Mr. J. A. Broun on Terrestrial Magnetism, 85 



duced to conclude that the variations of the secular change arc 

 not due to disturbance, though both apparently, obeying the 

 same law, are due to the same cause. This is the first appear- 

 ance of a law affecting the secular variations. 



I have endeavoured to show that the frequency of the appear- 

 ances of the aurora borealis obeys the same laws as the amount 

 of the magnetic disturbance. 



The general connexion betwixt the appearance of the aurora 

 borealis and magnetic disturbances was known in the time of 

 Graham, Celsius, and Wargentin. The fact that the needle was 

 even disturbed sometimes on the approach of an aurora (that is 

 to say, when invisible at the place of the needle), is noted by 

 Mairan as a result obtained by Wargentin (Mairan, p. 450, 2nd 

 edit.). It is to M. Arago that we owe a distinct statement, and 

 a considerable amount of evidence of the fact, that when there is 

 a magnetic disturbance in Europe, there is generally an aurora 

 borealis visible in higher latitudes, if not visible at the place 

 where the needle is disturbed. 



My own observations proved that there was always an irregu- 

 larity in the movement of the needle, even when the aurora was 

 most feeble ; in some cases slight disturbances caused me to 

 search for the aurora and discover its existence, when otherwise 

 it would probably never have been remarked. I have noticed 

 several times in the Makerstoun Observations a curious phjeno- 

 menon occurring with faint magnetic disturbance, which sud- 

 denly covered the whole sky as with a thin milky veil, that I 

 have designated milky aurora. I never, however, observed a 

 magnetic disturbance at Makerstoun without finding an aurora 

 also visible, if the state of the sky was favourable. The follow- 

 ing are the laws of aurora deduced from my own observations : — 



Diurnal law. — The maximum frequency in the winter months 

 occurs about 9 or 10 p.m. This fact is as old as the correspond- 

 ing one for magnetic disturbance. 



Monthly law. — As I lost no opportunity of searching for 

 aurorse during the years I had the direction of Sir Thomas 

 Brisbane's observatory, I conceived, if a monthly law existed, it 

 might be deduced from my observations. I think I have shown 

 strong reasons to conclude that the maximum frequency of the 

 aurora occurs near full moon. It is curious that this is the 

 result that Mairan finds should follow from his theory (Mairan, 

 p. 280). This is the epoch I have found for the maximum dis- 

 turbance also. 



Annual law of variation of frequency of the Aurora Borealis. — 

 This law, that the maximum frequency occurs at the equinoxes, 

 has been attributed lately to AI. JIausteen (I'roc. Roy. Soc. 

 vol. vii. )). 430). It was attributed by M. Kacmtz to Mairan. 



