PART II. POLAR MAGNKT1C PHENOMENA AND TERRELI.A EXPERIMENTS. CHAP. IV. 569 



(see p. 552); and we assumed, after discussing our experiments, that the rays again and again looped 

 back on themselves and described curves that more or less resembled epicycloids. Rays such as these 

 would be able to pass at the equator and nearest to the terrella with a velocity-component from east to 

 west, i. e. a direction the reverse of that of the primary rays. 



It is interesting here to call to mind (cf. pp. 82 & 83, Section I) that the negative equatorial storms 

 were explained by the bending round of rays in the vicinity of the magnetic equator, so that they encircled 

 the earth from west to east, while the positive equatorial storms were explained by rays with a compo- 

 nent motion from east to west, nearest the earth at the equator. 



We are not able to see, in the photographs as reproduced here, any distinct signs of shadows 

 cast by the two parallel wires upon the terrella, although, during the experiments, such shadows were 

 easily discernible, though always faint. 



In both precipitations, A and B, upon the horizontal screen, there occurred in certain positions of 

 the terrella, curved but parallel shadows of these wires. These shadows have been especially useful in 

 investigations for the purpose of coming to an understanding regarding the rays of group B, as we shall 

 presently see. 



111. Equatorial Rings of Light. In connection with the ray-phenomena just described, belonging 

 to group A, we will discuss a phenomenon which has already been mentioned several times, and called 

 equatorial rings. The phenomenon is described in "Expedition Norvegienne 1899 1900", p. 41 ; but un- 

 fortunately on that occasion the luminous rings were not photographed. There is, however, a photograph 

 of one in Section I of the present work, p. 80, fig. 37. 



The equatorial ring is formed of rays that curve round the terrella from west to east. Under spe- 

 cially favourable experimental conditions, the concentration of rays near the plane of the equator is so 

 great that the rarefied gas is rendered luminous. It is not only rays that move exactly in the plane of 

 the equator that form the ring, but more especially rays that move alternately above and below the plane 

 of the equator in its immediate vicinity. We will here point to photographs 5 and 6, fig. 203, where 

 ust such rays as these are made distinct by their precipitation upon the terrella about the equator. Even 

 the rays that come nearest to the terrella in the polar regions, and which thus, in their passage through 

 the plane of the equator, intersect it at large angles, will perhaps serve to produce the luminous ring, as 

 they bring about a powerful concentration of rays just at the magnetic equator. We have seen indeed 

 that the rays from one primary pencil, have numerous lines of intersection in the equator. When 

 such rays, by a suitable proportion between the magnetism and the stiffness of the rays, are free to move 

 a great many times round the terrella near the equator, the gas there becomes luminous, and we may 

 have the equatorial ring. As may be expected from what has been stated, the appearance of the ring is 

 almost a chance phenomenon; it is unstable, and many fruitless attempts may be made to induce it to 

 show itself. 



The three photographs forming fig. 207 were taken several years ago, and the experiments on 

 that occasion were made with a powerful influence-machine. The strength of the current with these 

 machines, however, is so small that the phenomena are not bright. The rarefied gas itself, moreover, 

 plays a very important part if the phenomenon is to be successful. It seems as if impurities were an 

 assistance. The experiments were made with a tension of about 6000 volts, and with the employment 

 of about 10 amperes upon the terrella No. 2, with a diameter ofiocm. The magnetic moment with this 

 current-strength was about 50,000 C. G. S. The ring is distinctly seen to be rather thin and broad, its 

 outer margin often extending far beyond the terrella. The inner margin of the ring often comes right 

 up to the terrella; but I have several times observed the ring standing unattached in the gas, with a 

 dark interval between it and the terrella. 



