November 17, 1916] 



SCIENCE 



721 



southeast of Charlottetown and on Prince 

 Edward Island at approximately latitude 

 46° 6', longitude 62° 53'. The auroral stream- 

 ers near the zenith may be regarded as ap- 



observed to remain visible in the laboratory 

 for 20 minutes. There is much evidence in 

 favor of the view that the meteor train is 

 phosphorescent nitrogen and formed in the 



-''r'-A~''. - '^^ % ~" '■ ""■,'" 



Fig. 3. The altitude of aurora borealis seen from Bossekop during the spring of 1913. Eaeh 



calculated attitude is marked by a dot and the several hundred simultaneous photographs of 



aurora from the stations — Bossekop and Store Korsnes — (mutual distance 274 kilometers) gave 



about 2,500 determinations of height, which are seen above. (Beprinted from Nature.) 



proximately parallel and their apparent focus 

 is of course due to perspective. 



HEIGHT OF THE AURORA AND THE METEOR 

 TRAIN ZONE 



One of the results of the study of meteor 

 trains has been the discovery of a definite 

 meteor train zone between 50 and 70 miles' 

 altitude. "When certain large meteors pass 

 into this zone, a train is observed to remain in 

 the track, apparently composed of self-lumi- 

 nous gas and which frequently remains visible 

 for half an hour. Nitrogen has been found to 

 assume a true phosphorescent state similar to 

 the afterglow of zinc sulphide. 7 It has been 



'C. C. Trowbridge, Phys. Beview, Vol. XXVI., 

 June, 1908. 



same zone in the atmosphere which is suscep- 

 tible to electrical discharges and results in 

 the aurora. 



In Fig. 4, curve A, the heights of 2,400 ob- 

 servations of the aurora made by Carl 

 Stormer's expedition are shown. Curve B 

 shows the heights of the middle portion of 30 

 meteor trains, and curve C gives the heights 

 of the lower ends of 21 meteor trains. 



The initial intensity of gas phosphorescence 

 has been found to be proportional to the third 

 power of the gas pressure; 8 hence it is to be 

 expected that meteor trains would show a pre- 

 dominance at slightly lower elevation than 

 the aurora, as indicated in Fig. 4. 



s C. C. Trowbridge, Phys. Rev., Vol. XXXII., 

 February, 1911. 



