INTRODUCTION. XX1X 
August 22" 20". There being a considerable natural vibration of the bifilar 
magnet, with little or no change of mean position, the following estimations were 
made at the extremities of the ares. 
5 Vibrations. Mean. 
7 33-0 9 350 2 20 
7 56:0 10 0-0 2 40 
, 2m 25-9 
8 22:0 10 25:0 2 3:0 
8 47°5 10 49-5 2 20 
9 10:5 11 140 2 35 
Mean, 1 vibration = 24s-58. 
The are of vibration was 11’ during the first 9 vibrations, 9’ at the 10th, and at the 
end of the 11th the magnet stopped. 
Oct. 8° 3". The magnet vibrated by a knife. 
2 estimations, mean of 4 vibrations, 1 vibration 255-22 
2 So Ree rc ove 1 -- 26528 
8 estimations, as below. 
¥ 3 Vibrations, Mean, 
16-2 35:0 78:8 
42:8 0°5 777 : 
8:8 27-0 78-2 iat 
35:0 530 78:0 
1 vibration = 26:06. 
Oct. 11% 23. Natural vibration. 
2 estimations, mean of 11 vibrations, 1 vibration 25:18. 
255 was used throughout in the observations. 
32. The point of the scale coinciding with the vertical wire of the fixed tele- 
scope is estimated to a tenth of a division, at 25° before the minute of observation, 
at the minute, and 25° after it ; the mean is deduced, as in the declination, from the 
a+2b6+e 
formula —j—, a, 6, and ¢ being the three readings. 
The mean thus obtained is corrected to the temperature of 26° Fahr., this 
being below the lowest temperature which has occurred within the observatory. A 
constant quantity of 300:0 has been added to all the readings, in order that the 
mean readings might be about 500; in this case the greatest changes will be con- 
tained within 0 and 1000. 
The values of 1 scale-division, in parts of the whole horizontal force, are given 
beneath ; they are also given at the foot of each page of the Daily Observations of 
Magnetometers. 
Tables have been formed giving the temperature-correction for every tenth of 
a degree Fahr. to a tenth of a scale-division, from the formula 
MAG. AND MET. OBS. VOL. I. h 
