180 



PRESSURE. 



Twenty four-hourlij jjeriod. 



This term reflects the characteristics which we have already noticed from our survey 

 of the actual curvas. The months with the largest whole day amplitudes are April and 

 August. The strange difference between June and the months on either side of it is also 

 shown by the large anip'itude of June and the small amplitudes of May and July. It will 

 a'so be noticed that the amplitudes and phases for the whole day period of months forming 

 pairs in figure 59 are very similar. 



Naturally the phases depend largely on whether the morning or evening maximum is the 

 most developed. Thus the phase for the months having the morning maximum highly deve- 

 loped — October to February — have phases varying about 250°, while the phase for those with 

 the eveniiig maximum devebped — April, June and August — is nearer 150°. 



For the year the amplitude of the whole day wave is only -002" and the phase 204°, 

 which brings the maximum of this wave at 16 hours 24 minutes, and the minimum at 

 4 hours 24 minutes. 



Twelve-hourly period. 



From the point of view of the physics of the atmosphere this is by far the most 

 important term of the harmonic series. 



Examining first the amplitude this shows a marked yearly variation with a maximum 

 in March and a minimum in August. The values of an are plotted as the bottom curve 

 of figure 59 and it will be seen that they lie very nearly on a sine curve having its maxi- 

 mum in February and its minimmn in August, the variations in cio therefore do not coincide 

 with the seasons, the maximimi occurring two months after midsuimner and the minimum 

 two months after midwinter. 



The phase of the half daily period is remarkable. During nine out of the twelve months 

 the phase lies between 140° and 1-50°, i.e., during these months the time of the maxima and 

 minima does not vary by twenty minutes.* Of the three remaining months, May, June and 

 November, the latter really departs from the above rule by accident due to a very abnormal 

 November in 1912. The three years 1902, 190.3 and 1911 combined give a phase for Novem- 

 ber of 148° and therefore we are justified' in saying that only two of the months. May and 

 June, have phases varying appreciably from the other months. Looking at the seasons this 

 result comes out even more forcibly, for the pha.se in the summer, autumn and spring only 

 varies by 1 degree from 147°, i.e., the phase in these seasons does not vary by more than 

 two minutes from the mean. On the other hand the phase for the winter months is only 

 116°, i.e., it varies from the mean phase in the other seasons by a whole hour. 



That this is not an accidental result is seen from the following statement of the phase 

 in the individual winter months. 



Table 101. 



Values of ^2- 



* A variation in A^ of 30° is equivalent to a variation of one hour iu the time of the maxima and minima. 



