APPENDIX II 



MAGNETIC VARIATION 



Whenever an opportunity occurred, as at Stanley, Falkland Islands, the variation was 

 recalculated, and with some unavoidable exceptions this was also done as a matter of 

 course at all places where a regular survey was made. 



The method was always the same. True bearings of a number of well-defined points 

 (usually six) at more or less equal distances round the compass were determined 

 from a solar observation and theodolite angles, or, as at Stanley, from a record of a 

 previous determination. A round of bearings by compass, in our case a 6 in. Dover 

 compass, was then taken and repeated left to right and in the opposite sense as often as 

 considered necessary. The difference between these and the true bearing is of course the 

 variation, and the mean of the differences tends to eliminate card and observational 

 errors. By taking the compass bearing both w'ays, the errors of friction of the pivot and 

 cap are cancelled. The final result, if carefully taken wdth a good compass, should be 

 reliable, although of course the compass is only marked to degrees, and sometimes half 

 degrees, and the minutes have to be estimated. The vicinity of any iron structure 

 should, of course, be avoided and iron implements or knives, if carried, laid down at a 

 distance. 



Variations were calculated in the Falkland Islands at Stanley (three times), and in 

 South Georgia at the following places: Undine, Larsen, Husvik, Leith and Prince Olaf. 



No abnormal magnetic variation was detected at any place visited during the time 

 covered by this narrative. 



Table of results 



