116 DETERMINATION OF ALTITUDES BY BAROMETER. 



the head of Des Chutes valley, Oregon Territory, at an elevation of ah / t four thousand feet 

 above the level of the sea, and a mean temperature of 50° Fahr. We were again fortunate in 

 having little abnormal error, and the table was admirably adapted to the observations taken on 

 the elevated plateau east of the Cascade range. 



Table No. 3 was deduced from three days' observations at Fort Beading, taken about the 

 middle of November, with a mean temperature of about 50° Fahr. Here a barometric storm 

 rendered it necessary to reject four days' careful observations, to avoid the effect of abnormal 

 error. During the three days, however, there was very little atmospheric disturbance ; and the 

 table was useful for the observations taken on my route from Vancouver, where the temperature 

 was uniformly low and the elevation generally inconsiderable. 



It is very interesting to compare these tables, especially the two for Fort Reading. As the 

 Des Chutes curve was obtained from observations at a place differing widely from the others in 

 altitude, mean temperature, and latitude, it is impossible to decide what part of its peculiarities 

 is due to each of these causes of variation. The Fort Reading curves, however, are deduced 

 from observations taken at the same spot, with a change of mean temperature only ; and it is 

 fair to suppose that the differences between them are due principally to this cause. These 

 differences are, that the November curve has a more rounded form and departs less from the 

 horizontal line representing the mean of the day, and that its points of maximum and minimum 

 are nearer together by about three hours. These results are precisely what we should expect, 

 from the great difference of 33° Fahr. in mean temperature, assuming the heat of the sun to be 

 the cause of the horary oscillation ; and they show conclusively that the horary variation is by 

 no means constant, even at the same locality, for all seasons of the year. For some hours, the 

 difference between the corrections in these two tables would affect the computed height of a 

 station more than forty feet. This clearly proves the great error, which has sometimes been 

 committed, of applying a table of horary corrections to observations taken at places of very 

 different mean temperatures, and far distant from the spot for which it was computed. 



4. For abnormal oscillation. — The abnormal oscillations of the barometric column are princi- 

 pally caused by general movements in the atmosphere, which are shown by repeated and 

 numerous observations to extend rapidly and progressively over very large tracts of country. 

 In the first volume of the third edition of Professor J. F. Daniell's Treatise on Meteorology, 

 will be found a very interesting article upon these oscillations. It is illustrated by diagrams, 

 constructed as explained above, which represent numerous barometric observations taken under 

 the direction of the Meteorological Society of the Palatinate. For Europe, at least, they con- 

 clusively prove this most important principle of rapid and wide extension. In the third volume 

 of this series of reports upon Pacific Railroad Explorations, will be found diagrams, prepared 

 by Captain A. W. Whipple, United States Topographical Engineers, to represent barometric 

 observations taken at different military posts by the Medical Department of the United States 

 army. They show that the same principle is true for our western country. Slight local storms 

 do not appear to produce much effect upon the height of the barometric column. The observa- 

 tions taken on our survey agree perfectly with these well known facts. One of our barometers, 

 No. 1068, was left at Fort Reading during our field work, with Dr. J. F. Hammond, United 

 States army, who kindly volunteered to have observations taken daily every third hour, from 

 6 a. m. to 9 p. m. I obtained a corresponding set of observations taken at San Diego, distant 

 over six hundred miles, and one month's observations at Benicia, distant about two hundred 

 miles from Fort Reading. Both sets were taken under the direction of medical officers of the 



