M. Arago on the Physical Constitution of the Sun. 207 



shall examine the stars which nearly touch each other ; and 

 this great astronomer will prove to us, that these stars which 

 are in some way coupled together, do not appear near each 

 other solely by the effect of perspective, but that they are in 

 a state of mutual dependence, and circulate round their com- 

 mon centre of gravity in a short time, which, in certain cases, 

 has been already determined. 



In observing that these double stars are of very dissimilar 

 colours, our thoughts naturally turned to the inhabitants of 

 the obscure and revolving planetary bodies which apparently 

 circulate round these suns ; and we would remark, not with- 

 out real anxiety for the works, the paintings, of the artists 

 of these distant worlds, that to a day lightened by a red light, 

 succeeds not a night but a day, equally brilliant, but illumi- 

 nated only by a green light. 



The comparison of the positions of the stars determined at 

 different epochs, proves that they have very erroneously been 

 csdledjixed ; that they move in space in different ways, so that 

 in the long-run the present form of the constellations will be 

 completely changed ; that the absolute speed of these stars 

 is unequal, but that it has been ascertained, with regard to 

 one of them, with complete certainty, that it moves 20 leagues 

 per second ; that the sun, in this respect resembling all the 

 other stars, is not immoveable, and draws after it the train 

 of planets with which it is surrounded. 



We must be struck with the inequality of the distribution 

 of the stars in the celestial sphere. There, we see more than 

 20,000 of them in a superficial space equal to the tenth part 

 of the apparent surface of the moon ; here, in a space of the 

 same extent, not a single luminous speck will be perceptible, 

 even with the best telescopes. 



After having carefully glanced over the luminous matter 

 scattered over such immense spaces, and which by its agglo- 

 meration, continued during centuries, seems to have produced 

 new stars, we would discuss the vast ideas of Wright, Kant, 

 Lambert, and William Herschel, upon the constitution and 

 dimensions of the milky way. In short, some additional steps 

 in astronomical conjecture, that is to say, in that branch of 

 science founded only on imposing probabilities and natural 



