836 IIUMBOLDT'S COSMOS : 



we should incline to the theory of disruption and unequal 

 projection of the fragments into space, from some force 

 within the body, acting when its primitive consolidation was 

 yet incomplete ; - such force as, on a smaller scale, we must 

 presume to have been concerned in raising the lofty mountains 

 and forming the enormous craters which characterise the sur- 

 face of the moon. Or might we enlarge the hypothesis and 

 suppose it possible that the vast neighbouring mass of Jupiter, 

 already consolidated, aided the action of internal forces in 

 bringing about the anomaly in question? * We pause here 

 however ; not solely from the uncertain ground on which we 

 are treading, but because this instance was brought forwards 

 merely as a possible illustration of the greater changes in 

 bodies far more remote from us. 



Closely related to the new and vanishing stars are those of 

 variable brightness ; of which our author treats at considerable 

 length, and with the advantage of a very valuable communi- 

 cation from Argelander on the stars of periodical variation.f 

 With our limited space we can refer to a few only of the facts 

 which have been accumulated on this curious subject. The 

 phenomenon of variability is more frequent in red stars than 

 in white ones. It exists apparently in stars of every magni- 

 tude. The total number of variable stars, with determinate 

 periods now known, is about 24 ; but many others will 

 doubtless be added by future observation. The periods of 



* The celebrated Kant, in his ' Natur-Geschichte des Himmels,' adopts the 

 idea that the smallness of Mars depends on the prodigious attraction of the mass 

 of Jupiter, acting when the planets were in course of formation. 



f We learn with great satisfaction that Argelander is composing a treatise on 

 this very interesting branch of astronomy. It cannot possibly be in better 

 hands. As an instance of his zeal in research, we may mention that he has 

 succeeded in tabulating above 100 observations on Algol, including a period of 

 fifty-eight years, during which there must have occurred not fewer than 7600 

 periods of variation of this remarkable star ; each marked by equal times of de- 

 creasing and increasing brightness, with a stationary interval of nearly three 

 days between. 



