TERRESTRIAL MAGNETIC INTENSITY. 23 



however, could hardly fail to be sensible, the variation of temperature being no less 

 than 36° Cent., and as cold tends to increase the apparent intensity, if no such in- 

 crease was observed, it might plausibly be argued that the real intensity had di- 

 minished. It must, however, be observed, that the observations lasted only in 

 general from one to two minutes, and that in so short a period (and depending on 

 a single value of the elapsed time) the acceleration due to the above-mentioned 

 cause would hardly be perceptible. Taking the mean result of the effect of tem- 

 perature ascertained by myself for No. I. and " Flat" needles, we find the factor 

 — .00037 applicable to the time for a decrease of 1° R. of temperature, which 

 agrees exactly with Hansteen's mean correction. If we apply this to Gay Lus- 

 SAC's observation we find a correction of — .0108 as a factor applicable to the 

 time, for the effect of — 36° Cent, of temperature. Yet large as this is, amount- 

 ing to i^tli part of the whole, the discrepancies of observation often amounts to 

 double that quantity.* StiU we admit with M. Kupffer that the probability de- 

 ducible from M. Gay Lussac's observations, is in favour of a slow diminution. 



39. The next series of observations includes those of Humboldt and Gay 

 Lussac, recorded in the Memoir es d'Arcueil,-\ which include observations made in 

 the Alps, though at no great heights ; and here no particular influence of height 

 was observed, nor was indeed looked for. t 



40. Since that period the subject seems to have met with Httle practical at- 

 tention, until the recent publication of M. Kupffer's " Voyage au Mont Elbrouz" 

 by the Petersburg Academy. From his observations with a needle by Gambey, 

 half a metre long, M. Kupffer attempts to deduce, not only the fact, but the 

 amount of the diminution with height, and this upon the authority of a single ex- 

 periment, and at no considerable elevation. § In fact, all the intricate corrections 

 which this delicate observation requires were little more than guessed at. The 

 difference of geographical position of the two stations (12' in lat. 38' in long.) was 

 allowed for by observations made with a different apparatus, — the effect of tempe- 

 rature was deduced from indirect experiments, far from presenting a mutual agree- 

 ment ; and the whole difference of level (4500 French feet) offered a very small 

 basis for so general a conclusion. But, besides this, there is an oversight in M. 

 Kupffer's deductions (first pointed out to me by Professor Necker of Geneva), 

 which tends yet farther to diminish the probability of his conclusions. The esti- 

 mate of the effect of geographical position on the magnetic intensity, M. Kupffer 

 conceives to be such, that the variation for 12' in lat. (diminishing from the lower 

 to the upper station) would exactly counterbalance the variation due to 38' in E. 



* See the details of the Observations in the Annales de Chimie. An. xiii. (1805), Tom. lii. p. 75. 

 t Tom. i. p. L J Ibid. p. 10. 



§ The observations were not made at the summit of Mont Elbrouz, as stated in the Annuaire du 

 Bureau des Longitudes, 1836, p. 288, but near the foot of it, and the difference of height of the stations 

 was less than 5000 English feet. The stations were " Pont de Malka," and " Hauteur de Kharbis." 



