of a system of winds of very great geographical extent, and this is probably one reason why the barometric fluctuations at the mountain stations are less than they are in the Mississippi Valley. The oscillations of the barometer are generally somewhat greater at Salt Lake City than at San Francisco, and generally they become very much magnified after crossing the Rocky Mountains. The depression of the barometer at the center of a great storm is mainly due to the geographical extent of the system of winds set in motion; and after a storm-center has reached the Mississippi River, there are no mountain barriers to prevent the formation of a system of circulating winds over an area 2,000 miles in diameter.

Areas of high barometer.

Each of the areas of low barometer described in my eighth paper was preceded by an area of high barometer on the east side at an average distance of about 1,000 miles, and was followed by an area of high barometer on the west side at about the same distance.

area.

In order to discover the circumstances under which areas of high barometer originate, and the relations which they bear to areas of low barometer, I selected from the published volumes of the Signal Service observations all those cases in which the barometer at any station rose above 30.65 inches. Sometimes at the same hour the barometer was above 30.65 inches at a considerable number of stations, all included within the same high In such cases, only one of the stations was employed, viz., the station at which the barometer was highest. These cases of high barometer are exhibited in the following table, in which column 1st shows the number of the high area; column 2d shows the date at which the barometer was above 30.65 inches; column 3d shows the greatest height of the barometer observed at that hour; column 4th shows the name of the station at which the given height was observed; column 5th shows the temperature at the given station at the given hour; column 6th shows the highest velocity of the wind which preceded each of these areas of high barometer; column 7th indicates the region where this high area appears to have originated; the abbreviations are the same as in the preceding tables, except Ark. which denotes Arkansas; column 9th shows the station where the barometer was lowest on the east side of the high area; column 8th shows the height of the barometer at the given station; column 11th shows the station where the barometer was lowest on the west side of the high area; and column 10th shows the height of the barometer at the given station.

These cases correspond to thirty-two different areas of high barometer, and thirty of these occurred during the six months from October to March, and none occurred during the four months from May to August. More than half of the whole number occurred during the months of December and January. Areas of unusually high pressure are thus found to occur at the same season of the year as areas of unusually low pressure.

In order to show how far these cases of high pressure are dependent upon latitude and longitude, I have divided the stations into three classes; one class including the stations east of longitude 86° from Greenwich; a second class including the stations between the meridians of 86° and 102°; the third class including the stations west of the meridian of 102°.

The observations at each of these stations cover the entire period of twenty-six months, with the following exceptions, viz., the observations at Pembina commenced in November, 1872; those at Halifax, in December, 1872; at Yankton and Fort Gibson, in April, 1873; at Chatham, in October, 1873; at Cape Rosier, Father Point and Sydney, in November, 1873; and at Brockville, in January, 1874. We see that cases of high barometer occur most frequently at the northern stations. About two-thirds of the whole number were north of latitude 44°, and only fifteen per cent of the whole number were south of latitude 40°. Only one of the cases occurred beyond the Rocky Mountains; and on the east side of the mountains high barometer is most frequent near the meridian 97°. After making allowance for the unequal period of observation at the different stations, we find that in the neighborhood of Dakota, cases of very high pressure are nearly twice as frequent as they are at a corresponding latitude near the Atlantic coast. This fact appears still more palpable if we make the comparison for pressures as high as 30-9 inches. In the region between longitude 86° and 102° the barometer rose as high as 30.9 inches in fifteen cases, while only one such case occurred in the district east of longitude 86°, and none occurred in the district west of longitude 102°.

The low temperature generally attending these high pressures is very remarkable. The lowest temperature observed in any case was 36° Fahrenheit at LaCrosse. In seven cases the ther