568 GENERAL NOTES ON THE CLIMATR. 
In January, 1854, the cold was intense at Fort Benton and at St. Mary’s valley. It was still 
more severe, however, at all points near this latitude eastward as far as Fort Mackinac, and 
nearly the same at Fort Kearney, in latitude 41°. On thirteen days the thermometer was below 
zero at Fort Benton; at Fort Snelling the corresponding number of days was twenty-three, and 
at Fort Kearney fourteen; the lowest single readings being respectively 24, 36, and 16 degrees 
below zero. In February, the remaining month of winter, the mean temperature was 11° above 
that of Fort Snelling, and 4° below that of Fort Kearney. The low extremes at these posts were 
— 6°, —20°, and + 2°, respectively. The post was 4° colder than St. Mary’s for the same month. 
In each of the months of spring, nearly the same relation of temperatures was preserved to 
stations eastward. March and April were each 5° warmer at Fort Benton than at Fort Snelling, 
and April was warmer than at Fort Kearney. In May there was very little difference at the 
three posts. March only was colder than at St. Mary’s, and the mean of the three months was 
greater at Fort Benton than west of the mountains. 
The summer of 1854 had nearly the same mean temperature for each of the three posts pre- 
viously compared; that of Fort Benton being 72.8°, Fort Snelling 72.1°, and Fort Kearney 73.4°. 
St. Mary’s valley has a mean of 69.6°. The extremes of temperature range as high here, also, 
as at stations eastward on the plains, the summer maximum being 100°. 
The corresponding record at Fort Pierre first appears in July, with a mean temperature of 
76.9° against a mean of 73.6° at Fort Benton, and the differences for August and September are 
not large. The high summer temperatures of the plains at the sources of the Missouri are a 
very decided and well established feature. 
The range of temperature in successive months and for less periods has unusual characteristics 
at Fort Benton, as exhibited in the high temperatures of December, 1853, and April, 1854. In 
several instances warm days occurred in the winter months in connexion with high southwest 
winds. It is remarked by Mr. Doty that these winds are also attended with a fall of barometer, 
though this does not appear in the record asa marked result. In May, 1854, there were two 
days on which snow fell, and the temperature was below the freezing point. 
For the year the mean temperature compares very nearly with that of Fort Kearney, which is 
quite similar in the corresponding seasons, and it is three or four degrees above that of Fort 
Snelling, the excess occurring in winter and spring. 
The hygrometric observations during the autumn of 1853 were taken in-doors, and have no 
value as measures of the dryness of the atmosphere. They were resumed in April, 1854, and 
appear to be correctly taken to the close of August. The differences they give for the readings 
of the dry and wet thermometers are quite large; and though a mean of these gives but a rude 
approximation towards the true results of proportion of moisture in the air, some intelligible idea 
may still be obtained from them. Arranging these mean differences, for the several hours from 
May to August, we have the following results : 
Mean differences of readings of the wet and dry thermometers at 
7 a.m. 2 p.m. 9 p.m. Mean for 
month. 
° ce] ° ° 
May. pace t 1654200. tosis wewse cdecidnns 5.7 13.3 71 8.7 
June .... --.. 0. co ce cosnee anncne ence ences 7.4 16. 1 8.0 10.5 
Silly, -seeness d0esscstt -eecs uses eseue- 12.2 20.7 12.2 15.0 
August ..---- dO. 2022 none cone cone enon eoe= 15.6 23.5 11.0 15.7 
Tf correctly observed, these show a climate similar to that of New Mexico, and the arid por- 
tions of the plains, and of California. The maximum of single differences in the Atlantic States 
rarely exceeds 16°, and on the mean for a month does not reach 8°. In this case, the observa- 
