1872.] 



26-day Period of the Earth's Magnetic Force. 



423 



The results for 1844 and 18/0 are, I think, very near the truth, and 

 they confirm, within 2 days, the epoch with a period of 26 days ; and 

 whatever value may be given to the epoch for 1852 compared with that 

 for 1844, will diminish the probability of the value 26-J-. On the other 

 hand, 1852 compared with 1870 satisfies exactly the period of 26-J- days; 

 but it would also be satisfied equally well by a period of 26*02 days ; so 

 the interval is, in this case, too great to decide between the two values. 

 The other results, however, seem to me sufficient, and, neglecting the long 

 intervals, we have from 



1844-45, period = 25*96 days. 

 1852-50, period = 25*79 „ 

 1868-70, period = 25*90 „ 



and the mean 25*88 days is probably within a tenth of a day of the truth. 



As far, then, as the existence of a period of near 26 days is concerned, I 

 think there cannot be the slightest doubt ; the examination of great masses 

 of observations has confirmed my belief in it, as it has Dr.Hornstein's. But 

 we know nothing certainly as to its cause : it appears to be most probably 

 connected with the sun's rotation ; but in what way this may act nothing 

 is known. The single periods show great breaks, and what may be termed 

 accidental minima, in opposition to the minima belonging to the period : 

 these accidental minima are connected with great disturbances, probably 

 allied to the solar eruptions or to causes which generally produce spots and 

 protuberances. It is to these accidental minima that the smallness of the 

 coefficient a in the term for 1845 is due. "We might suppose that the sun 

 during its rotation produces an action on the magnetic or electric ether in 

 motion, which, as far as it acts on our magnets, may be supposed in greater 

 quantity or more condensed in certain parts of the earth's orbit and in 

 certain years ; and, as has been supposed in the case of the frequency of 

 the solar spots, this ether may also be acted on by the planets, and produce 

 an irregularity in the length of a few successive periods. These supposi- 

 tions are made merely to show that we are perhaps not in possession of all 

 the conditions of the problem, without which perfect exactness in the cal- 

 culations is impossible. 



In conclusion, I refer again those interested in the subject to plate xxvii. 

 in the Transactions of the Royal Society of Edinburgh, vol. xxii., where 

 the daily means of horizontal force are projected for four stations on the 

 earth's surface, all of which agree in showing the same movements, some 

 of which have an amplitude of '002 of the whole horizontal force (the 

 Astronomer Royal's result for 1870 gives a mean value of nearly the half 

 of this), and with intervals of about 26 days. 



I hope to be able to confirm this result hereafter by ten years' observa- 

 tions at Trevandrum *. 



* I regret that the papers containing my calculations for the first three years are not 

 at present in my possession, or I should have stated the result here. 



