48 



NATURE 



[January 12, 1922 



to the Bossekop photograph. The arrow-heads 

 radiating from each star show tlie directions of the 

 ?tar's declination (5) and altitude {h) circles and 

 if the auroral parallax (/?). The numerals 39, 43, 



K. B. 



Fig. I.— Auroral curtain photographed at Korsnes and Bossekop. 



46 represent degrees of altitude. The azimuths of 

 the stars, i8o°-89-5°, etc., are also shown. The 

 dotted lines represent the parallaxes in magnitude 

 and direction. 



Fig. 2.— Auroral band photographed at Bossekop and Korsnes. 



The 33G key diagrams occupy plates 33 to 87. 

 Finally, there are charts numbered i to 32, in- 

 cluded in plates 90 to 104, which show the geo- 

 graphical positions of the points on the earth 



NO. 2724, VOL. 109] 



vertically under all the auroral points dealt with. 

 The observational data are thus presented in 

 an extremely systematic fashion. Chap. 4,. 

 PP- ^.S7~2i2, includes a mathematical investigation 

 of the variation in the inclination of the trajectory 

 of an electrified corpuscle to the direction of mag- 

 netic force, the magnetic field having a potential. 

 In the applications the earth's field is supposed to 

 be given with sufficient accuracy by the first-order 

 Gaussian terms. The corpuscle is supposed to 

 have emanated from the sun, and its course is 

 considered after it has come to within 500 km. of 

 the earth's surface. The inclination of the tra- 

 jectory to the magnetic lines of force tends to 

 increase as the corpuscle approaches the earth. If 

 the angle attains to 90° the corpuscle retreats. 

 The energy may be absorbed while the corpuscle 

 is approaching or while it is retreating, or some 

 may remain after the retreat has carried the cor- 

 puscle outside the atmosphere. What happens is 



Fig. 3. — Measurements of points on Fig. 2. 



shown to be largely dependent on the constitution 

 of the atmosphere. The author assumes that 

 throughout the troposphere, the upper limit of 

 which is put at 10-5 km., air is constituted exactly 

 as at the earth's surface. But higher up in the 

 stratosphere the several constituents, nitrogen, 

 oxygen, argon, neon, helium, krypton, xenon, and 

 hydrogen, behave independently. The author 

 essentially follows Mr. Jeans, rejecting Wegener's 

 hypothetical gas geocoronium. According to his 

 calculations, p. 173, helium is the most important 

 gas between heights of 110 and 200 km., hydro- 

 gen preponderating at greater, and nitrogen at 

 lower, levels. These are practically the only gases 

 that count when we consider how far down an 

 auroral ray can come in the atmosphere. Separate 

 calculations are made of the absorption of cathode 

 and )8-rays on one hand, and of a-rays on the 



