126 Disturbance of the Needle during an Aurora. 



parison with the variation on the 10th, is assisted in the case of the 

 needle out of doors by the curve of mean variation being traced in a 

 dotted line upon fig. 1, plate I. In following the curves we see the 

 increase from the morning minimum, at, or before, 8J A. M., to the 

 day maximum at 3, P. M., and the subsequent descent towards the 

 evening minimum, with an irregularity in the progress of the varia- 

 tion, observable at 7, P. M. on the 10th of July. The lines now 

 cease to have even the most general resemblance. The descent to- 

 wards an evening minimum at 9, P. M., which appears upon the line 

 of mean variation, concurring with the descent attending the aurora, 

 might be supposed to have produced the suddenness of the effect no- 

 ted, were it not that a similarly rapid fall occurs between half past 

 eleven and twelve o'clock, contrary to the direction of the line of 

 mean variation. It is possible that the coincidence in the direction 

 given by the diurnal variation with that of the change attending the 

 aurora, between eight and nine, and their opposition between half past 

 eleven and twelve, is the cause of the first minimum being lower than 

 the second, supposing this to have been at 12 o'clock, since the differ- 

 ence between the second minimum and the mean is greater than that 

 between the first and the mean, by four minutes. At 1, A. M. of the 

 11th, the needle appears to have regained about the mean position, 

 the variation being rather more than the proportional mean for that 

 hour, supposing the increase from twelve to one, to be the same as 

 that from eleven to twelve. Midnight is about the hour of the night 

 maximum, which, however, not unfrequently occurs, after as well as 

 before, that time. 



The conclusions deduced from the observations on the needle 

 within doors, coincide, generally, with those obtained from the one 

 without. 



Having made material improvements in the needles referred to in 

 this paper, and completed two small observatories in the yard attach- 

 ed to my dwelling, expressly for magnetic observations, I purpose to 

 follow out this subject, and in connexion with it to study the effect 

 of meteorological changes, and, as necessary to the solution of these 

 questions, to preserve an account of the diurnal variation, in such a 

 way that the requisite corrections for the temperature of the needles 

 may be applied. Thus I hope to be able to make a contribution to- 

 wards determining a question to which the attention of men of science 

 has been particularly called by the recent discussions in the British 

 Association for the Advancement of Science. 



Philadelphia, December 24th, 1833. 



