GEOGRAPHICAL DISTRIBUTION. 175 



July is due to the abnormally high July pressure fouud in 1899, and the high summer 

 values are due, as in the case of Framheim, to the unusually high pressure during the 

 summer of 1911-12. 



Summary. — From this discussion of the monthly valu,iS of pressure at the three Ross 

 Sea stations we see that whether we take only the actual observations made at Framheim 

 and Cape Adare or the values for those stations deduced from the four years' observa- 

 tions at McMurdo Sound we arrive at the same general type of annual pressure variation 

 at all three stations. There seems little doubt therefore that in the Ross Sea area pressure 

 is highest in December and lowest in September or October. The pre.ssure falls fairly 

 regularly from December to July, remains more or less constant until October and then 

 rises very rapidly to its maximum in December. 



The annual variation of pressure has been determined in other parts of the Antarctic, 

 but except at Snow Hill wh.ere observations were made for twenty months, the data are 

 only for twelve months or less at each station. It has already been pointed out that the yearly 

 variation based on a single year's observations in the Antarctic can lead to very erroneous 

 conclusions. It would therefore be unprofitable to compare the individual results from other 

 stations, and a general comparison can best be made when the meteorology of the Antarctic 

 as a whole comes to be reviewed. 



Geographical distribution of pressure. 



We have now determined the mean annual pressure at the throe stations in the Ross 

 Sea area to be : — 



Cape Evans 29-26" 



Cape Adare 29- 19" 



Framheim 29-14" 



The winds at Cape Evans are so affected by the surrounding mountains that they give 

 us little information of the actual pressure gradient in the neighbourhood of the station. 

 There can however be little doubt that the air motion over the west of the Barrier is 

 mainly from the south, therefore we must assume that the isobars over that part of the 

 Barrier run more or less parallel to the Western Mountains with the highest pressure close 

 to the mountain range. The resultant wind direction at Framheim during the ten months — 

 April 1911 to January 1912 — was from S. 74° E. The exposure for winds at Framheim was 

 almost perfect, we must therefore conclude that the isobars run very aiiproximately in this 

 direction with the low pressure to the north. If the wind velocity is proportional to the 

 gradient the isobars must be closer together near Cape Evans than near Framheim, for the 

 resultant wind velocity during the ten months mentioned above was 10-7 miles an hour at 

 Cape Evans and only 3-1 miles an hour at Framheim. The isobars over the Barrier and 

 the south, of the Ross Sea shown on figure 58 have been drawn to conform with these condi- 

 tions. The distribution of the pressure near Cape Adare cannot be judged from the wind 

 direction and pressure at that station. The pressure distribution shown over the north of 

 the Ross Sea in figure 58 has been deduced from the general pressure distribution shown 

 in the weather maps contained in Volume II. 



The pressure distribution shown in figure 58 is probably correct in its main outlines, 

 but the distance between the isobars is only to be taken as indicating the gradient, for in 

 the absence of free wind ob.servations at Cape Evans and Cape Adare it is impossible accu- 

 rately to determine the gradient. The diagram shows that the mean pressure over the Barrier 

 is higher than over the Ross Sea, and there is every indication that the pressure is iower 



