The influence of the winds is likewise of considerable importance in determining the greater or lesser severity of an Arctic climate. Thus the northerly winds which prevail in Baffin’s Bay and Davis’s Straits during the summer months, and fill the straits of the American north-eastern Archipelago with ice, are probably the main cause of the abnormal depression of temperature in that quarter; while, on the contrary, the southerly winds that prevail during summer in the valley of the Mackenzie tend greatly to extend the forest of that favored region nearly down to the shores of the Arctic Sea. Even in the depth of a Siberian winter, a sudden change of wind is able to raise the thermometer from a mercury-congealing cold to a temperature above the freezing-point of water, and a warm wind has been known to cause rain to fall in Spitzbergen in the month of January. The voyages of Kane and Belcher have made us acquainted with the lowest temperatures ever felt by man. On Feb. 5, 1854, while the former was wintering in Smith’s Sound (78° 37’ N. lat.), the mean of his best spirit-thermometer showed the unexampled temperature of -68° or 100° below the freezing-point of water. Then chloric ether became solid, and carefully prepared chloroform exhibited a granular pellicle on its surface. The exhalations from the skin invested the exposed or partially clad parts with a wreath of vapor. The air had a perceptible pungency upon inspiration, and every one, as it were involuntarily, breathed guardedly with compressed lips. About the same time (February 9 and 10, 1854), Sir E. Belcher experienced a cold of -55° in Wellington Channel (75° 31’ N.), and the still lower temperature of -62° on January 13, 1853, in Northumberland Sound (76° 52’ N.). Whymper, on December 6, 1866, experienced -58° at Nulatto, Alaska (64° 42’ N.). Whether the temperature of the air descends still lower on advancing toward the pole, or whether these extreme degrees of cold are not sometimes surpassed in those mountainous regions of the north which, though seen, have never yet been explored, is of course an undecided question: so much is certain, that the observations hitherto made during the winter of the Arctic regions have been limited to too short a time, and are too few in number, to enable us to determine with any degree of certainty those points where the greatest cold prevails. All we know is, that beyond the Arctic Circle, and eight or ten degrees farther to the south in the interior of the continents of Asia and America, the average temperature of the winter generally ranges from -20° to -30°, or even lower, and for a great part of the year is able to convert mercury into a solid body.