SHORT MEMOIRS ON METEOROLOGICAL SUBJECTS. 471 



According to our theory, the existence of these winds on these borders 

 suffices to explain the high pressure that is experienced in the central 

 regions; it seems to us, then, very probable that the variation of the 

 pressure with latitude ought to be attributed to the permanent currents 

 that the earth's atmosphere presents. The pressure should increase 

 then from the equator to the parallel of 30° in the whole region where 

 the trade-winds of the equatorial zone prevail ; it should decrease from 

 the parallel of 40^ to the pole over the whole region where the return 

 current or the contra-trade of the temperate zone prevails. 



It is on the parallel where the interchange is made of the second cur- 

 rent for the first that the pressure attains its maximum value. 



(7) Explanation of the Increase of the Monthly Oscillations of the Barome- 

 ter from the Equator to the Pole. 



One of the most prominent facts that is presented by the observation 

 of the barometer is the rapid increase of the irregular oscillations in 

 proportion as we go away from the equator. The difference of the 

 monthly extremes, which at the equator scarcely surpasses on the 

 average two or three millimeters, is ten millimeters at the parallel of 

 30°, and becomes thirty and even forty millimeters in the polar regions.* 

 This enormous variation of the monthly barometric oscillation was indi- 

 cated a long time ago by De Saussure, who has said that any hypothe- 

 sis destined to explain the barometric oscillations ought first of all to 

 take account of their increase with the latitude. Let us see if our theory 

 gives any satisfaction as to this point. 



In the formula which gives the transverse barometric variation for 

 an atmospheric current of the velocity F, and which is 



Ar (J ' 



the sine of the latitude enters as a fiictor. An atmospheric current of 

 constant intensity produces, then, on its borders variations of barometric 

 pressure which are very unequal, according as it is develoj^ed near the 

 pole, or near the equator. If we represent by unity the variation of the 

 barometer in the polar regions, it becomes 0.71 at the latitude of 45<^, 

 0.50 at the latitude of 30o, and 0.2G at the latitude of IS© 



Let us apply to the numbers cited by Kiimtz a correction for the lati- 

 tude calculated according to these considerations; in other words, let 

 us divide by sin X the monthly oscillation of the barometer in order to 

 reduce it to that which would have been produced at the pole of the 

 earth by a current of the same intensity. We shall obtain a series of 

 numbers which increase with the latitude, but in a much less rapid 

 degree than previously. The monthly oscillations reduced to the pole 

 become ten to fifteen millimeters for the equatorial regions, twenty 

 millimeters for the parallel of 30°, and thirty-five to forty millimeters 

 for the polar regions. 



* Kiimtz, Meteorologie, p. 281. 



