ON COMPARISON AND REDUCTION OF MAGNETIC OBSERVATIONS. 227 



G. B. Airy > from the two periods 1841-47 and 1848-57. The results for 

 these two periods, however, it must be said, differ widely between them- 

 selves, the range deduced from the first being nearly thrice that deduced 

 from the second period. 



Thinking more light desirable in the face of these discordances, I got 

 out the annual inequality for the mean of the five years 1887-91 from the 

 results published annually in Table XI. of the Greenwich ' Magnetical 

 and Meteorological Observations.' Proceeding as in Table X., i.e. taking 

 means for the first of each month, and applying the secular correction 6''4 

 deduced from the Greenwich tables, I obtain an inequality whose 

 resemblances to that shown in Table X. are not more conspicuous than 

 the divergences. 



The largest easterly declination appears in April-May, the largest 

 westerly declination in September -October, and the range, 0'"6, is even 

 smaller than in Table X. Against these comparative agreements must, 

 however, be set the fact that the first three months of the year show an 

 easterly departure from the mean. 



The divergences in the results obtained for the annual inequality of 

 declination do not, of course, necessarily imply that any of them are erro- 

 neous. The phenomena at any one station might not unnaturally present 

 considerable variations — at least in range — from year to year ; and it is 

 conceivable that local influences may be more effective in this than in 

 other phenomena. It has also to be borne in mind that the data em- 

 ployed at the several stations were selected on different principles. Still, 

 I am doubtful whether any more definite conclusion should be drawn 

 than that the annual inequality of declination near London is at present 

 a very small quantity. 



§ 20. The annual inequality of horizontal force shown in Table X. is, 

 comparatively speaking, large and unmistakable ; its range is a large 

 fraction of the secular variation. In this instance there is a very fair 

 agreement with the results given on pp. 166, 167 of Lloyd's ' Treatise ' for 

 Dublin on the mean of the years 1841-50, the most conspicuous differ- 

 ence being that the Dublin range was some 50 per cent, in excess of that 

 given by Table X. 



The only previous determinations, so far as I know, of the annual 

 variation of the horizontal force at Kew are those of General Sabine '^ and 

 Dr. Balfour Stewart,^ for the epochs 1857-62 and 1863-68 respectively. 

 The former found the horizontal force, corrected for secular change, to be 

 on an average about -00012 C.G.S. units higher in summer than in 

 winter, while the latter found no difference. This divergence might be 

 attributed to the epochs considered being different ; but I feel consider- 

 able doubt as to the data employed having being adequate. They appear 

 to have been in both cases simply the results of the absolute observations, 

 uncorrected by reference either to the magnetograph curves or to the 

 diurnal variation. 



In conclusion, I have much pleasure in acknowledging my indebted- 

 ness to Mr. T. W. Baker, chief assistant, and Mr. R. S. Whipple, librarian 

 at Kew Observatory, for explanations as to the methods of standardising 

 the magnetic curves at Kew, and for other valuable information and 

 assistance. 



' Phil. Trans, for 1863, p. .311. ^ Pldl. Trans, for 1863, pp. 298, 299. 



' Proc. Roy. Soc, vol. xviii. 1870, pp. 238, 239. 



Q2 



