1863.] 625 



in the solar-diurnal variation, having its epochs coincident, or very 

 nearly so, with the sun's passage of the equator, and dependent 

 consequently on the earth's position in its orbit. The sun's action 

 in producing this semiannual inequality is shown to be character- 

 istically different from that which is manifested in the solar-diurnal 

 variation itself, pointing apparently to a difference in the mode of 

 the sun's action in the two cases. 



The sixth section contains a tabular view of the " Lunar-diurnal 

 Variation " at Kew, in each of the five years during which the photo- 

 graphic record has been maintained there ; this is followed by a 

 comparison with similar results at other stations on the globe, and a 

 statement of the principal points of agreement or of difference which 

 are shown thereby. 



III. " Results of the Magnetic Observations at the Kew Obser- 

 vatory, from 1858 to 186.2 inclusive/' No. II. By Major- 

 General EDWARD SABINE, P.R.S. Received June 18, 1863. 

 (Abstract.) 



This paper is a continuation of the preceding one. It consists of 

 two sections, the seventh and eighth. In the seventh section the 

 author discusses the secular change and annual variation of the decli- 

 nation; and in the eighth section, the annual variation or semi- 

 annual inequality of the inclination and of the horizontal and total 

 magnetic force. 



Seventh Section. The positions of the horizontal magnet at 24 

 equidistant epochs in the day, ^tabulated from the photograms of the 

 Kew declinometer, with the omission of the disturbed observations, 

 as described in the former paper, are grouped in weekly means, 

 forming 52 mean values, corresponding to the number of weeks in 

 the year. A Table is given of these weekly values, comprehending, 

 in five columns, the five years from January 1858 to December 1862 

 inclusive, and from these a sixth column is formed, representing the 

 mean declination in each of the 52 weeks of a mean or typical year, 

 corresponding in this instance to the year 1860. The mean declina- 

 tion obtained from all the weekly results in the five years, and corre- 

 sponding to its middle epoch July 1, 1860, is 21 39' 18"'l ; and from 

 a comparison of the mean declinations corresponding to July 1 in the 



