SURGES. 189 



(c) The result of (a) and (/;) is a steady decrease in the rapidity with which the pres- 

 sure changes take place as high latitudes arc attained. 



The physical explanation of those relationships will be taken up in chapter VI. 



Pressure surges. 



The pressure changes which we have studied so far have been actual changes of the 

 barometer from hour to hour and we have shown that the pressure variations are equivalent 

 to waves having a period of five or six days. When the pressure changes are plotted on a 

 much smaller time scale, it is seen that the waves we have so far considered are themselves 

 superposed on waves of a much longer period. These long waves, which I propose to call 

 pressure surges, are clearly seen on plate I. On this plate the mean daily pressures at Cape 

 Adare, Cape Evans and Framheim have been plotted and the points connected by a thin 

 line. The irregularities on the thin line are due to the pressure waves which we have already 

 considered, and as we wish to study the longer pressure waves we must eliminate them. 

 This is most conveniently done by arithmetically smoothing the curves. If instead of 

 plotting against each day the mean pressure of the day we plot the mean pressure of ten 

 days about it, small irregularities are smoothed out while the resulting curve follows closely 

 the larger waves. Thus the thick curves on plate I have been obtained by plotting the 

 mean pressure from the 1st to the 10th of any month against the 5th, the mean pressure for 

 the ■2nd to the 11th against the 6th and so on. Th.e waves on this thick line are the 

 surges which we are now- about to investigate. 



Comparison of pressure surges in different parts of the world. 



A glance at the plate shows that the same surges affect all three stations in the Ross 

 Sea area and later on we shall discuss the relative intensity of the surges at each station. 

 Before we do this, however, it seems desirable to enquire whether surges of this nature are 

 confined to the Antarctic or are recognisable in other parts of the world. 



As far as I know a similar investigation has not been made previously, therefore it is 

 necessary to calculate the ten-day means for the stations at which the comparison is to be 

 made. If the pressure surges are to be compared at several stations this necessitates calcu- 

 lating 365 ten-day means for each year for each station. If data for several years at several 

 stations are to be used, and several years' observations would be necessary for a complete 

 discussion, this will entail a very large amount of computing. I was unable to undertake 

 such a large amount of work, but the important r6!e these pressure surges play in the 

 pressure changes in the Antarctic constrained me to do the necessary calculations to compare 

 the surges for one year at least at several typical stations in all parts of the world. 



The year chosen for this comparison was from March 1902 to February 1903, as during 

 the greater part of this period there were observations in the Antarctic at the following widely 

 separated stations : Hut Point, Snow Hill, the Gauss Station and on Kerguelen. The stations 

 outside the Antarctic chosen to compare with these were the following ; — 



Wellington and Adelaide, which with Kerguelen are typical of the Southern Ocean. 

 Seychelles as t3'pical of a tropical ocean. 

 Bombay as typical of a monsoon climate. 

 Greenwich as typical of the northern temperate zone. 

 Irkoutski as typical of the centre of a large continent. 

 Stykkisholm as typical of the great Icelandic centre of action. 



Vardo as typical of noith polar regions, data from stations further north during the 

 period considered not being available. 



