4 REMARKS ON METEOROLOGY XSB BAROMETRIC RESULTS. 



difficult to determine on account of its variableness, and ■which is probably due, in part, to a 

 want of perfect adjustment, at the time of observation, of the index point ; partly to personal 

 error, and in part to the unequal bore of the tubes, causing an increase or decrease of capillary 

 attraction^ for it not unfrequently happens that two barometers, at a pressure of about 30 inches, 

 will read with a given difference uniformly ; but when subjected to a less pressure — 27 inches, 

 for example — other things being equal, this difference will often vary as much as 0.010. This 

 fact is observable even during the diurnal range of the barometer, though to a less appreciable 

 extent. 



The correction for horary oscillation was obtained from frequent hourly observations taken 

 for that purpose. Those used for determining the hourly corrections for the hours of the day, 

 to apply to observations made in connexion with the survey, are given in Table No. 1. In 

 Table No. 2 are the observations for the twenty-four hours at the places indicated. For the 

 construction of these curves, see Flate 1, Figs. Nos. 1 and 2. By reference to these tables and 

 diagrams, it will be perceived that this phenomenon is marked by irregular oscillations as to 

 the maximum and minimum point. This is probably due to the peculiarity of the season of 

 winter on the California coast, it being the period when greater fluctuations of the atmosphere 

 take place, and also to the want of a sufficiently long series of observations. In several instances, 

 during remarkable changes of atmospheric condition, the diurnal oscillations were completely 

 broken up. An example of this occurred on the 1st and 4th of January, 1855, at Camp 14', at 

 the headwaters of the Salinas river, when a great storm occurred. For a comparison of this 

 storm at several other places along the coast, see Flate No. 5. At Camp 17, January 16 and 

 17, the atmosphere was in a remarkably quiescent state, and the curve for those days is 

 probably very near the true measure for the time and place, the tropical hours occurring at 

 10 A. M. and 4 p. m. — (See Flate 1, Fig. 3.) 



The correction for abnormal oscillation, by far the most important element of error in barometric 

 discussion, and hitherto the most difficult of measurement, was obtained by a careful comparison 

 of nine months' daily simultaneous observations at Benicia barracks and at the mission of San 

 Diego, taken under the direction of the medical department of the United States army, with those 

 taken during the progress of the survey. * This comparison was made in accordance with the 

 plan of Captain A. W. Whipple, to whom the credit of the suggestion with reference to the 

 determination of barometric heights is due, and its results are satisfactory and confirmatory of 

 the plan pursued by him in the discussion of the meteorological observations pertaining to his 

 exploration near the thirty-fifth parallel. The diagrams (Flafes Nos. 2 — 10) will illustrate the 

 intimate relations of this abnormal effect upon the mercurial column for places widely remote. 

 By a combination of the observations at the above named places, and with those along the line 

 of survey, a scale of correction was obtained, which, when applied to each observation, has 

 produced interesting and accurate results. The manner in which this combination was made 

 may be briefly stated and the plates explained. The Benicia and San Diego observations for 

 nine months, from October, 1854, to June, 1855, inclusive, were plotted in their relative posi- 

 tions; the hours being by regulation, sunrise, 9 A. m., 3 p. m., and 9 p. m. These were sufficient 

 to indicate the direction of the curve of pressure from one day to another ; a line was therefore 



"There is a paragraph on this subject in Kaemtz Meteorology, p. 292, where the sensible parallelism of the curves is 

 recognized, hut no satisfactory results are given, nor any reference to their application to obtaining correct altitudes for an 

 extended barometric profile. 



