May 24, 1888] 



NATURE 



85 



tion of a low segment near the horizon, a dark segment 

 was produced by contrast. Sometimes the luminous zone 

 was broken, and then dark spots or irregular spaces were 

 produced in the same way. These dark spaces were 

 frequently seen tinted with a faint rosy light. 



(5) The movement of the arches is ordinarily reported 

 to be from north to south, at places situated to the south 

 of the maximum zone, and, from the opposite direction, 

 at places within the maximum zone. Thus, at different 

 stations between the latitude of Rome and the latitude of 

 Bossekop, 696 per cent, of the auroral arches have moved 

 from the north ; at Mossel Bay, Franz-Josef Land, and 

 Discovery Bay, on the contrary, 62-5 per cent, have moved 

 from the south. At Cape Thordsen it was of course expected 

 that the most part of the auroral arches would move from 

 the south. Yet this was not the case. On the contrary, 576 

 per cent, moved from the north. The movements were, of 

 course, almost the same in both directions. 



(6) The anomalous forms of arches were very frequent, 

 and were made a matter of accurate investigation. Some- 

 times an auroral arch presents the form of a sinuous band, 

 or resembles a brilliant curtain with deep folds. At other 

 times the arches appeared as regular spirals. Seen from 

 the outside of the earth, or from above, the spirals were 

 almost all contorted in a direction contrary to the motion 

 of the hands of a watch, and the undulations folded as 

 an S. The motion was, in 80 per cent., from west to east. 

 The folds of the auroral draperies had very different 

 dimensions on different occasions. Sometimes a regular 

 arch showed only a slight undulation ; at other times, 

 only a part of an immense auroral drapery was seen 

 above the horizon, as a pseudo-arch. 



(7) Often, waves of light are running along the arches, 

 and then the rays or beams are apparently in vivid 

 motion. This appearance of the aurora is known in 

 England as " the merry dancers." In 103 cases the waves 

 were running from west to east, and in 101 cases from 

 east to west. The mean angular velocity per second was 

 38'6. For a mean vertical height of the aurora of 100 kilo- 

 metres above the earth's surface, or 222 kilometres from the 

 observer's eye, this gives the immense velocity of about 

 2-5 kilometres per second. The light of the aurora was 

 often suddenly changing as to the distribution and in- 

 tensity of light, but the geometrical form of the whole 

 phenomenon was only slowly varying. The rays were 

 sometimes observed to have a slow proper motion from 

 west to east, or vice versa. 



(8) As to the classification of the auroral forms, the 

 author rejects that of Weyprecht. The different forms 

 of the aurora in the classification of Weyprecht are, in 

 fact, only different views or projections, as, for instance, 

 the forms III. = beams or rays, and IV. = corona. The 

 corona results, according to the rules of perspective, when 

 a large number of separate beams parallel to each other 

 and to the direction of the dipping-needle seem to con- 

 verge to one point, viz. the magnetic zenith. A regular 

 and fully-developed arch consists, as we have said before, 

 of a long fringe of rays, and so on. The author considers 

 only two different forms of auroral light, viz. zones, or 

 horizontal layers of light ; and arches, composed more 

 or less of distinct rays parallel to the dipping-needle. 

 The arches present themselves in four different condi- 

 tions : (1) arch, or a regular band ; (2) band, or drapery; 

 (3) spiral ; and (4) pseudo-arch. 



(9) The light of the aurora is, according to the author, 

 of two kinds: (1) the yellow light, entirely monochro- 

 matic, and showing in the spectroscope the well-known 

 yellow line of Angstrom ; (2) the crimson or violet light, 

 resolved in the spectroscope into several rays and bands, 

 spread over all parts of the spectrum. In the following 

 table we give (I.) the lines observed by the author, (II.) 

 the lines observed by several authors before the year 

 1884, and (III.) the spectrum of lightning, according to 

 the observations of Herschel, Vogel, Schuster, and the 



author. The unity for wave-length is, as usual, the 

 o-ooooooi of the millimetre. 



1. 

 6306 ± 

 5776 ± 

 5664 ± 

 5568 ± 



5353 ± 

 5264 ± 

 5228 ± 

 5001 ± 



4837 ± 

 4707 ± 

 4642 ± 

 4236 ± 



7'3 

 30 

 30 

 i-6 



3o 



25 

 27 

 42 

 107 

 5"i 

 3 '3 

 67 



11. 

 6294 ±64 

 5776 ± 30 

 5664 ± 30 

 5570 ± 09 



5353 ± 3'3 

 5280 ± 1 8 

 5226 ± 3*2 

 5003 ±27 

 4862 ± 1-5 

 4702 ±29 



4636 ±2-4 

 4286 ±44 



III. 



6300 



5685 



5338 

 5260 



5004 

 4860 



4632 



There were twelve other extremely faint auroral rays 

 to be seen occasionally, but their position could not be 

 exactly observed. 



As to the further discussion of the different auroral 

 spectra and their supposed connection with different 

 auroral forms, we must refer to the original paper. 



(10) No sound was ever heard from the auroral light. 

 The feeble rustling noise sometimes heard was observed to 

 come from the loose agile surface-layer of snow driven to 

 and fro by the lightest wind over the underlying layers. 

 Nor was a " smell of sulphur" observed. 



(1 1) As to the height of the aurora, it may first be men- 

 tioned that the aurora was never seen to descend below 

 the mountains or the lower clouds. Only two or three 

 times it is possible that the light was seen below the 

 upper clouds. Yet sometimes the auroral light was seen 

 to be reflected from the surface of the snow. Direct 

 measures of the parallax from the end of a short base 

 (573 metres), by means of auroral theodolites of Mohn's 

 construction, gave an average height of 551 kilometres ; 

 from observations of the corresponding amplitudes and 

 heights of the arches, according to Bravais' method, 

 577 kilometres ; and by several other observations and 

 calculations, about 60 kilometres was found to be the 

 probable mean height of the aurora. 



(12) As to the annual and diurnal periods of the aurora, 

 no annual variation in the frequency could be proved. 

 The apparent daily period gave a maximum at 8h. 50m. 

 Gottingen time, or 9h. 13m. local time, in the evening ; and 

 a minimum at exactly the same hour in the morning. 

 This apparent period must be corrected for the influence 

 of the quantity of clouds and for the influence of the 

 twilight. If F represents the apparent frequency of the 

 aurora, and Q the quantity of clouds in tenth parts of the 

 whole sky, there was found F = 1 - 0*0730 Q, in taking 

 for unity the apparent frequency when the heavens were 

 totally clear. 



Further, the apparent frequency when the sun was 

 io° 47' below the horizon was the half of the true fre- 

 quency, and the influence of the sun's light was sensible 

 as far as to a depth of the sun of 17 45' below the hori- 

 zon. Once only the aurora was seen when the sun was 

 not more than 5 25' below the horizon. 



Taking into account these sources of error, the true 

 daily range has a maximum at 3I1. 3m. p.m., and a 

 minimum at 8h. 3m. a.m. local time. 



Finally, there was also a well-marked daily range in 

 the form of the aurora. The most brilliant phase of the 

 phenomenon occurred at 4I1. p.m. ; the aurora then ap- 

 peared as a complete regular arch. On the other hand, 

 the minimum brilliancy took place at 9I1. a.m. ; the arches 

 then were resolved into whirling fragments. 



Upsala, April. H. Hildebrandsson. 



NOTES. 

 The general arrangements for the Bath meeting of the British 

 Association have now been made. The first meeting will be 

 held on Wednesday, September 5, at 8 p.m. precisely, when 



