i82 



PRESSURi;. 



words instead of the maxima and minima occurring at each place at the sa e local time, 

 they occurred at each place at the same absolute time. Thus in north polar regions the 

 barometer oscillation apjiears not to be due to a wave travelling round the circles of latitude, 

 but the whole cap of air appears to oscillate outwards from the Pole. 



In table 102 are collected together all the observations of the twelve-hourly barometer 

 oscillations available from high southern latitudes.* 



Table 102. 



Constants of (he half daily harometer oscillation in the Antarctic. 



It is clear from this table that the times of maxima are not constant in either local 

 time or Greenwich time. The oscillations cannot therefore be duo to waves travelling round 

 the earth parallel to the circles of latitude, which would cause the maxima to occur at the 

 same local time at all stations ; nor to the oscillation of the polar cap which would cause 

 the maxima to occur everywhere at the same Greenwich time. It is possible, however, that the 

 oscillation observed at each station might be due to the sum of two oscillations, one round 

 the earth and the other to and from the Pole, for the sum of two oscillations of the 

 same period is another oscillation of the same period but with a different amplitude and phase. 



In order to investigate this solution I was led to review the work done previously on 

 the twelve-hourly barometer oscillation in the northern hemisphere and the results of this 

 work have been published in the Quarterly Journal of the Royal Meteorological Society, 

 Volume XLIV, No. 185, p. 1, 1918. 



It was shown in that work that the observed amphtude and phase of the twelve-hourly 

 barometer oscillation in all parts of the world can be explained by the existence of two 

 atmospheric vibrations — 



(a) a vibration around the world parallel to the circles of latitude, and • 



(6) a vibration between the equator and the poles parallel to the circles of longitude. 

 The former is expressed by 



h sin (2a;+B) 

 and the latter by 



c sin (2X+C-2X) 

 in which x is the local time, B and C are constants, and h and c are functions of the latitude. 



* Observations are also available for Port Charcot, but they are so widely different from those obtained at 

 the neighbourmg station of Snow Hill that they are clearly wrong ; they have therefore not been used in this 

 discussion. 



t Anales do la Ofiicina Meteorologica Argentin.i. In this publication the value.s given on page 63 for the 

 Fourier coefficients are wrong, the values used here have been calculated from the data given in table 44, page 62. 



