88. REPORT—1848. 
reckon all places where the winter is too warm and the summer too cool, as 
belonging to a sea climate ;—and all places where the winter is too cold and 
the summer too hot, as belonging to a continental climate ;—the thermic nor- 
mals will give the baundary lines between these two species of climate. An 
inspection of the maps of the several months will show whether a place 
belongs always to one or other of the above classes, or whether it changes 
its character in this respect in the course of the year. 
The greatest winter cold is known to fall in North Asia and North America: 
on examining the map for January, we see that these two coldest localities form 
a connected cold region ; for the thermic normals by which they are bounded 
pass, on the Pacific side, along the west coast of America and the east coast 
of Asia, and unite in Behring’s Straits; and on the Atlantic side, when they 
can be traced no further towards the north, they point exactly to the pole. 
Now it is in the nature of things that a thermic normal must pass through 
the pole, for as that point includes in itself all degrees of longitude, it must 
necessarily correspond to the definition of a point of normal temperature. 
The whole of Europe is included in January in the warm space, for the 
thermic normal coincides almost exactly with the boundary between Europe 
and Asia; Greenland is also included, but of North America only the narrow 
strip of coast on the Pacific, beyond the Rocky Mountains. In the tropical 
region the sea is everywhere warmer in winter, therefore the interior of 
Africa forms an insulated cool space in opposition to the warm West Indies 
and the coasts and islands of the Pacific and Indian Oceans. Java and the 
Sunda Islands have at that season, as compared with the West Indies and 
Polynesia, a continental climate. We see therefore that these names are un- 
suitable when comparing places under different latitudes ; for it would sound 
strange to say that Moscow has a sea climate, and that Singapore and Batavia 
have a continental climate. 
In conformity with the shape of the cold spaces, all the January isothermals 
have their longer axes in a line from America to Asia, passing from the 
middle of North America beyond the pole to Mandschury. 
The terrible January cold of Yakutsk is not corresponded to by any 
equally cold point in North America. If therefore we assume for this month 
two poles or maxima of cold, we must assign to them different intensities. 
But this is not necessary ; the course of the curves appears rather to indicate 
a connected narrow tract from Yakutsk to New Siberia. 
But it may be said, how is it possible that if the isothermals for the whole 
year curve round two separate poles of cold, these poles should not also ap- 
pear in the several portions of the year? It may be observed in reply, first, 
that the examination cannot be pursued into the higher latitudes for all the 
months of the year with equal exactness ; but that besides, that may be true 
on the annual mean which yet has no reality at any single portion of the year. 
The following example will illustrate this. 
A mass of Jand within the tropics, when the sun is vertical, so increases 
the heat, that under these circumstances continental stations show tempera- 
tures such as are never met with at sea. Now although these continental 
masses are cooler than the sea in the winter, yet this cooling is less than 
the disproportionate heating before spoken of, and the mean of the whole 
year is therefore above the normal. ‘Thus the greater breadth of Africa 
north of the equator, and the expanse of India, cause the line of maxi- 
mum mean annual temperature not to coincide with the equator but to run 
north of that line. We will now suppose the imaginary case of two belts of 
tropical land at equal distances on each side of the equator, which latter shall 
