172 COSMOS. 



When the needle, by its sudden disturbance in its horary 

 course, indicates the presence of a magnetic storm, we are 

 still unfortunately ignorant, whether the seat of the disturbing 

 cause is to be sought in the earth itself, or in the upper regions 

 of the atmosphere. If we regard the earth as a true magnet, 

 we are obliged, according to the views entertained by Friedrich 

 Gauss (the acute propounder of a general theory of terrestrial 

 magnetism), to ascribe to every portion of the globe measur- 

 ing one-eighth of a cubic metre, (or, 3^ of a French cubic 

 foot) in volume, an average amount of magnetism equal to 

 that contained in a magnetic rod of lib. weight.* If iron 

 and nickel, and probably also cobalt (but not chrome, as has 

 long been belie ved),f are the only substances which become 

 permanently magnetic, and retain polarity from a certain 

 coercive force, the phenomena of Arago's magnetism of 

 rotation and of Faraday's induced currents show, on the 

 other hand, that all telluric substances may possibly be made 

 transitorily magnetic. According to the experiments of the 

 first-mentioned of these great physicists, water, ice, glass, 

 and carbon affect the vibrations of the needle entirely in the 

 same manner as mercury in the rotation experiments. J 



lately that I discovered, for the first time, that as early as at the close of 

 the sixteenth century, and consequently hardly twenty years after Robert 

 Norman had invented the inclinatorium, William Gilbert, in his great 

 work De Magnete, proposed to determine the latitude by the inclination 

 of the magnetic needle. Gilbert (Pliysiologia Nova de Magnete, lib. v. 

 cap. 8, p. 200) commends the method as applicable " aere caliginoso." 

 Edward Wright, in the introduction which he added to his master's 

 great work, describes this proposal as " worth much gold." As he fell 

 into the same error with Gilbert, of presuming that the isoclinal lines 

 coincided with the geographical parallel circles, and that the magnetic 

 and geographical equators were identical, he did not perceive that the 

 proposed method had only a local and very limited application. 



* Gauss and Weber, Resultate des magnet. Vereins, 1838, 31, s. 146. 



t According to Faraday (London and Edinburgh Philosophical 

 Magazine, 1836, vol. viii. p. 178) pure cobalt is totally devoid of mag- 

 netic power. I know, however, that other celebrated chemists (Heinrich 

 Rose and Wb'hler) do not admit this as absolutely certain. If out of two 

 carefully-purified masses of cobalt totally free from nickel, one appears 

 altogether non-magnetic (in a state of equilibrium), I think it probable 

 that the other owes its magnetic property to a want of .purity; and this 

 opinion coincides with Faraday's vieAv. 



J Arago, in the Annales de Chimie, t. xxxii. p. 214; Brewster, 

 Treatise on Magnetism, 1837, p. Ill ; Baumgartner, in the Zeitschrijt 

 filr Phys. und Mathem., bd. ii. s. 419. 



