336 



Prof. Hansteen on the Polar Lights, 



The hours mentioned are after the chronometer giving the 

 time of Christiania, to which we must add about one hour to 

 make it equal to that of Torneo. — By a previous investigation 

 on the influence of temperature on the time of oscillation of 

 this cylinder, I have found that when the normal temperature 

 to which all observations are reduced is fixed at = /, the tem- 

 perature during the observation at = t', the time of oscilla- 

 tion observed at = T, the reduction of the time of oscillation 

 to the normal temperature is = — T . {t' — t) . 0*00047 *.) If 

 we reduce the above time of oscillation, after this formula, to 

 the mean temperature t = + 80°-8, we obtain the following 

 time of 300 oscillations with Dollond's cylinder in a tempera- 

 ture of + 80°*8, at Haparanda, near Torneo. 



If the evening observation of the 26th of August is omitted, 

 we see that in the mean of Torneo, the least intensity (least 

 time of oscillation) takes place between 9 o'clock A.M. and 

 10 o'clock P.M., and the greatest intensity (shortest time of 

 oscillation), in the evening, as at Christiania. The mean of 

 the three times of observation is =88l""79 



26th Aug. 9^ P.M. was found =887^8^ 



Irregularity .... 6'''04 



* This reduction is found different for cliffercnt cylinders, and, as it ap- 

 nears, larger for those of smaller than those of irreater diameter. 

 ' ^ The 



