Aurora Borealis on the Magnetic Needle. 195 



correct, why was not the Aurora seen from 7** till lO'^ ? but 

 even admitting that it was the termination of an Aurora of the 

 day, why did the magnetic needle at Paris not feel its influ- 

 ence on the evening of the 17th ? 



Another case of exactly the same kind occurred on the 4th 

 November. An Aurora of' great beauty^ with numerous and 

 very bright rays, appeared at Leith in the evening ; but the 

 magnetical needle at Paris is then perfectly quiescent. From 

 nine till two p. m. it was much agitated, and M. Arago again 

 supposes that this " was most likely the remains of an Aurora 

 of the day." 



2. We next proceed to a prediction which is thus announ- 

 ced : " In this same month of August, on the night of the 

 2lst, the morning of the 22d, the night of tfie ^6th, and par- 

 ticularly on the night of the 29th, great anomalies were ob- 

 served in the extent of the oscillations of the needle. On all 

 these occasions the sky was I believe clouded at Leith. If not, 

 and the observers there did not see the Aurora, for instance, 

 on the night of the 29th of August, we will be obliged to ad- 

 mit that there exist other causes of which we are still ignorant, 

 which excite considerable influence over the magnetic needle." 



" In reply to this remark," says Dr Coldstream,* " I may ob- 

 serve, that the notes in our Journal of the state of the sky on 

 the night of the 21st, and morning of the 22d, are not satis- 

 factory ; that the night of the ^6th was particularly clear, with 

 bright moonshine, and that much cloud prevailed on the 29th, 

 and so that had an Aurora existed we could not have seen it." 

 Here, then, we have an agitation of the magnetical needle at 

 Paris on fmr days, the 21st, 22d, 26th, and S9th August 

 1825, and yet Aurorae, so far as we can find, were seen 

 neither in Scotland or England. The night of the 26th was 

 particularly clear, and though clouds prevailed on the 29th at 

 Leith, yet the predicted Aurora, had it existed, might have 

 been seen elsewhere, and must have been recorded. Before 

 concluding this part of the subject, we may add, that an Au- 

 rora, " which played with considerable brilliancy," was seen 

 at Leith at 10"^ p. m. on the 11th September, and we believe 



• See this Jovrnal, No. x. p. 90. 



