2l8 



NA TURE 



[Al'RIL 



21, I9IO 



■see similar formations of light to those which occur during 

 my experiments with kathode rays around a magnetic 

 terrella. 



In my work, " The Norwegian Aurora Polaris Expedi- 

 tion, 1902-1903," descriptions will be found in several 

 places of these phenopiena, but to elucidate the subject 

 here I append a few new illustrations, which show very 

 plainly the shape of these formations of light. 



Figs. I and 2 show how the rays are drawn in in belts 

 around the magnetic poles of the terrella. corresponding 



February, p. 57), and it is not impossible that indications 

 of an alteration in those parts of the comet's tail nearest 

 the planet may be noticeable. 



We may then possibly expect to find traces of the rays 

 being drawn in towards the polar regions of Venus in a 

 manner similar to that demonstrated by the experiment 

 shown in Fig. 4, or a more or less distinct bending of 

 the comet's tail, assuming Venus to be magnetic. 



The probability of such being visible must, however, be 

 admitted to be small, as the central line or the tail, if it 





with the polar-light zones on the earth. They are taken 

 looking along and perpendicular to the magnetic axis. 

 Fig. I shows the spiral rings of light around a magnetic 

 north pole, corresponding to the south pole of earth 

 magnetism. \\'e find these belts of light sometimes, as 

 here, with a tolerably even strength of light like a con- 

 tinuous band, and at other times we find the rays con- 

 centrated in three limited streaks, with well defined 

 positions around the magnetic poles of the terrella. 



Fig. 3 also shows an equatorial ring. This phenomenon 



of light is magnificent, 

 but unstable ; it is difficult 

 to produce ; it may 

 suddenly appear and sud- 

 denly vanish, as the rays 

 which run round the 

 terrella at the equator are 

 difficult to obtain suffici- 

 ently concentrated for the 

 rarefied gas to illuminate 

 them. At the lower part 

 of Fig. 3 and on Fig. 4 

 a characteristic pointed 

 tongue of light will be 

 seen, which is drawn in, 

 and shows the manner in 

 which the rays here come 

 into the terrella. The 

 magnetic equator is in- 

 dicated on the terrella by 

 a dark line. 

 It may now be imagined that analogous formations of 

 light may be observable around the earth of the rays 

 from the comet's tail on May 18-19. ^^e downward rays 

 in the -Polar regions will, it is true, be difficult to observe 

 in northern parts owing to the northern declination of the 

 sun, but in Antarctic regions there may be more hope 

 of doing so, and the phenomenon would then probably 

 appear somewhat similar to the aurora australis. At night, 

 in low latitudes, one may conceive the possibility of a 

 ring like the equatorial ring being observable. 



About May 2 the comet will be in the vicinity of Venus 

 (see Bulletin de la Sociitd astronomique de France, 



NO. 21 12, VOL. 83] 



Fig. 4. 



Fi ;. 3. 



is directly away from the sun, will be at a considerable 

 height above the planet ; but I desire, nevertheless, to direct 

 the attention of astronomers to these conditions, as Venus, 

 if as strongly magnetised as our earth, must be expected 

 to e.xercise a noticeable influence on the tail of the comet 

 at a distance of several million kilometres, especially if 

 the rays in the tail are easily deviated by magnetic force. 



This phenomenon may, in case it is present, be 

 determined by astronomical observations of the comet's 

 tail and Venus in the period May 1-3, and I beg, there- 

 fore, to ask astronomers, in the interests of science, to 

 make arrangements for the necessary observations, if 

 possible, and to favour me with a short account of the 

 results. Kr. Birkeland. 



Universitetets Fysiske Institut, Christiania, 

 March. 



Neutral Doublets at Atmospheiic Pressure. 



In his papers on magneto-kathode rays, Prof. Righi 

 assumes the presence of neutral doublets, formed of a 

 positive and negative ion in more or less stable combina- 

 tion. Sir J. J. Thomson has independently put these in 

 evidence very clearly in his experiments on positive rays. 

 Working independently, we have made some observations 

 which point to the existence of such doublets at atmo- - 

 spheric pressure. Ionised gas is drawn through two 

 insulated metal tubes ; along the axis of each a thiok 

 insulated wire electrode is fixed. These wires can be con- 

 nected in turn to a Dolezalek electrometer, and the current 

 between tube and electrode measured. The ionised gas Js 

 produced by splashing mercury, or by heating lime or 

 aluminium phosphate on a strip of platinum foil. With 

 a certain blast, in one case, the current reached its satura- 

 tion value on the first electrode with a voltage of 320, 

 being then 130 in arbitrary units. Raising the voltage to 

 656 did not increase this by one division ; the extreme read- 

 ings at the intermediate voltages were 128 and 130, so 

 that the observations were quite regular. Nevertheless, 

 with 656 volts on the first electrode a current can be 

 detected at the second, this in some cases amounting to 

 as much as 10 per cent, of the original. 



Lime gives a large excess of negative, mercury, and 

 aluminium phosphate of positive, ions, but in each case 



