, „ prjofessor fobbes's expekiments on 



19. In the case of No. 1., the magnetism has been considered as stationary 

 throughout the period 1832-1835, with which we arc now concerned. In the 

 case of the Flat Needle, this cannot be assumed, nor can we admit the change to 

 have been uniform. It seems probable that much movement, and especially al- 

 ternations of temperature, accelerate the loss of magnetism, that loss having been 

 (greatest in 1832, when most of the following obsei-vations were made. This is 

 confirmed by a more minute inspection. Observations were made at Geneva on 

 the 20th August 1832, imd again on the 10th November, and between these dates 

 the whole of the alpine series is contained : Now the vai-iation of the logarithms 

 for that period is no less than .00452, or at the rate of .02001 per annum ; whilst 

 we have seen that dm-ing the period from June 2. 1832 to May 7. 1833, which 

 includes the above, the mean change was only at the rate of .00736 per annum. 

 It is clear then that, in order to render the observations of 1832 comparable with 

 one another, we must assiune a much higher rate than the mean, for the months 

 from June to November. Admitting some Mttle doubt as to the Geneva compa- 

 risons due to the monthly change of intensity, and the great difference of tem- 

 perature in the two cases ; I think that I shall best satisfy the conditions by as- 

 suming the log. time to have increased by .00100 for each month from June to 

 November, leaving .00684 — .00500 = .00184 for the whole of the remaining six 

 months down to May 1833, during which the needle was in a state of almost per- 

 fect repose. 



20. The mode of allowing for this is the following. All determinations of 

 intensity are relative, referring to some intensity as a standard ; but I have taken 

 the horizontal intensity at Paris as unity (which is to that at the magnetic 

 equator as 4788 to 10,000 according to Humboldt).* Hence, since the squares of 

 the times of 100 vibrations are inversely as the magnetic forces, the terrestrial 

 horizontal intensity at a station A is to that at Paris, or 1, as the square of the 

 time observed at Paris (which time we may call Tp) is to the square of the tune 

 obscr^-ed at station A (or T^). Hence, 



Intensity at A = ( y^ ) 



If the magnetism of the needle change, we must therefore find by interpolation 

 the time of vibration at Paris for the particular epoch of observation. 



same external case in which they came fi'otn Norway. This arrangement I have not changed, but in 

 packing tliem I have taken pains to place the opposite poles nearest one another, an arrangement which 

 seems to have been attended with good effect ; and to shew that needles may lie within an inch or two of 

 one another without material injury, when we see the stationary condition of No. 1, and the diminishing 

 rate of variation of the " Flat" Needle. 



• Deduced from the measure of total intensity 1.3482 at Paris, given in the Memoires d'Arceuil, 

 torn. i. multiplied by the cosine of the dip (there also given) 69° 12'. 



