32 Staff-Commander E. W. Creak on the Magnetical [May 8, 



The extreme deflections of the magnet (always regarding the mean 

 position of the north end as being the magnetic north) took place, at 



Floeberg Beach. 



Westerly f December , 10 ^ Easterly f December > 1 a.m., 



-< January, noon, , < January, 9 P.M., 



extreme. J ' extreme. «" 4 



L February, 1 p.m. I February, midmgbt. 



Westerly 

 extreme. 



October, 11 p.m., 

 November, 11 p.m. 

 December, midnight, 

 January, 10 P.M., 

 February, 11 p.m., 



Discovert/ Bay. 



October, 11 a.m., 

 November, 10 A.M., 

 December, 10 A.M., Easterly 

 January, 1 P.M., extreme. 

 February, noon, 



March, noon. [_ March, midnight. 



As Van Rensselaer Harbour and Port Bowen, as well as Floeberg 

 Beach and Discovery Bay, happen to lie in a region within which 

 observers hitherto report their inability to discover any connexion 

 between appearances of the aurora and magnetic disturbances, the 

 mean hours of extreme deflection of the magnet for the whole period 

 are here given for comparison : — 



rNoon Van Rensselaer Harbour. . 2 a.m. -» 



Westerly J Noon Floeberg Beach Midnight I Easterly 



extreme. ] 10 a.m Discovery Bay .. „. 11 P.M. j extreme. 



*■ 11-49 a.m... Port Bowen 10 p.m. J 



As it is probable that magnetic disturbances have had a certain 

 degree of control ever the determination of these hours of extreme 

 deflection, it now becomes necessary to give results freed from such 

 disturbance. 



The method of eliminating disturbances adopted in the discussion 

 of these observations, has been to select certain days of principal dis- 

 turbance, and to reject the observations of the whole of such days 

 from the mean. 



The observing officers recorded 32 days of undoubted disturbance. 



An examination of the hourly scale readings on these days shows 

 that the motions of the magnet were very irregular from hour to hour, 

 and that there were other occasions when similar movements took 

 place and might likewise be considered days of disturbance. Thus 

 from the 37 days of observation at Floeberg Beach, 18 days were 

 rejected as disturbed, and at Discovery Bay 69 days out of 156. 



But another method, which consists in deciding upon a separating 

 value and rejecting all hourly observations exceeding that value, may 

 also be said to have been satisfied, for out of 2,088 hourly scale read- 

 ings at Discovery Bay, all those that exceeded the mean hourly value 

 for the whole month by 30' of arc have been rejected, with the excep- 

 tion of 15 which differed 40' of arc. 



