g2 Double Stars and tht 



pared with that which really exists, proves tliat there are not 

 in the heavens, two stars of the first magnitude, sufficiently 

 near each other to render it probable that their distance 

 ought to be considered as fortuitous. In return the magni- 

 tude which we find in the first, presents us with examples of 

 remarkable proximity. Who is not acquainted with the three 

 brilliant stars of the second magnitude, in the belt of Orion, 

 the two exterior of which are distant from the middle one, 

 the one, only twenty-six, and the other eighteen minutes ? 

 Calculation proves that there are one thousand four hundred 

 to one against the probability that this nearness is accidental. 

 The constellation of the southern cross is still more remark- 

 able. We there find, within the space of fifteen degrees 

 square, (which does not comprehend the two thousand sev- 

 en hundredth part of the celestial vault,) one star of the 

 first magnitude, two of the second, one of the third, and 

 one of the fourth ; and the doctrine of probabilities shows 

 twenty thousand to one against an accidental combination 

 of this nature. We have good reason therefore to presume 

 that these stars are dependent one upon another. 



These conjectures are confirmed when we examine the 

 stars to the sixth and seventh magnitudes, relative to their 

 distribution in the celestial vault. According to a calcula- 

 tion of probabilities, founded on the number of these stars 

 which are found in the catalogue of Harding, the case in 

 whicli two among them should be distant from thirty-two 

 seconds to a minute, ought not to exist more than once and 

 a half, while we are in fact acquainted with fifteen exam- 

 ples. There ought not to be more than six or seven couples 

 from the first to the seventh magnitude, in which the two 

 stars forming the couple, should be distant from one to two 

 minutes, and there are fifteen already known. If we wish 

 to calculate on greater distances for stars of the sixth mag- 

 nitude, we shall find that there ought not to be more than 

 seven or eight couples distant from two to five minutes, while 

 in fact there are eighteen. From five to ten minutes, prob- 

 abilities give us twenty-seven or twenty-eight couple, and 

 we are acquainted with thirty-six. We find more than cal- 

 culation would grant, even at ten or fifteen minutes, namely 

 twenty-five instead of twenty-two. We may then regard with 

 ereat probability a considerable number of pairs of stars from 

 the first to the sixth magnitude, in which the two stars are 

 distant from each other from one to fifteen minutes, as be- 



