Effect of an Aurora Borealis on the Magnetic JVeedle. 115 



the same line. The height of the thermometers contained in the 

 boxes with the needles is given. 



From the table just given it appears that on the 18th of May, the 

 westerly variation, as shown by the horizontal needle out of doors, had 

 two distinct points of minimum, the former at 9 A. M., and the latter 

 between 9J and 10^ P. M., and two points of maximum, the first at 

 3 P. M., and the second at some period, not ascertained, after 1 1 

 P. M. The same variation shown by the needle within doors, had 

 its minima at 11 A.M., and from 6 to 7 P. M., its maxima between 

 2J and 4^ P. M., and at some hour of the night which was not as- 

 certained. The temperatures of the two needles being very differ- 

 ent, the effect of changes of temperature should be ascertained, to 

 render the results strictly comparable ; a remark which suggests the 

 object, in part, of the observations upon these two needles, so differ- 

 ently situated. 



In the dip we find a minimum at 2J P. M., as the only point very 

 decidedly marked : there is an apparent maximum at 7, which the 



* As this needle was intended for ditferential results, the dip here recorded is not 



to be supposed the true dip for tliis plate. 



