252 CARNEGIE INSTITUTION OF WASHINGTON. 



Interferometer Observations of Capella. 



Part of the time from December to April was devoted by Mr. 

 Anderson to an application of Michelson's interferometer method to 

 the measurement of the spectroscopic binary Capella. The method, 

 first desciibed by Michelson in 1890-91 and employed by him in a 

 determination of the diameters of Jupiter's satellites, has not since 

 been applied to astronomical problems. Preliminary trials by Pro- 

 fessor Michelson having shown that atmospheric disturbances probably 

 would not offer serious interference, it was decided to test the method 

 with the 100-inch telescope, and the problem chosen was the measure- 

 ment of close double stars or spectroscopic binaries. 



In the ordinary use of a telescope, light from a celestial body is 

 allowed to fall on the entire objective; when a telescope is used as an 

 interferometer the light traverses only two relatively small areas of the 

 objective, which are usually rectangular in shape and symmetrically 

 situated with respect to the center. If the object observed is a single 

 star of exceedingly small angular diameter, the observer sees in the 

 focal plane a series of perfect interference fringes. If, on the other 

 hand, the object be a uniformly luminous disk of angular diameter a, 

 the distinctness of visibility of the fringes will be poor if the distance 

 between the two areas of the objective is in the neighborhood of 

 X/a, and indeed will be zero if this condition is exactly fulfilled. If the 

 object be a double star whose components are separated by an angular 

 distance a, the fringes will be invisible when the distance between the 

 areas is X/2a and the line joining the areas is parallel to the line joining 

 the two stars, provided the two component stars are equally bright. 

 If the components are of unequal brightness the visibility will be a 

 minimum under the same conditions. 



The interferometer, consisting of a tube about 27 inches long, having 

 two rectangular apertures of variable separation at one end and an eye- 

 piece at the other end, was mounted at the Cassegrain focus. It could 

 be rotated about its own axis for determination of position angle. It 

 was at once found that Capella showed the disappearance of the fringes 

 to be expected in the case of a double star of nearly equal components, 

 and a series of four complete and two incomplete observations was 

 seciu-ed between December 30, 1919, and April 23, 1920. These obser- 

 vations show that both angular separation and position angle can be 

 measured to an accuracy of about 1 per cent. It was also found that 

 seeing as poor as 1 on a scale of 10 is not very harmful, though with 

 better seeing the observations are easier to make. Since the spectro- 

 scopic elements of the system of Capella are already well known, the 

 present observations give us a good knowledge of the true orbit, the 

 distance, and the mass of Capella. The semi-major axis of the true 

 orbit is 0^05249 and the parallax 0!f0600. The masses of the compo- 



