HORARY AND AIR TEMPERATURE CORRECTIONS. 73 
of several at the summit or elsewhere, with the nearer camps. The coincidence of results by 
single and successive steps has been so satisfactory as to warrant the conclusion that the grades 
and altitudes of these passes are quite accurately determined. The correct use of such data in 
the joining of intermediate with main lines, and in the correspondence of single observations 
with the mean of observations, is the surest test of barometric survey, and discrepancies cannot 
be wholly removed. The principal cause of these discrepancies is in the non-periodic variation 
of atmospheric pressure, for which no constant or correction can be given, except by reference 
to continued readings for a month or more at some station near. Mean results best eliminate 
this error, and they are therefore preferred in the order of their number, or of the number of 
days they cover. In the ascent of the Arkansas, the uniform grade of the stream permits the 
use of four or five successive camps as a near result, and two or three non-periodic variations 
are thus eliminated. Subsequently, to Great Salt lake, a less error from this cause is likely to 
occur, as this variation is least in August and September. 
For the survey eastward from Great Salt lake to the valley of Green river, the field readings 
are compared with those of the same date by the other barometer at Great Salt Lake City: these 
may be regarded as nearly simultaneous, (though the hours were not always the same,) from the 
fact that the same days, and portions of the day, were compared. For the remainder of the line 
to California, no mode of correction of this error existed, except in the preference of the means 
embracing the greatest number of observations; yet the errors have probably no practical im- 
portance. 
The observations generally sustain the checking and criticism which rank the results as a 
determinate survey, with a near approximation to absolute accuracy, and, as among themselves, 
sufficiently conclusive of grades and points of comparison. 
Horary corrections of the barometer. 
Hour. Inches. Hour. Inches. 
6 a. m. . 007 3 p. m. —. 030 
7 a. m. . 020 4 p. m. —. 045 
8 8. m. +. 030 5 p. m. —,. 050 
a. m. +. 040 6 p. m. —. 030 
10 a. m. . 050 7 p. m. —. 020 
11 a. m. = 055 8 p. m. —. 005 
12 m. . 025 9 p. m. —. 000 
l p.m. +. 00 10 p. m. +. 010 
2 p. m. —. 015 11 p.m. . 020 
Scale of corrections for extreme air temperatures. 
Low temperatures. High temperatures. 
At 35° add 25° At 95° subtract 15° 
At 40 add 23 At 93 subtract 13 
t45 add 21 590 subt 
At 47 add At 88 subtract 10 
At 50 add 18 At 85 68 
At 53 add 15 At 83 subtract 7 
At55 add 13 At 80 subtract 5 
At 57 add 10 At 78 subtract 3 
At 60 add 5 At 75 subtract 2 
Norx.—The measures of this correction are variable to some extent with the season, and apply nearly to departures from the 
mean of the month in which observations are made. Itis here given as required for summer months, in which most of the 
observations in field surveys have been taken. 
10 b 
