224 



On the Variation of the Magnetic Needle. 



The mean of all the above observations, (excluding the morning 

 observations of five and six o'clock) is 5° 40' 34" W. These ob- 

 servations shew that the north end of the needle has in the morn- 

 ing a motion eastward amounting to from one to three minutes, 

 when the declination is usually less than at any other hour of the 

 day, and may therefore be called the minimum. This minimum 

 during the winter, is attained about nine o'clock, but during the 

 summer months commonly as early as seven. The needle then 

 gradually deviates to the w T est, and attains its greatest westerly bear- 

 ins about two o'clock in the afternoon, when the declination is great- 

 er than at any other hour of the day, and may therefore be called 

 its maximum. This maximum declination is attained during the 

 winter months, about three o'clock ; and during the summer, com- 

 monly as early as one. From this time, the needle again returns to 

 the eastward, till it attains its original bearing about ten o'clock, 

 and then continues nearly stationary until the next morning. The 

 mean of the observations at nine o'clock in the evening, is a little 

 less than at ten, agreeing with the results of other observers, who 

 had remarked an evening minimum. The difference in this case is 

 however, so slight, that it might be presumed accidental. 



The following table exhibits the differences between the minimum 

 and maximum of each month. It is remarkable that the amount of 

 this variation is less in July than in either of the preceding or fol- 

 lowing months, a circumstance which seems to have been first ob- 

 served by Colonel Beaufoy, in 1318, and which was confirmed by 

 the observations of five years. It appears somewhat improbable 

 that such a coincidence should be accidental. 



1834. 



November, - 

 December, - 



1835. 



January, 



February, 



March, 



April, 



4' 51 

 3 9 



4 27 

 2 52 



5 25 

 7 33 



May, 



une 



J 



July, 



August, 



September, - 

 October, 



November, - 



- 12' 10" 



- 11 11 



- 10 3 



- 13 30 



- 14 3 



- 10 39 



- 7 39 



The following table will show to some extent how far these chan- 

 ges are connected with variations of temperature. 



