20 LETTER OF MR. BLODGET ON BAROMETRICAL OBSERVATIONS, 
av 
These reductions were made by D. G. Major, Esq., under the direction of Mr. Blodget, who 
furnished the tables and formule, and, at my request, the feline communication referring 
to their application: 
*€ SMITHSONIAN INSTITUTION, July 16, 1854. 
‘Dear Str: At your request, I make the following note of the direction given to the reduc- 
tion of the barometric observations made on your line of survey from the Pimas villages, on the 
Gila river, to El Paso. They were reduced in part by the aid of new constants and new modes 
of correction, which require some explanation. 
** Determinations of elevation by the barometer are subject to error from two causes not con- 
sidered in the formule and tables usually employed, or those providing for this determination, 
though simultaneous observation on a vertical line. 
**'The survey of an extended surface-line necessarily involves liability to those errors, and it 
cannot be accurately made by the principles applicable to vertical and simultaneous comparisons 
alone. 
“The sources of error are, first, variable constants of atmospheric pressure, both in the 
changes for the day and among the months; and, second, non-periodic changes, or variations 
without regularity or definite recurrence. 
** To avoid the first error, or that arising from horary variations of pressure, a scale of cor- 
rection for the observations made at each hour has been applied, reducing each reading to a 
mean position for the day. This scale is a new one, of larger range of variation than that ap- 
plicable in the eastern United States and in Europe; and it was determined from hourly obser- 
vations made by the survey under Lieutenant Whipple, corrected and confirmed by reference to 
the results of hourly observations by the boundary survey, which were made accessible for this 
purpose by Major Emory and Mr. Chandler. The scale is given in connexion with the compu- 
tation and results. 
** To correct the work for non-periodic variations of pressure, the principal camps are d 
to each other consecutively, and each to a principal camp; correcting the determination by the 
mean of these results. Each camp is also referred directly to the sea-level, assuming a mean 
pressure at sea-level in those latitudes of 30.050 inches, with the CHER SN corrected to the 
reading at freezing-point, and the air temperature at 649, These results agree very nearly with 
those obtained by the first and preferable mode. 
** The intermediate stations and minor camps are first referred to the nearest principal camp; 
and the line formed by successive differences from such point of departure is corrected, if found 
not to agree with the single difference determined from camp to camp by a "—e correc- 
tion of the intermediate elevations. The result of elevations are still liable to error, from a 
measure of non-periodie variation that cannot be determined, but they are the best ebk in 
this description of survey, without simultaneous observations at stations very near each other. 
**'The correction for monthly variation of pressure would be very small at the date of this 
survey. 
“It is proper to say that the greatest error probable in the determination of the absolute 
elevation of any camp by those methods cannot exceed one hundred feet, and the error of any 
grade would be wholly unimportant. 
** Respectfully, yours, 
| “L. BLODGET. 
“ Lieutenant PARKE, 
** In charge of Survey of Line from Gila river to El Paso.” 
