32 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 51 



than at the equator; again, on the Sea of Okhotsk and at Cape Horn 

 it is six Paris lines (0.53 inch) smaller than at latitude 23. 5 . At 

 any point in the interior of a continent whose altitude above sea 

 level is known by geometrical measurements (e. g., railroad level- 

 lings), the observed mean air pressure can be reduced to the value 

 appropriate to the sea level vertically below it and thus gradually 

 an empirical expression can be found for the pressure at sea level 

 and, therefore, for the atmospheric pressure in general, as a func- 

 tion of the longitude, latitude, and altitude above sea level, or the 

 distance from the center of the earth. Individual contributions that 

 I have made to this subject leave no doubt that above the solid 

 land, as also above the ocean, the mean air pressure at any lev 

 layer differs very much according to the latitude and longitude. 



The first of the above stated two fundamental assumptions (v 

 that the atmosphere is in a state of rest relative to the globe' 

 also decidedly negatived by ordinary observation. In one p 

 tion of the atmosphere, lying between the parallels of +25 an> 

 — 2 5 the air is at every minute and, therefore, on the averagt 

 of all time, in that state of strong steady flow that we call the trad t 

 wind; in other words, therefore, the average or permanent condition 

 of 0.4225 of the total mass of the atmosphere is a regular flow thai 

 is certainly not to be ignored. In the remaining portions of ,h 

 atmosphere, however, the movements are less steady as to time. 

 But when the successive motions of the air, during one or rr n 

 whole years at any place are combined into one resultant, the 

 general this resultant differs from zero and in such a way tha 

 direction and velocity depend upon the coordinates of the loca 

 and are independent of the years or number of years for which 

 computation was made. 



Therefore, after eliminating the influence of periodic and accid*- 

 tal circumstances and in direct opposition to the above-gi" 

 assumptions of the physicists, we find that the earth's atmosph' 2 '- 

 shows the following phenomena: 



(1) A current whose direction and velocity are independent of 

 the time and which, therefore, at every place depend only on the 

 coordinates of locality. 



(2) At any level surface (or one that is perpendicular to the result 

 ant of the explicit forces) the atmosphere is under a pressure that 

 varies with the coordinates of the points of this surface, but is con- 

 stant as regards the time. 



These two observed facts contradict the results of the ana" sir 

 of Laplace only because in place of a certain assumption fl nis 



