1885.] on the Solar Corona. 211 



coronal light itself by onr air, but tbe amount of this scattered light 

 over the corona must be less than what is seen over the dark moon. 



That the sun is surrounded by a true gaseous atmosphere of rela- 

 tively limited extent there can be little doubt, but many considera- 

 tions forbid us to think of an atmosphere which rises to a height 

 which can afford any exiDlanation of the corona, which streams several 

 hundred thousand miles above the photosphere. For example, a gas 

 at that height, if hundreds or even thousands of times lighter than 

 hydrogen, w^ould have more tlian metallic density near the sun's 

 surface, a state of things wliich spectroscopic and other observations 

 show is not the case. The corona does not exhibit the rapid con- 

 densation towards the sun's limb which such an atmosphere would 

 present, especially when we take into account the effect of perspective 

 in increasing the apparent brightness of the lower regions of the 

 corona. There is, too, the circumstance that comets have passed 

 through the upper part of the corona without being burnt up or 

 even sensibly losing velocity. 



There can scarcely be doubt that matter is present about the sun 

 wherever the corona extends, and further that this matter is in the 

 form of a fog. But there are fogs and fogs. The air we breathe, 

 when apparently pure, stands revealed as a dense swarming of millions 

 of motes if a sunbeam passes through it. Even such a fog is out of 

 the question. If we conceive of a fog so attenuated that there is only 

 one minute liquid or solid particle in every cubic mile, we should still 

 have matter enough, in all probability, to form a corona. That the 

 coronal matter is of the nature of a fog is shown by the three kinds of 

 light which the corona sends to us. Eeflected solar light scattered 

 by particles of matter solid or liquid, and secondly light giving a 

 continuous spectrum, which tells us that these solid or liquid particles 

 are incandescent, while the third form of spectrum of bright lines, 

 fainter and varying greatly at different parts of the corona and at 

 different eclipses, shows the presence also of light-emitting gas. This 

 gas existing between the particles need not necessarily form a true 

 solar atmosj)here, which the considerations already mentioned make 

 an almost impossible supposition, for we may w^ell regard this thin 

 gas as carried up with the particles, or even to some extent to be 

 furnished by them under the sun's heat. 



It will be better to consider first the probable origin of this 

 coronal matter, and by wdiat means it can find itself at such enormous 

 heights above the sun. 



There is another celestial phenomenon, very unlike the corona 

 at first sight, which may furnish us possibly with some clue to its 

 true nature. The head of a large comet presents us with luminous 

 streamers and rifts and curved rays, which are not so very unlike, 

 on a small scale, some of the appearances which are peculiarly 

 characteristic of the corona.* We do not know for certain the con- 



* See " Comets," Eoyal Institutiou Procfcdingr--, vol. x. p. 1 . 



p 2 



