MOUNT WILSON SOLAR OBSERVATORY. 



215 



JUPITER'S NINTH SATELLITE. 



Using the 1915 observations, an improved determination of the 

 orbit of Jupiter's ninth satelUte was made by Mr. Nicholson, and an 

 ephemeris was computed for the 1916 opposition. The satelKte was 

 found near the computed place on four negatives taken by Mr. Shapley 

 with the 60-inch reflector. Its position was measured on these plates 

 and the orbit corrected by the aid of the observations at three opposi- 

 tions, 1914, 1915, and 1916. The positions of three new asteroids 

 were also obtained from the plates. 



MEASUREMENTS OF PARALLAXES AND PROPER MOTIONS. 



At the 80-foot focus of the 60-inch reflector 402 plates with 535 

 exposures were taken by Mr. van Maanen, as follows: 



The measures and reductions for 20 parallax fields have been com- 

 pleted, thus giving a total of 54 finished fields. Most of the stars 

 measured are of the later spectral types, of magnitudes 5 to 7, and of 

 proper motions less than 0''5 annually. The results for the first 50 

 fields have been discussed for systematic errors in several ways: 



1. The parallaxes of the 396 comparison stars, arranged according 

 to brightness, seem to indicate a systematic error of +0''001 in the 

 parallaxes of the principal stars. 



2. The neglect of the quadratic terms of the coordinates of the 

 comparison stars has caused no systematic error. 



3. A comparison of five fields observed in common with other obser- 

 vers shows no appreciable error. 



4. A comparison of the proper motions in right ascension derived 

 here with those given by Boss indicates a systematic error of — O'^OOl. 



5. A comparison of the Mount Wilson parallaxes with those derived 

 from van Rhijn's tables gives a difference of +0''007. 



6. A comparison of 47 of the parallaxes with those determined by 

 Mr. Adams by his spectroscopic method indicates an error of +0''003. 



7. A comparison with the mean of the parallaxes determined by 

 other observers for 68 stars of the same spectral type, magnitude, and 

 proper motion points to a systematic error of — 0''001. 



The only large difference is revealed by the comparison with van 

 Rhijn's tables; there are several reasons, however, for believing that 

 this can not be wholly due to the Mount Wilson parallaxes. From all 

 material available it is probable that the latter have no systematic 

 error exceeding 0''003. 



