MAGNETIC OBSERVATIONS AT THE KEW OBSERVATORY. 



445 



Semiannual Inequality. — In the elaborate and very valuable discussion of the magnetic 

 observations at Girard College, Philadelphia, 1840-1845, contained in the thirteenth 

 volume of the 'Smithsonian Contributions to Knovpledge' (1863), Dr. Alexander 

 Dallas Bache, For. Mem. R.S., announced the existence of a semiannual inequality in 

 the lunar-diurnal variation of the Declination as shown in two particulars, (Ij that the 

 amplitudes of the deflections both to the east and to the west, and in both the 12-hourly 

 divisions of the (lunar) diurnal variation, are less in the six months from October to 

 March than in the six months from April to September ; and (2) that the lunar hours 

 at which the deflections pass through the zero are earlier by more than an hour in 

 October to March than in April to September (Smithsonian Contributions, vol. xiii. 

 part 3, pp. 11-13). In the case of the Horizontal Force, Dr. Bache finds a similar 

 semiannual inequality in the hours of maxima and minima, but in respect to their 

 amounts he remarks that the range in' April to September is but slightly greater than 

 that in October to March, the difference being by no means so marked a feature as in 

 the Declination (Smithsonian Contributions, vol. xiii. part 6, p. 72). The lunar-diurnal 

 variation of the Vertical Force at Girard College does not appear to have been examined. 



The above results were derived by a treatment of the observations conducted according 

 to the same general principles, in regard to the separation of the larger disturbances, 

 which had been previously introduced by myself in the discussion of the British Colonial 

 magnetic observations. 



Following Dr. Bache's example in this particular investigation, I directed Mr. Magrath, 

 who is charged with the superintendence (under myself) of the reduction of the magnetic 

 observations at the Woolwich Office, to separate the lunar-diurnal effects at Kew into 

 two portions, one of which should contain the months from April to September inclusive, 

 and the other the months from October to March inclusive, being the division of the 

 year adopted by Dr. Bache. The subjoined Table (No. V.) exhibits the semiannual as 

 well as the annual means of the lunar-diurnal variation of the Declination at Kew taken 

 from the photograms commencing January the 1st, 1858, and terminating December 31, 

 1864. The Table is divided into two portions solely for convenience in printing. 



Table V. — Annual and Semiannual Means of the Lunar-diurnal Variation of the 



Declination at Kew. 



