of Terrestrial Magnetism. 447 



agree with those of Mr. Broun at Makerstoun, as may be seen in 

 the results for 1846, p. 87, plate 1. The following are the re- 

 sults with respect to the frequency and magnitude of the per- 

 turbations in the different months of the year : — " The mean 

 value of a perturbation is a maximum in the equinoctial months, 

 less in the winter months, and a minimum in the summer 

 months." At Hobarton the difference between the summer and 

 the equinoctial months is scarcely perceptible. The proportions 

 of the frequency and of the magnitude of the perturbations in 

 each month, relatively to the sum of those observed in a year, 

 come out a minimum in the winter months, a maximum in the 

 equinoctial months, and intermediate in the summer months. 

 These conclusions, however, may be somewhat varied by the use 

 of different systems of reduction, depending chiefly on the de- 

 finition of an extraordinary perturbation. This is not the case 

 in the laws of the perturbations taken with reference to the day, 

 because in that case all the methods of reduction bring out the 

 same result. Thus the same result is seen in the curves given 

 by Mr. Broun, whose method of reduction is different from that 

 used by Colonel Sabine. 



But a most singular fact, which has been discovered in these 

 researches, is the start which the mean annual values of the 

 perturbations take between the years 1845 and 1846, when they 

 are almost doubled. This being a fact of the greatest importance", 

 Colonel Sabine has endeavoured to place it beyond doubt by the 

 best possible proofs. The following is the table given by him 

 (p. 115) 



The circumstance of the last three years having ratios almost 

 double of the first three, appears not to be accidental, particu- 

 larly when we observe that in the two observatories, which are 

 almost antipodal, the same fact appears, and that during all the 

 six years the same instruments were employed. Besides, in the 

 same years the observed diurnal excursions of the declination, of 

 the inclination, and of the total force, 1 ave sensibly increased 

 beyond the limit of any probable error. Lastly, the same 

 fact, in regard to the diurnal excursion of the declination, 

 results from the observations of Dr. Lamont at Munich in 

 Bavaria. Further observations may throw much light on this 



