22 5 
influence in the solar rays. 
place, could not be ascertained with precision ; and the much 
greater difference between either of these and the other two 
may possibly be attributable to the same source, though I shall 
have occasion at the conclusion of this paper to point out 
another, and, I conceive, the true cause of this difference. 
Had my object been to give a correction for difference of 
temperature, to be applied generally to the intensities de- 
duced from the times of vibration, a closer agreement in the 
results would have been desirable ; but although such is not 
my object, I wish, from these experiments, to point out the 
necessity of such correction, when we would deduce the 
magnetic intensity from the times of vibration of a needle in 
different parts of the earth, where the temperatures during 
the observations are necessarily different. 
The observations which I have detailed are, I think, quite 
conclusive as establishing the fact, that the rays of the sun 
had a tendency to check the vibrations of the needle, parti- 
cularly those in the last table ; since here, on the needle 
being exposed to the sun, the terminal arc was reduced from 
nearly 14 0 to 8±°, and again on excluding its rays that arc 
was increased to nearly 1 3 0 : the only question is whether 
this would take place under all circumstances ; that is, whe- 
ther the influence be exerted in all cases directly on a mag- 
netised needle and independently of changes in its tempera- 
ture and intensity, or whether it was an effect produced on 
the metal of the box by a change in its temperature. That 
the effect is independent of changes in the temperature and 
intensity of the needle will appear from comparing the first 
observation in the sun with the first of the second set in the 
shade ; the temperatures and times of vibration are very 
