PROPER MOTION OF THE STARS. 253 



time, therefore, before the beginning of his last and painful 

 illness, led Bessel, the greatest astronomer of our time, to the 

 conviction " that stars whose variable motion becomes appa- 

 rent by means of the most perfect instruments, are parts of 

 systems confined to very limited spaces in proportion to their 

 great distances from one another." This belief in the exis- 

 tence of double stars, one of which is devoid of light, was so 

 firmly fixed in Bessel's mind, as my long correspondence with 

 him testifies, that it excited the most universal attention, 

 partly on his account, and partly from the great interest 

 which independently attaches itself to every enlargement of 

 our knowledge of the physical constitution of the sidereal 

 heavens. " The attracting body," this celebrated observer 

 remarked, " must be very near either to the fixed star which 

 reveals the observed change of position, or to the sun. As, 

 however, the presence of no attracting body of considerable 

 mass at a very small distance from the sun, has yet been 

 perceived in the motions of our own planetary system, we are 

 brought back to the supposition of its very small distance from 

 a star, as the only tenable explanation of that change in the 

 proper motion which, in the course of a century, becomes 

 appreciable." 10 In a letter (dated July, 1844) in answer to 

 one in which I had jocularly expressed my anxiety regard- 

 ing the spectral world of dark stars, he writes: "At all 

 events, I continue in the belief that Procyon and Sirius are 

 true double stars, consisting of a visible and an invisible star. 

 No reason exists for considering luminosity an essential pro- 

 perty of these bodies. The fact that numberless stars are 

 visible, is evidently no proof against the existence of an 

 equally incalculable number of invisible ones. The physical 

 difficulty of a change in the proper motion, is satisfactorily 

 set aside by the hypothesis of dark stars. No blame attaches 



10 Schum. Astr. Nachr., nos. 514-516. 



