SOUTH WIND.

315

could scarcely have failed in coining to the conclusion that asoutherly wind was blowing west of the Mississippi. It isaltogether difficult to imagine, on any other supposition, howsuch a high temperature could exist there on that day, seeingthe weather was so cold two days previous. We are not leftto conjecture, however, for direct observations show that anaerial current, about GOO miles in breadth, was flowing fromthe Gulf of Mexico, and raising the temperature of the air overthe territories to the west of the Mississippi. The arrowsindicating the directions of wind on the 9th, and the figuresthe temperatures, are laid down in Plate 2, at a few stationson the small chart of the weather at sunrise.

It may be here observed, also, that since the cold weatherprevails over such large tracts of the American continent, thechange from cold to warm weather must be much less suddenin the North-Western territories than from hot to cold—thenorth-west winds at once lowering the temperature. Thus,as Fort Snelling, in Minnesota, is about 1400 miles from FortBrown, in Texas, on the liio Grande, the warm air at thesouthern station could only bo translated to the northern inthirty-five hours, though it travelled at the rate of forty milesan hour. Owing to this cause, we need not wonder thatthe south winds are very cold on the Illinois prairies for thefirst day they blow. In these regions, the thermometertherefore, usually rises much slower than it falls, and it willalso be found that the barometer rises much faster than itfalls.

But, on the 10th November, some very remarkable changestook place. At Fort Snelling, in Minnesota, the mostnortherly and westerly station, snow seems to have begun tofall at 1 a.m., and continued till 8 a.m. The temperaturefell from 58° at 9 p.m. of the 9th, to 22° at sunrise of the 10th.At two stations in Iowa, west of the Mississippi, the tempera-tures also fell to 27° and to 24°. St. Louis, in Missouri, beingfarther to the eastward than Fort Snelling, the temperaturewas 46° at sunrise, but fell to 35° by 9 p.m. Though FortSmith, Arkansas, is about 700 miles distant from Fort Snelling,and about one degree of longitude further west, yet thechange of temperature and wind seem to have been simultaneous