3G8 SCIENTIFIC RECOKD FOR 1884. 



4. As soon ns tho oartli curroiits in l)t)tli linos hoooino stronij, the ninjj- 

 netic iiistiHunonts doviato from tlioir ordinary positions, and these per- 

 turbations increase with the strength of the earth currents, but uot in 

 any regular proi)i>rtiou. 



5. Tlve changes in the south-to north line i)reccded by about live min- 

 utes the changes in variation of declination, and similarly the current 

 in the east-to-west line preceded the changes in the horizontal intensity, 

 so that the earth current would a[)pear as the primary cause. He con- 

 cludes that the earth currents are the primary cause of many pertur- 

 bations, but not the cause of the periodic variations in the nuignetic 

 elenunits. (Z. 0. M. 6\, xix, p. 510.) 



353. Prof. K. Schering, of Gottingen, criticising Professor Wild's re- 

 sults, states that the method is not thoroughly reliable, and expresses 

 his conviction that we cannot determine a dilterence of potential b(^ 

 tween two points of the earth, nor even ])rove its existence, if we use 

 ground plates which by contact with the earth become electritied in any 

 unknown manner. He recommends the method introduced by W. Sie- 

 mens in the iirogrammeof the ("Jerman North Polar expedition for ISSl*- 

 '83 as the only one to be used in determining the currents induced by 

 the changes taking place in the magnetic force of the earth. [Z. 0. O. 

 M.y XIX, p. 552.) 



354. The great magnetic storm and auroral display of November 17, 

 ISSlI, is abundantly discussed, with numerous contributions of observa- 

 tions, throughout vols, xxvii and xxYiii oi' Xafure. 



355. T. W. liackhouse, of Sunderland, collects together all accessible 

 observations on the spectrum of the aurora, and- shows that the nuniber 

 of striking coincidences between it and the nu)dified air spectrum make 

 the snggestion that they are identical one worthy of consiiicration. 

 (Nature, XXVIIT, ]). 20!).) 



356. O. Jesse discusses the auroral arc observed October 2, 1882, and 

 places it at an altitude of 122.2 kilometers, with a j^robable error of 4.5 

 kilometers. (Z. 0. O. M., xviii, p. 238.) 



357. O. Jesse, in some remarks on the determination of the altitude 

 and position of the aurora, states that the very rapid changes going on 

 during" a dis})lay make it important to have a method of determining its 

 altitude by observatious made at one i)lace. lie has, therefore, devel- 

 oped such a method, which is published in full in the Astronomische 

 Nachrichteu, No. 2510. This method is based npon the assnmi)tion that 

 the rays constituting the aurora lie on the surface of a cone whose apex 

 [s in the interior of the earth, where the direction of any one ray and the 

 oscnlating magnetic axis of the earth intersect each other. The ai)])li- 

 catiou of this nu'thod requires that we should measurxi for as many rays 

 as possible the angle under which the ray or its prolongation intersects 

 the horizon; also, the azimuth of the point in which the ray cuts (ho 

 horizon, together with an accurate determination of the time; further- 

 more, for some of the rays the apparent position of their highest exten- 



