102 



Bear. The southern hemisphere presented fewer cases, 

 In the full catalogue of double spheres, including those 

 observed by American astronomers, the number reaches 

 about 11,000. 



The best time for observation of double stars is consi- 

 dered to be from midnight to early dawn. 



One of the most remarkable features of the double 

 star is its colour. In the case of single stars we know 

 only three colours white, yellow and red, with varying 

 shades, while amongst double stars we meet with every 

 possible variety pure white, yellow, orange red, pale red, 

 bright red, whitish blue, greenish blue, sky blue, pale green, 

 full green, grey and purple. The most interesting com- 

 binations are presented by double orbs of different colour. 

 In 1892 an analysis of these colours was made by the 

 German astronomer Neffkong, when the following com- 

 binations were revealed: 10 pair white with yellow; 

 7 pair white with red; 7 pair white with ash-gray; 

 3 pair white with green; 8 pair white with blue; 2) 13 

 pair yellow with red; 13 yellow with ash-gray; 3 - 

 yellow with green, 62 yellow with blue; 3) 2 red with 

 green, 1 red with blue. 4) 5 green with ash-gray; 14 

 green with blue; 5) 2 yellow with yellow of different 

 shade; 6) 1 blue with blue of different shade. 



Amongst these stars those of a blue or green colour 

 are the most interesting, since these colours are never 

 met with in single stars. Many astronomers even regard 

 these cases as mere optical illusions, but the result of 

 scrupulous study in all the chief observatories of the old 

 and new world compels us to believe that blue stars 

 actually exist. We have indeed examples of a white star 

 with its blue satellite, just as we have of a yellow star 

 with its red satellite. There is one easy method of pro- 

 ving that the colour of double stars is no optical illusion, 

 proceeding from a contrasted effect; it is only necessary 

 with the optical screen to shut out one of the pair from 

 view, when, if the other retains its colour, the phenome- 

 non is evidently not illusory. But the decisive proof in 



