[ 36 ] 



III. On the Diurnal Variations of Magnetic Elements, as 

 depending on the Method of Tabulation. By William 

 Ellis, F.R.A.S., of the Royal Observatory, Greenwich*. 



THE interesting paper of Messrs. Robson and Smith, "On 

 the Diurnal Variation of the Magnet at Kevv," which 

 appears in the August number of the Philosophical Magazine 

 (page 140), includes a comparison between the diurnal in- 

 equality of the declination-magnet as determined at the 

 Greenwich and Kew Observatories in the years 1870, 1871, 

 1872, 1883, 1886, and 1887. 



The Kew results for the years 1883, 1886, and 1887 were 

 determined by the authors of the paper by Wild's method, 

 the hourly ordinates for about five quiet days in each month 

 having been measured. Those for the years 1870, 1871, and 

 1872 had been previously determined by Mr. Whipple, Super- 

 intendent of the Kew Observatory, and by him communicated 

 to the 1886 Report of the British Association Committee on 

 Comparing and Reducing Magnetic Observations; and were, 

 as Mr. Whipple kindly informs me, calculated from the records 

 of four selected quiet days in each month of the three years. 

 The results for the whole six years are thus in every respect 

 comparative. 



The Greenwich results with which those for Kew were 

 compared were deduced, not from selected days, but from the 

 records for all days excepting those, comparatively few in 

 number, on which either the magnet was excessively disturbed, 

 or the record from some accidental cause was wanting. 

 Whenever results for all days are hereafter mentioned, it will 

 be understood to be with this limitation. For the days thus 

 employed a pencil line was drawn by hand through each 

 photographic trace so as to represent the general form of the 

 curve, neglecting its petty irregularities, but giving con- 

 sideration to deviations of marked character. The hourly 

 ordinates were measured from such pencilled line, and it is 

 on the measures so made that the diurnal inequalities con- 

 tained in the Greenwich volumes, and employed by Messrs. 

 Robson and Smith, depend. 



Comparing together the Greenwich and Kew results for 

 declination, found in the way described, Messrs. Robson and 



* Communicated bv the Author. 



