January 12, 1922] 



NATURE 



47 



the use of a current meter, but that other factors 

 also come into play. Otherwise it would be diffi- 

 cult, for instance, to understand how certain pelagic 

 species of fish {Myctophum glaciale and M. 

 Dofieini, Stomias boa, etc.) can at all seasons occur 

 in far greater quantities — have a maximum of 

 lensity — in the Alboran Sea (the westernmost part 

 F the Mediterranean, between Spain and Morocco) 

 than either west or east of there, despite the fact 

 that the surface layers are in constant move- 

 ment towards the east. This is actually the case, 

 as was first shown by A. V. Tanins: and Vilh. Ege 

 on the basis of material from the Thor expeditions. 

 The Dana Expedition has proved the same thing. 

 ■Comparatively few specimens occur west of Gib- 

 raltar and east of Oran, but in the Alboran Sea 

 itself great quantities of all three species were 

 found, so that the contents of a single net might 

 ^how, for instance, more than 1500 specimens, espe- 

 • ially Myctophum glaciale. 



In conclusion T cannot refrain from emphasising 

 the extreme importance an intensive studv of the 



Straits of Gibraltar and adjacent waters would 

 have for general — physical and biological — ocean- 

 ography. When, at the commencement of October, 

 I was obliged to leave this area in order to take up 

 the other tasks allotted to the Dana Expedition, it 

 was with the conviction that the expedition would 

 in all probability have l^een able to do more for 

 the cause of oceanography in general by keeping 

 station at Gibraltar dining the ten months we have 

 for work, than by ^ruising about the ocean. Being 

 so convinced, I venture to hope that British 

 naturalists may soon take up this important task, 

 which Great Britain, with Gibraltar as a base, has 

 unique opportunities for dealing with. A research 

 vessel stationed at Gibraltar would take but half an 

 hour to arrive on the scene of operations, the meet- 

 ing-place of two deep seas. The saving in time 

 and coal, and the unparalleled opportunities of 

 utilising all favourable weather conditions for 

 oceanographical Avork, are self-evident. 



(On board the Dana, at San Vicente, Cape Verde 

 Islands, November t, 1921.) 



Photographic Studies of Heights of Aurora. 

 By Dr. C. Chree, F.R.S. 



THE two publications referred to below, 1 by 

 l*rof. Carl Stormer, of Christiania, merit the 

 attention of all interested in the physics of the 

 atmosphere. As is generally known, Prof. Stormer 

 discovered a satisfactory method of measuring the 

 height and position of aurora by means of photo- 

 graphs taken simultaneously at the two ends of 

 a long base. The photographs include two or 

 more stars, the exact positions of which in space 

 are ascertainable, the precise time of taking the 

 photographs being known. The difference between 

 the positions of the aurora relative to the stars in 

 the two photographs enables the necessary cal- 

 culations to be made. 



The first memoir gives a very full account of 

 photographs taken in the spring of 1913 at two 

 Norwegian stations, Bossekop (B.) and Store 

 Korsnes (K.), 27-5 km. apart, near latitude 70° N. 

 •Some of the results have been already discussed in 

 a series of papers enumerated on p. 7, which have 

 appeared in different publications, especially 

 J errestrial Magnetism and Electricity, the 

 Astrophysical Journal, and the Paris Cotnptes 

 rendus. But the present memoir, besides sum- 

 marising these, contains much new matter. In 

 chaps. I and 2, pp. 8-37, there is a description 

 of the apparatus and equipment and of the methods 

 of observation. This is intended to be sup- 

 plementary to descriptions already given, but de- 

 scribes various improvements and simplifications. 

 Chap. 3, pp. 38-156, is a complete journal in 

 chronological order of all the 336 pairs of photo- 

 graphs discussed. Besides the date and hour and 

 lime of exposure, values are given of the parallax 



1 Carl Stormer : " Rapport sur une expedition d'aurores bor<5ale.« k 

 Bossekop et Store Korsnes pendant le printemps de Tannic ign." Geofysiske 

 iPublikationer, vol. i, No. 5. Pp. 269-(-ioa plates. (Kristiania, ro2i.) 



" Kxemples de rayon* aur ^raux dipassant des altitudes de 500 kilometres 

 aii-dessus d« la terre." Geofysiske Publikationer, vol. 2, No. a. Pp. 5 + a 

 f>lates. (Kristiania, 1921.) 



of each selected auroral point [i.e. the angle sub- 

 tended at the point by the 27.5-km. base), its astro- 

 nomical co-ordinates (altitude and azimuth), and 

 several calculated data, including the height of 

 the point above the ground, and the distance from 

 Bossekop of the point itself and of the correspond- 

 ing point on the earth's surface vertically under it. 

 The vertical heights vary from 87 to 323 km., 

 the horizontal distances from Bossekop from 5 to 

 780 km. Some of the more notable auroras are 

 discussed in considerable detail. The 336 pairs of 

 photographs appear in plates i to 28, each plate 

 containing twelve B. (Bossekop) and the corre- 

 sponding twelve K. (Korsnes) photographs. To 

 each pair of photographs there answers a diagram 

 showing the stars used in the calculations, the 

 positions of the auroral points, usually distin- 

 guished by numerals, and dashed lines to represent 

 the parallaxes. 



We reproduce two pairs of B. and K. photo- 

 graphs. One (Fig. i) represents an auroral 

 curtain in which twenty-one points were measured. 

 The nearest point (towards the apparent tops of 

 the photographs) was at a horizontal distance of 

 99 km. from Bossekop, the most remote point 

 (near the lower left-hand corner) at a distance of 

 265 km. The heights measured varied from gp 

 to 130 km. Fig. 2 represents a band having the 

 right-hand edge exceedingly sharp and luminous. 

 The twelve points measured are shown in the key 

 diagram (Fig. 3). Their heights varied only from 

 102 to 108 km. The horizontal distances from 

 Bossekop of points i and 12 were respectively 61 

 and 178 km., and their parallaxes were 13- 1° and 

 7-7°. The stars used were a, /3, and ^ Aurigae. 

 C, and Co represent the positions relative to the 

 stars of the centres of the plates for Bossekop and 

 Korsnes. The other details as to the stars refer 



NO. 2724, VOL. 109] 



