TERRESTRIAL MAGNETIC INTENSITY. 





Table 



II. 







laitial Semi- Arc = a = 10° 







Initial Semi- Arc = a = 20° 



No. of Vibrations observed = n = 100, 







No. of Observations 



observed = n = 300 



m Additive Log. 







m 



Additive Log. 



70 9.99978 







70 



9.99969 



80 9.99976 







80 



9.99960 



90 9.99972 







90 



9.99953 



100 9.99970 







100 



9.99946 



110 9.99967 







110 



9.99939 



120 9.99966 







120 



9.99933 



130 9.99962 







130 



9.99926 



I have every reason to think this correction to be accurate on the whole : 

 the agreement of the two modes of observation being in general very close. 



15. III. Temperature. This extremely important correction it is very diffi- 

 cult to determine. Without an accurate estimation of it, it w6uld be vain to at- 

 tempt to decide whether or not the magnetic energy varies with height ; because 

 at great elevations the temperature being always diminished, the intensity would 

 appear too great (the magnetic energy in iron being increased by cold, and dimi- 

 nished by heat). I therefore endeavoured to compare the intensity of the needles 

 employed, within the range of temperatures usually observed. The apparatus 

 employed was of this kind. The needle was first allowed to take the temperature 

 of a heated room and vibrated. Then every thing else remaining the same (and 

 of course any local attraction which might exist being unaltered) the apparatus 

 was placed in a cylindrical glass jar, with ice in the bottom, placed in a dish of 

 ice, and covered with a glass plate also covered with ice. A steady temperature, 

 but little above the freezing point, was thus attained, and the oscillation again ob- 

 served. These experiments were repeated many times. One series was under- 

 taken at Geneva in October 1832, another at Edinburgh on four different days of 

 August 1834. Those for the needle. No. 1, were conducted with the most scru- 

 pulous care, nearly 5000 vibrations having been counted for this purpose alone. 

 One set was discarded as differing too much from the others, and the remainder 

 agreed very closely, although made under such different circumstances, and at 

 such different times. The result adopted for Needle No. 1. gave an increase in 

 time of .00045 for a diminution of temperature of 1° Reaumur, and vice versa ; for 

 the Flat Needle (determined from two concordant series, both observed at Geneva 

 on different days) .00030. From these results the following tables were calcu- 

 lated, giving the reduction in each case to 0° of Reaumur (which being the scale at- 

 tached to the instrument, was always observed in these experiments.) This seems 

 preferable to referring to any other arbitrary temperature, upon which observers 

 do not generally agree. 



