METEOROLOGY. 



109 



sion is to be determined. For these observations he proposes a binoc 

 ular attached to a circle liaviiiff a horizontal axis wliich revolv(!s aroniid 

 a vertical axis, so that the I)iiiocular can liav<} the i)lane of its axes 

 brought into the plane of the auroral beams. Many observations show 

 that the auroral rays arranj^e themselves pre eminently along the mag- 

 netic j)arallels in curves tliat cut the magnetic in(;ri(lians at right angles; 

 they also show that tin; auroral phenomena as arranged along t.lic. mag- 

 n<;tic parallels follow the curves of equal total magnetic intensity. (Z. 

 0. G. M., XIX, p. 405.) 



358. ])r. II. EUama, of Haarlem, gives a summary of the observations 

 on tiie aurora made by the International Polar party sent by the gov- 

 ernment of the Netherlands to the Sea of Kara. The aurora was in 

 general very unsteady and variable, so that drawings could rarely be 

 made. Arcs frequently occurred, but soon changed to the forin of 

 bands; the dark band below the arc was never distinctly seen. The 

 aurora was invariably visible when the heavens were clear and the riiooti 

 not too bright; the Jinest appearances oc{;urre<l at about 1 <)..';() i'. m., 

 when the corona most frequently occurred; the bauds were strongest 

 about 9.30 p. M. The auroras were generally on the north side of the 

 heavens, and only reached to the south of the zenith after a very biight 

 display; the highest point of the arcs did not lie in the magnetic gia- 

 dient, but in the astronomical meridian. Out of 80 observed arcs and 

 GO bands the summits of 04 and 41, respectively, lay in the astronomi(;al 

 meridian. Ekama has comi)Uted the altitude of his auroras by the for- 

 mula given by Nordenskiold in his account of the Vega's wintering in 

 Behring's Sound, lS78-'70; the results of nine nights agree closely, and 

 give for the distanc<; of the arc above the, earth's surface O.O.'i.'J of the 

 earth's radius; for the radius of the nearly horizontal arc itself 0.".'J4 ; 

 and for the di.stanee of the center of this ( ircle below the earth's surface 

 0.02 of the earth's radius, which value agrees with that of Norden>kiold. 

 The auroras were gcjnerally very faint ; the arcs had no color. Ihe wave 

 length of the peculiar auroral line was determine<l by the spectro.scope 

 at 5.j0.."i;> miilionths of a millimeter; no other lines were visible; even 

 a wholly red aurora gave only this yellow-green line. No connection 

 was found between the i»olar bands of clouds and the auroral light. Out 

 of 2().'3 j)olar bands the vanishing or converging points wc-re distributed 

 as follows: 



The direction of the polar bands j>robably depends on the direction 

 of the wind prevailing at the cloud level. IZ. 0. O. .U., xix, p. 482.) 

 S. Mis. ?/i 24 



