May 17, 1883] 



NA TURE 



61 



be confused for a moment by any one who has had an 

 opportunity of comparing them simultaneously. 



During the period which has elapsed since 1871 my 

 efforts have been bent on the closest study of the aurora 

 borealis, and an accidental discovery that a Geissler tube 

 may produce certain forms of light without being in direct 

 connection with either pole of an electric battery further 

 stimulated my attention. This led to the result that the 

 electric current emanating from the pole of an electric 

 apparatus, while the other is connected with the earth, 

 can be made to traverse a layer of air of ordinary density 

 without producing any light at all ; but when, on the 

 other hand, it encounters a layer of very thin air, the 

 luminous phenomenon will at once appear. These re- 

 searches led me to construct an auroral apparatus for 

 demonstrating this phenomenon as it appears in nature. 

 The knowledge I have gained of the aurora during my 

 continued labours and the above-mentioned observations 

 in particular, made me conclude that an attempt to pro- 

 duce the auroral phenomenon in the very lap of nature 

 by aiding the action of her own forces, ought to give 

 important results and also originate a method for the 

 future study of the phenomenon. In consequence of this 

 idea I made the proposition to the International Polar 

 Congress which I attended at St. Petersburg in 18S1, to 

 erect a station at Sodankyla to follow up my researches, 

 although the results might perhaps be of negative value. 



The Experiments on Oraticnturi. — For the purpose of 

 carrying out the experiment suggested above, the Finnish 

 Polar Expedition was comparatively well equipped, the 

 Physical Faculty of the University at Helsingfors having 

 contributed instruments as well as wire ; but the former 

 had to be slightly altered for the research in view. Cir- 

 cumstances, however, occurred which prevented my work 

 being commenced until the end of November, 1882 ; and 

 after having examined the country and made some pre- 

 liminary experiments from the steeple of Sodankyla 

 Church, situated close to our observatory, I determined 

 to erect the apparatus on the summit of the Oratunturi 

 Mountain, some twelve miles from the observatory. The 

 top was well suited for the purpose, although surrounded 

 by rising copses of wood, which should, according to 

 theory, be rather detrimental to such experiments. The 

 summit was determined by barometrical measurements 

 to be 1070 feet above the town of Sodankyla. To lead 

 the wire from Oratunturi to Sodankyla was, in conse- 

 quence of the great distance, impossible, as there were 

 neither wires nor insulators sufficient, and I had therefore 

 to carry out the experiments as follows : — 



On the summit of Oratunturi (lat. 67° 21', long. 27° 

 I7''3 E. of G.), about 540 metres above sea-level, I laid out 

 the instrument which I have named an "utstrbmnings" 

 apparatus, i.e. a "discharging" apparatus. It consisted 

 of a bare copper wire 2 mm. in diameter, fitted at every 

 half metre with points soldered thereon. The copper 

 wire was laid out in entwined squares, while the points 

 were raised on poles 7.\ metres high, provided with insu- 

 lators, so that each inner coil was about 15 m. from the 

 outer one. The apparatus covered a surface area of 

 900 square metres. From the inner end of this wire 

 an insulated copper wire on poles with telegraph insu- 

 lators led to the foot of the mountain, where a hut of pine 

 branches was erected ; here the wire was connected with 

 a galvanometer, and from this another wire led to a zinc 

 disk in the "earth," i.e. in a flowing spring. The eleva- 

 tion of the apparatus above the zinc disk was 180 metres, 

 and the direction of the conducting wire from the mountain 

 north-west to south-east. 



From the first day the apparatus was finished, viz. 

