418 



Mr. J. A. Broun on the 



[June 20, 



Magazine* for August 1858, I made the following remark : — "The result 

 which I have now obtained from three years' observations near the mag- 

 netic equator, it appears to me, is wholly independent of the moon, and is 

 due to the sun's rotation on its axis. If we could suppose that the solar 

 magnetic poles are fixed, it might then be possible to determine accurately 

 the time of the sun's rotation by means of the movement of our magnets. 

 If, on the other hand, the poles are in motion, as I conceive they are, we 

 shall have to employ another period than 25*325 days, as obtained from the 

 solar spots. The period to be employed will of course be found by careful 

 examination of the observations and by trial" [of different times]. These 

 views resulted from a discussion of observations of the horizontal force 

 made at Trevandrum in the years 1855, 1856, and 1857. 



In February 1861 a paper by me was read to the Royal Society of 

 Edinburgh, containing conclusions derived from an examination of all the 

 observations of the horizontal force which had been published, made between 

 1842 and 1848, the whole having been recorrected for temperature by my 

 own method, for the purpose of this and other investigations. Founding 

 chiefly on the simultaneous daily means of observations made at Makers- 

 toun, Trevandrum, Singapore, and Hobarton during the years 1844 and 

 1845 (the only years for which a complete diurnal series existed for 

 Makerstoun), I arrived at the conclusion, from an examination of a series 

 of successive periods, shown nearly equally well at all the four stations, 

 that there was a period of nearly 26 days, probably due to the sun's rota- 

 tion on its axis, the mean of the whole number of periods being 25*96 

 days *. The observations, however, seemed to show that the single periods 

 had a variable length, though this might be due to some extent to irre- 

 gular disturbing causes, yet dependent, perhaps, on the period of the year, 

 and it might be on other arguments, such as the position of the planets. 



In the examination of observations made by me in India, I had sought 

 for the greatest amplitude on trial of periods of 26 to 27 days, but post- 

 poned a complete investigation till the observations of a greater number of 

 years should enable me to follow more certainly the apparent variations in 

 the length of the single periods ; this other investigations have hitherto 

 prevented, and I am glad that Dr. Hornstein has taken up the subject, 

 though evidently ignorant of my previous conclusions. 



Mr. Airy's communication on this subject is very important ; its con- 

 clusion refers to a period of 26-J days ; and although the evidence may be 

 imperfect for such a period, it seems to me that the Greenwich results, when 

 projected, give considerable grounds for concluding that a period of near 

 26 days exists. In what follows, I limit myself wholly to the observations 

 of the horizontal force, as I have found that element, when accurately cor- 

 rected for temperature, best fitted to show the period in question. 



The Astronomer Royal has noted a probable inaccuracy in the amount 

 of the correction employed for the secular change ; as this correction is of 

 * Trans. Eoy. Soc. Edinb. vol. xxii. pi. xxvii. and p. 543. 



