INSTRUMENTS AND METHODS. 95 



on Wind Vane Hill varied between 1'25 and 1'50 times the velocities recorded by the 

 Dines anemometer at the Hut, the ratio varying from time to time as small changes in 

 wind direction made appreciable differences in the amount of shielding. On a few occasions 

 the Dines record has been used for supplying missing velocities when the cup anemometer 

 was out of order. 



From February i, 1911, to September 30, 1912, the electrical cup anemometer gave 

 a practically unbroken record of the hourly velocity of the wind. It has been found possible 

 to fill in the few lacunae either from the records of the Dines anemometer or bv reasonable 

 interpolation. 



Wind Vane.— A self-recording wind vane was installed dming February, 1911, on the top 

 of Wind Vane Hill. From the firet trouble was experienced in this position. It was exceed- 

 ingly difficult to change the paper during high winds, and the first five days' record was lost 

 owing to the blowing away of the paper when it was taken oft' the drum. The exposure 

 however was so good on the top of this hill that I was very reluctant to move the instru- 

 ment, but when low temperatures set in, the stopping of the clock became so frequent that 

 it became necessar}' on May 16, 1911, to remove it to the hut, where it was possible to 

 have the clockwork inside. The hut was however in tli6 lee of Wind Vane Hill and the wind 

 direction was affected to some extent by this obstacle. The change in direction was really 

 very small ; but it was just sufficient to cause the predominant wind to be changed from 

 S.E. to E.S.E., and a certain proportion of the E.S.E. winds just crossed the dividing 

 line into the E. division. This small break in the contiimity of the series was unfortunate, 

 but practically it has no large significance for, as will be shown later, the wind direction 

 was so largely affected by Mount Erebus, that for all piactical pui-poses the winds may be 

 divided simply into nort.herly and southerly, the former blowing from the Ross Sea to the 

 Barrier, and the latter fi'om the Barrier to the Ross Sea. 



The self-registering wind vane was nf)t entirely satisfactory as it did not turn very freely 

 and the time scale was so small that it was difficult to fix with accuracy the time when 

 the wind clianged. Also it gave other troubles so that there was a disappointingly large 

 number of missing records. Owing however to the great preponderance of winds from the 

 directions about E.S.E. and N.N.W. it was possible to fill in a large proportion of the 

 missing record with tolerable certainty. 



There is a more or less complete hourly record from February 7, 1911, to the end of 

 August, 1912, and in addition readings of a small vane exposed on AVind Vane Hill were taken 

 each morning from January 13, 1911, to December .30, 1912. 



Wind Velocity. 

 Mean Wind Velocity. 



The members of our expedition who had been previously in th.e Antarctic with Scott 

 or Shackleton, constantly remarked on the great amount of wind experienced in 1911 as 

 compared with that of their previous visits. The following year, however, was verj' much 

 worse, and the mean wind during 1912 at Cape Evans was the highest recorded up 

 to that time during twelve consecutive months in any part of the Antarctic. The large 

 variations from year to year and from month to month are shown in the following table in 

 which the mean velocities for each month are given for the four years 1902, 1903, 1911 

 and 1912. 



