90 BAROMETRIC OBSERYATIONS. 
The mean of the observed air temperatures is used in these cases also, as avoiding, to some 
extent, a source of error in extremes of surface temperature; for which, in single observations, 
a table of corrections is appended. 
All the observations were also corrected for horary variation of atmospheric pressure through 
the day, thus bringing each to the true mean position for the day. 
The accompanying scale of horary corrections gives the values employed for each hour. 
They are derived primarily from well-determined curves of daily variation of pressure for 
the eastern United States, but with material and important modifications and additions estab- 
lished by the observations of other surveys in the interior of the continent, principally by that 
of Lieutenant Whipple. By the observations through the winter months at Great Salt Lake 
City, the measures of this horary scale are shown to be less for that season, and to conform then 
more nearly than in summer to those observed in the eastern United States and in Europe. For 
the months occupied in the field-work of this survey, however, and for the districts traversed, the 
measure of the correction here employed is fully confirmed. At the sea-level, or so near it, as 
both extremities of the line are, the measures of horary variation again fall off 10 those belonging 
to well-known districts ; yet, as no determinations of importance occur at these extremities, it is 
not necessary to give the scale belonging to them. 
A correction previously found to be required for extremes of air temperature has been so well 
determined by the comparison of survey by levels, and with the barometer at the passes of the 
Sierra Nevada, surveyed by Lieutenant Williamson, that a scale of corrections sufficiently precise 
for practical use has been constructed. Where the error from this cause could not be eliminated 
by the use of mean temperatures, this scale has been employed in the determinations here made. 
The measures given for this correction belong to extremely arid climates, and to elevated dis- 
tricts, requiring modification in the position of the point where no correction is required; also 
in different seasons. - As it affects great elevations in these arid districts by an extreme amount 
of at least 150 feet, it is too important to be neglected, notwithstanding a discretionary use of the 
v^' is usually necessary. 
The reduced observations at stations on the Plains, from Pawnee fork to camp 33, (the first camp 
after crossing the Arkansas river), were referred to the mean barometric reading noted by Dr. 
Engelmann at St. Louis for July, 1853—the month in which they were made. The altitude of 
his station above the Gulf of Mexico, as determined by him from a long series of observations, 
was added to make up the entire altitude. For these stations and dates,the results thus 
obtained are very nearly identical with those computed by direct reference to the barometric 
mean at the level of the sea for the latitude. 
For altitudes beyond this point direct comparisons of each camp are made to an assumed mean 
barometric reading at the level of the sea, in this latitude, of 30,000 inches—the barometer 
corrected to 32° Fahrenheit, and a mean air temperature taken of 579. "The constant belong- 
ing to the latitude and climate of the Gulf is 30.050 inches of the barometer, and 64? of air 
temperature, which would add unduly to the altitudes. The principle is assumed that the con- 
stants of pressure and temperature employed belong to the latitude, and that the resulting deter- 
minations of elevation belong correctly and alike to both the Gulf and the Atlantic and the Pacific. 
There are no well-determined mean readings of the barometer on the Pacific coast, yet the most 
of. those recently made in California give the impression that a slightly greater mean atmospheric 
_ pressure exists there than in the same latitudes of thc Atlantic. ‘The constant has not, however, 
been altered for any portion of this line. 
Abe. discussion of observations at the principal passes (oe been in part upon simultaneous 
) tions at an hour's interval in time and distance. "The slower movement of one barometer, 
', ren. le red i it necessary in many cases to take successive readings of the best one, cor- 
ry variation, for determinations of successive differences. "The termini of these 
descent were also checked by comparison of preferred results, as of the 
