20 



SCIENCE 



[N. S. Vol. XLII. No. 1070 



ily determined before 1880 at not more than 

 twenty. 



In the 'eighties, progress became more 

 rapid. Gill, the Astronomer Eoyal for the 

 Cape, in conjunction with a young Ameri- 

 can astronomer, Elkin, determined with 

 great accuracy, though with only a small 

 4-inch heliometer, the distance of nine stars 

 of the southern hemisphere. These stars in- 

 cluded a Centauri, and the bright stars 

 Sirius and Canopus. These results were 

 communicated to the Royal Astronomical 

 Society in 1884. The work of GiU and 

 Elkin did not stop there. After some years, 

 a very fine 7-inch heliometer was obtained 

 at the Cape, and with it, between 1888 and 

 1898, the parallaxes of seventeen stars were 

 determined by Gill and his assistants with 

 very great accuracy. The stars observed at 

 the Cape consisted of the brightest stars of 

 the southern hemisphere, and of the stars 

 with the greatest proper-motions. The re- 

 sults were remarkable. The stars with 

 large proper-motions were nearly always 

 comparatively near — say within one million 

 times the sun's distance. On the other 

 hand, some of the very brightest stars, par- 

 ticularly Canopus, the brightest star in the 

 sky after Sirius, were at vastly greater dis- 

 tances. 



Meanwhile Elkin, who had been appointed 

 director of the Yale Observatory in 1884, 

 carried out with a 6-inch heliometer, between 

 the years 1885 and 1892, a determination of 

 the distances of the ten brightest stars of 

 the northern hemisphere. After these were 

 finished the Tale observers, Elkin, Chase 

 and Smith, embarked on the ambitious 

 program of the determination of the dis- 

 tances of 163 stars of the northern hemis- 

 phere which show large proper-motion. 

 They have added forty-one southern stars 

 to these, and thirty-five stars of special in- 

 terest. The results of all these observa- 

 tions were published in 1912. They have 



not, in most cases, the high accuracy of the 

 Cape observations, but, nevertheless, are of 

 great accuracy, and appear to be free from 

 any considerable systematic error. A third 

 important series of observations was made 

 by Peter with a 6-inch heliometer at Leip- 

 zig. These were commenced about 1890, 

 and continued until the death of Professor 

 Peter in 1911. The parallaxes of twenty 

 stars were determined with the same high 

 accuracy as the Cape observations. 



Observations with the heliometer require 

 both skill and industry. To secure the 

 needful accuracy measures must be made 

 in four different positions of the instru- 

 ment, so that possible small systematic 

 errors may be eliminated by reversal. 

 Great care is required in the adjustments 

 of the instrument, particularly in the accu- 

 rate determination of the scale-value at dif- 

 ferent temperatures. The possibility of ob- 

 taining satisfactory results with less labor 

 was considered by Kapteyn, in view of the 

 successful determination of the parallax of 

 Gr. 34 by Auwers. From 1885 to 1887 he 

 made observations with the transit-circle 

 at Leyden of fifteen stars for the purposes 

 of determining parallax. The observation 

 consisted in observing the time when the 

 star the parallax of which was sought and 

 two or three neighboring stars crossed the 

 meridian. Observations are made at the 

 two most favorable epochs — say every night 

 in March, and every night in September — 

 to determine whether the star has changed 

 its position relatively to its neighbors in the 

 interval. The difficulties are twofold. The 

 purely accidental error of observations of 

 transits is considerable as compared with 

 the small quantity which is sought. Be- 

 sides this, the star of which the parallax is 

 required is probably brighter than the com- 

 parison stars, and special precautions are 

 required to guard against personal errors 

 of the observer. 



