MAGNETIC OBSERVATIONS. 99 



care is taken to obtain corrections for temperature in the 

 needles that are employed both before and after the ascent. 

 The neglect of such a correction has led to the erroneous 

 result deducible from Gay-Lussac's experiments, that the 

 magnetic force remains the same to an elevation of more 

 than 22,000 feet, 14 whilst conversely the experiment showed 

 a decrease in the force on account of the shortening of the 

 oscillating needle in the upper cold region. 15 Faraday's 

 brilliant discovery of the paramagnetic force of oxygen must 

 not be disregarded in the discussion of this subject. This 

 great physicist shows that in the upper strata of the atmo- 

 sphere, the decrease in the intensity cannot be sought merely 

 in the original source of the force, namely the solid earth, 

 but that it may equally arise from the excessively rarefied 

 condition of the air, since the quantity of oxygen in a cubic 

 foot of atmospheric air must differ in the upper and lower 

 strata. It seems to me, however, that we are not justified 

 in asbuming more than this that the decrease of the para- 

 magnetic property of the oxygenous parts of the atmosphere 

 which diminish with the elevation and with the rarefac- 

 tion of the air, must be regarded as a co-operating modifying 

 cause. Alterations of temperature and density through the 

 ascending currents of air may further alter the amount of 

 this influence. 10 Such disturbances assume a variable and 

 specially local character, and they operate in the atmosphere 

 in the same manner as different kinds of rocks upon the 

 surface of the earth. With every advance which we may 

 rejoice in having made in our knowledge of the gaseous 

 envelope of our planet and of its physical properties, we at the 

 same time learn to know new causes of disturbance in the 

 alternating mutual action of forces, which should teach us 

 how cautiously we ought to draw our conclusions. 



The intensity of the terrestrial force, when measured at 

 definite points of the surface of our planet, has, like all the 

 phenomena of terrestrial magnetism, its horary as well as its 

 secular variations. The horary variations were distinctly 



14 Annales de Chimie, t. Hi, 1805, pp. 8687. 



15 Arago, in the Annuaire du Bureau des Longitudes pour 1836, 

 p. 287 ; Forbes, in the Edin. Transact, vol. xiv, 1840, p. 22. 



16 Faraday, Exper. Researches in Electricity, 1851, pp. 53, 77, 2881, 

 2S61. 



H 2 