 December 5, there appeared almost every night a yellow- 

 white luminosity around the summit of the mountain, 

 •while tin such luminosity was seen around any one of the 

 others! The flames were very variable in intensity, and in 

 constant oscillation as those of a liquid fire. Three times 



it was tested, 2^ miles off in south-east, by a Wrede 

 spectroscope (small size with two prisms), and it returned 

 a faintly continuous spectrum from D to F, in which the 

 auroral line X = 5569 with soft variable intensity was 

 observed. The galvanometer steadily gave the de- 

 flexions of a positive current from the " utstrbmnings " 

 apparatus to the earth. The deflexion was so exceed- 

 ingly variable that the needles were in constant motion 

 when the circuit was closed. A Leclanchd's element of 

 ordinary size gave a deflexion which varied according 

 as the positive pole was turned against the mountain or 

 against the earth, but it was always measurable. 



It was, however, impossible to continue the observa- 

 tions for any length of time, as hoar-frost quickly deve- 

 loped on the wires and the poles in large quantities, 

 whereby the insulation became affected. The conducting 

 wire from the mountain was of copper insulated with 

 waxed cotton, and 08 mm. in diameter. This became so 

 covered with hoar-frost that it broke under the weight in 

 spite of the short distance between the supporting poles, 

 viz. 25 metres. It was, therefore, necessary to examine 

 its entire length before the experiments could be com- 

 menced, and as the temperature, as a rule, was under 

 — 30° C, our work was greatly impeded. 



Although the deflexions, by their great variation in the 

 electric current and perhaps, from changes in the electric 

 forces or the imperfect insulation, cannot be of great 

 scientific value, I subjoin them : — 



December 13, 1882. — After several alterations of the 

 galvanometer, and the right sensitiveness having been 

 obtained, the result was, with open circuit : scale- 

 reading 1 of position of equilibrium (a) 36i'5 ; (b) 3627 ; 

 with closed circuit (c) 457 '8 [° r 95 '8 P arts °f the 

 metre = 3°'5, every part of the metre being 2''2], 

 This deflexion constantly varied, and often descended 

 to 30 parts of the metre, to rise suddenly again. The 

 temperature was that day comparatively high, viz. — io° 

 to — 12 C. The air was hazy. 



December 19, 1882. — Plane of equilibrium, 468X); with 

 closed circuit, 471 '6 ; deflexion = 36 parts of the metre. 



A Leclanche"s element placed in the circuit gave — 



1. riane of equilibrium with open circuit 4^9'4 



2. With the carbon pole against the earth 476'4 



3. With the carbon pole against the mountain 4^7'9 



Or the current from the atmosphere = 275 parts of the meter. 

 And the current from the element = 4^25 ,, ,, 



The insulation was here worse than in the previous 

 case, and the temperature very low, viz. - 35 C. 



From December 27 to 29 several experiments were also 

 made, under which the deflexions of the current were 

 greater or less than under the above. 



With regard to the construction of the galvanometer 

 it may be mentioned that the instrument consisted of an 

 ordinary wooden frame constructed for the coiling of the 

 wire, while a pair of astatic needles with a mirror could 

 freely swing within the frame thus : one within and the 

 other above the coil. At first the frame was coiled with 

 copper wire o'5 mm., insulated with indiarubber, but 

 during the experiments this wire was replaced by another 

 0'4 mm., and insulated with silk impregnated with stear- 

 ine. The arrangement of the needles was altered several 

 times, but under the above-recorded experiments they 

 were hung on two fine threads of cocoon silk about 

 20 cm. long and about 1*5 mm. apart. The deflexions 

 were read with a tube and scale from a distance of o'S 

 metre. 



The Experiments at Pietarintunturi. — On this moun- 

 tain (lat. 68° 32'-5, long. 27° 1 7'"3 E. of G.), 950 metres 

 above the sea, a smaller " utstrbmnings " apparatus was 

 erected in two parts, so that the inner one covered about 

 80 and the outer 320 square metres. In other respects it 



1 These figures refer to the reading of a divided metre-scale viewed by 

 reflexion in the mirror of the galvanometer by the ''subjective " method 

 through a reading-telescope.— [Ed.] 



