African coast, and that thus the 
outhern tropic is rarely crossed in the Atlantic in mid-ocean. 
To the south of the Cape of Good Hope, the isothermals flatten and are 
uch crowded: this crowding is still more striking in March, when the 
sothermals of the torrid zone have their concavities, and those of the tempe- 
ate and frigid zones their convexities, in the meridian of the Cape. 
It has only been possible to determine directly the position of the line of 
i° Reaumur, or $2° Fahrenheit, in the southern hemisphere, for the four 
yionths of December, January, February, and March. It is comparatively 
mut little inflected ; these determinations however can only be regarded as 
pproximate, if we consider that the drift-ice of the Antarctic regions, 
eing everywhere exposed to the uninterrupted action of an open ocean, 
Ithough it may consist of more compact masses, yet ‘can never form into 
uch extensive fields as the ice of the northern seas, and from its state of 
lisruption is far more variable in its place in different years. The boldness 
with which Captain Ross broke through the zone of moveable ice which he 
net with in the place where Dumont D’Urville had found an open sea, was 
‘ecompensed by finding beyond it a sea free from ice, which permitted him 
© advance to farhigher latitudes ; from a comparison of the different voyages, 
ye arrive however at a conviction, that before reaching the barrier of fixed ice 
hhe temperature dependent on the position of the moveable ice may vary very 
considerably in different years. Ifwe were enabled to sketch the isother- 
mals of a year, we might perhaps find an increase of temperature beyond the 
noveable belt of drift-ice. By the combination of the results of different 
single years, there may appear a local curvature which in the mean of many 
years would soften off into simpler forms. We may explain in this manner 
the apparently contradictory statements of different circumnavigators on the 
‘emperature of the southern hemisphere. From not being acquainted with 
the part to be attributed to non-periodic variations, the observed atmo- 
spheric relations on each occasion have been regarded as the normal ones. 
_ |t was overlooked that a traveller visiting Berlin in January 1823, would have 
found there the mean January temperature of Godthaab (in Greenland), of 
Bear Island, and of Moscow; and in January 1834, a temperature higher 
than the mean January temperature of the plains of Lombardy. 
In February the isothermals in Northern Asia begin to move northwards, 
while in North America they are still moving southwards. In Baffin’s and 
Hudson’s Bay they become still steeper than before, while in Siberia they 
begin to flatten. Near the thermal limits between the northern and southern 
hemispheres, the temperature of 22° Reaumur, or 814° Fahrenheit, is found in 
two separate spaces, one in the interior of South America, the other in Cen- 
tral Africa, where it extends to Australia, the larger part being in the southern 
hemisphere, but in Guinea extending to 10° north of the equator. In the 
southern hemisphere the distribution has altered but little ; in Australia the 
‘ ee and west sides continue to be cooler than the middle until after the 
ginning of March. 
___ In March the spaces in America and Africa, enclosed by the isothermal of 
_ 22°, or 814° Fahrenheit, have united ; the inflection in the middle of the At- 
- Jantic still recalls their separation in February. The flattening of the Asiatic 
 eurves has become still more decided, and shows itself unmistakeably in the 
