360 THE STELLAR UNIVERSE. 



effect. A periodic star could be caused only by considerable and permanent 

 spots. 



2. Newton conjectured that the variation of brightness might be produced 

 by comets falling into distant suns and causing temporary conflagrations. 

 Waiving any other objection to this conjecture, it is put aside by its insuf- 

 ficiency to explain the periodicity of the phenomena. 



3. Maupertius has suggested that some stars may have the form of thin flat 

 disks ; acquired either by extremely rapid rotation on an axis, or other physical 

 cause. The ring of Saturn affords an example of this within the limits of 

 our own system, and the modern discoveries in nebular astronomy offer other 

 examples of a like form. The axis of rotation of such a body might be subject 

 to periodical change like the nutation of the earth's axis, so that the flat side 

 of the luminous disk might be present more or less toward the earth at differ- 

 ent times, and when the edge is so presented it might be too thin to be visible. 

 Such a succession of phenomena are actually exhibited in the case of the rings 

 of Saturn, though proceeding from different causes. 



4. Mr. Dunn* has conjectured that a dense atmosphere surrounding the stars, 

 in different parts more or less pervious to light, may explain the phenomena. 

 This conjecture, otherwise vague, indefinite, and improbable, totally fails to ex- 

 plain the periodicity of the phenomena. 



5. It has been suggested that the periodical obscuration or total disappear- 

 ance of the star, may arise from transits of the star by its attendant planets. 

 The transits of Venus and Mercury are the basis of this conjecture. 



The transits of none of the planets of the solar system, seen from the stars, 

 could render the sun a periodic star. The magnitudes even of the largest of 

 them, are altogether insufficient for such an effect. To this objection it has 

 been answered that planets of vastly greater comparative magnitude may re- 

 volve round other suns. But if the magnitude of a planet were sufficient to 

 produce by its transit these considerable obscurations, it must be very little in- 

 ferior to the magnitude of the sun itself, or, at all events, it must bear a very 

 considerable proportion to the magnitude of the sun ; in which case it may be 

 objected that the predominance of attraction necessary to maintain the sun in 

 the centre of its system could not be secured. To this objection it is an- 

 swered, that although the planet may have a great comparative magnitude, it 

 may have a very small comparative density, and the gravitating attraction de- 

 pending on the actual mass of matter, the predominance of the solar mass may 

 be rendered consistent with the great relative magnitude of the planet by sup- 

 posing the density of the one vastly greater than that of the other. The den- 

 sity of the sun is much greater than the density of Saturn. 



6. It has been suggested that there may be systems in which the central 

 body is a planet attended by a lesser sun revolving round it as the moon re- 

 volves round the earth, and in that case the periodical obscuration of the sun 

 may be produced by its passage once in each revolution behind the central 

 planet. . 



Such are the various conjectures which have been proposed to explain the 

 periodic stars and as they are merely conjectures, scarcely deserving the name of 

 hypotheses or theories, we shall leave them to be taken for what they are worth. 



TEMPORARY STARS. 



Phenomena in most respects similar to those just described, but exhibiting 

 no recurrence, repetition, or periodicity, have been observed in many stars. 

 Thus, stars have from time to time appeared in various parts of the firmament ; | 



* See PhiL Trans., vol. 52. 



