MONTHLY MEANS. 



Table 9i. 

 Monthly pressure in McMurdo Sound. 

 Reduced to 32°F., sea-level and gravity at 45°. (Inches. 



IG7 



(a) Feb. 1902 to Jan. 1903. 



(b) Feb. 1903 to Jan. 1904. 



The above values have been plotted in figure 54 and the months of the same year connected 

 up by a characteristic line. The resulting diagram is exceedingly instructive showing as it 

 does the tremendous variations of the pressure for the same months from year to year 

 compared with the normal yearly variation. It also shows the impossibility of forming even 

 an approximate idea of the yearly variation from one year's observations. Thus the year 1903 

 showed the maximum pressure in June while in 1912 the minimum pressure occurred in this 

 month. Similarly November was the month of maximum pressure in 1912 and the month 

 of minimum pressure in 1903. 



When the mean values of each month for the four years are taken the irregular changes 

 from month to month are still large (thin curve in figure bHa) and it is clear that many 

 more years would be required before a smooth curve showing the true yearly variation would 

 be obtained. 



When one has to deal with a curve subject to such large accidental variations it is 

 legitimate to endeavour to eliminate them by smoothing the curve. I have tried several 

 methods of smoothing this curve and find that the most satisfactory result is obtained by 

 applying the formula b=-^— | — -■ The numerical results are given in the last line of table 94 



and the thick line in figure b5a shows the resulting curve. In its main features this curve 

 is probably correct. 



The numbers given in the last line of table 94 have therefore been taken throughout 

 this work as the normal values of the monthly pressure in McMurdo Sound. 



The result that the pressure is lowest during the cold winter months and highest during 

 the warm summer months is very surprising, but there can be little doubt of the reality 

 of the phenomenon. 



