62 



Captain Lefroy's Second Report on 



Table I. 



I do not offer these instances as conclusive, but they are 

 somewhat remarkable ; and I may state, that, having marked the 

 Gottingen hour of the first appearance against every observation, 

 the great majority give direct support to the inference I have 

 drawn, and there are few or no instances contradicting it. The 

 question will soon be decided if the time of commencement of 

 each display is recorded, and a note also made of the latest hour 

 at which it may have been noticed that there was no aurora. 

 For example, the observer takes a look out at / p.m., no aurora; 

 again at 8, aurora, which is duly entered. In this connexion, 

 the fact that there was no aurora at 7 is almost as important as 

 the fact that the phsenomenon was visible at 8, and should be duly 

 entered. On the 29th of September, 1851, at &^ 30"^ p.m., there 

 was no trace of aurora at Toronto ; at &^ 36"*, a brilliant, heavy 

 serpentine band occupied the northern sky. In this instance, 

 and in various others, the time of appearance is fixed to five or 

 six minutes, and if at any northern station it happens to have 

 been fixed with anything like the same exactness, the question 

 will be answered. 



In March 1851, we have evidence of aurora every night save 

 three, 13, 17 (full moon), 19; these, however, were pretty gene- 

 rally clear nights. Registers for April 1851, have reached me 

 from a few stations only, but as far as they go give evidence of 

 aurora every night save four, 4, 14, 15 (full moon but clouded 

 everywhere), and 21. The 16th of December is the only instance 



