PART II. POLAR MAGNETIC PHENOMENA AND TERRELLA EXPERIMENTS. CHAP. V. 655 



Dr. BARNARD, after his examination, in 1907, of the illumination of the dark side of Saturn's rings, 

 ,uggests the explanation(') that the rings are auto-luminous; but he rejects the idea by conjecturing that 

 mch a hypothesis would not be compatible with the presumed physical composition of the rings. 



I think it will be quite possible to satisfy all the results of the observations hitherto made of these 

 ings by a hypothesis entirely different from the above-mentioned meteoric theory. 



On p. 613 of the present volume, I have described some experiments that have served as a starting- 

 joint for an explanation of the zodiacal light. 



Round a highly magnetic globe, 8 cm. in diameter, in a vacuum-tube with a capacity of 70 litres, 

 have produced a ring with a diameter of up to 34 cm., and long luminous rays in the polar regions 

 if the globe, the whole bearing a considerable resemblance to pictures of the sun during an eclipse. 



Now if the discharge-current, which in the above experiment was from 10 to 30 milliamperes, be 

 educed to i milliampere or less, the polar phenomena cease, and the ring becomes exceedingly thin 

 nd sometimes assumes an appearance almost exactly like that of Saturn's rings. 



Round the magnetic equator of the globe, and touching it, a luminous zone appears, then a dark 

 pace, which, farther from the globe, is gradually formed into a flat, dimly-luminous ring resembling the 

 rape ring of Saturn. This dimly-luminous ring farther away increases in strength and a light-ring 

 ippears. 



Fig. 245. 



Fig. 245, i shows the rings from the side, and fig. 245, 2 a little from above, thus making the 

 ark space between the globe and the ring distinctly visible. Fig. 245, 3 shows, in addition to a brightly 

 iminous ring, a fainter ring outside the former, and separated from it by a dark division that might 

 nswer to CASSINI'S division in Saturn's ring. Fig. 235 shows that by a special arrangement it has 

 een possible to get as many as 5 rings, one outside another, round the globe. In this case, however, 

 ic rings are not flat, as the outer ones are in the form of cylinders, which increase in height with 

 icir distance from the globe. When the magnetisable globe is not magnetic, but is still a cathode, it is 

 ften seen surrounded by several luminous spherical envelopes. It is perhaps these that, when the globe 

 ; magnetised, become changed in shape and flattened. 



How are the phenomena of Saturn's rings to be explained, supposing the rings to be due to 

 imilar electric radiation from the planet, the latter being considered to be magnetic? 



With regard to physical investigations of the power of an electrically luminous gas, and of radiant 

 latter, to absorb and diffuse solar light, we have mentioned some few known facts on page 623, to 

 /hich the reader is again referred. 



I think there are also here good reasons for admitting that in the radiant matter which we suppose 

 > have been radiated by Saturn, there is a comparatively very great number of electrons of dispersion, 



(') Astrophys. Journ., vol. XVII, 1908, p. 39. 



