THE MEASURE OF STELLAR SPACES 335 



hands each minute of time ; a thousandth of a " second " is less 

 than a twenty-millionth part. 



One has perhaps a little better idea of the accuracy attained 

 when he thinks of looking through a slit at the far end of a tube, 

 the slit the breadth of a thin knife-blade say, the hundredth 

 of an inch the tube more than half a mile long. The slit would 

 represent a tenth of a second of arc, of which one degree would 

 cover thirty feet. These are the units, the subdivisions of the 

 gauge, with which Bessel and his successors worked. 



Observations taken from August 1837 to October 1838 led 

 him to a parallax for 61 Cygni of fW of a second. This meant 

 nearly six hundred thousand mean distances of earth to sun. 

 It meant that light, travelling six hundred and sixty million 

 miles per hour, would require nine and a half years to cover the 

 intervening space. The distance has since been slightly re- 

 duced. But nearly seventy years of minute observation, by 

 many different astronomers in different parts of the earth, 

 have changed Bessel's result but little. He had at least reached 

 a figure which could stand. The long puzzle was at an end. 

 Here at last was a star whose distance was known. 



Curiously enough, as so often happens, the solution would 

 have come from other hands within not more than a year. Com- 

 bining the observations of Henderson in 1832 with others made 

 in 1839, it was found that there was a star much nearer than this 

 No. 61 of the Swan. This was the now familiar alpha Centauri, 

 the third brightest star in the heavens, but unseen by our 

 northern eyes. Its parallax was originally calculated at o'.gi, 

 or nearly a second. Subsequent reckonings have reduced this 

 somewhat ; it is now set down at 0^.75. This indicates a dis- 

 tance of 4.4 light-years. Seventy years of observation, again, 

 have failed to disclose any more proximate sun. 



About at the same time, Struve at Dorpat announced a 

 parallax of a quarter of a second for Vega. Somewhat earlier 

 he had estimated that of the alpha star in the Lyre at between 

 o".i and o".2 of a second. After two years of observation, 

 Peters, at Pulkowa, gave o".i as the parallax of Polaris. This 

 figure is slightly too great ; it is probably not more than o".o6, 

 answering to a distance of forty-four light-years. This is still 

 about at the limit of certain parallactic measure. It is pain- 

 ful work ; less than a hundred are yet known with sufficient 

 precision to permit of any confidence in the results. This is 



