February 21, 1902.] 



SCIENCE. 



285 



ation without introducing an assumption 

 with respect to one or the other. It has ac- 

 cordingly been assumed that the latitude 

 variation can be represented by two peri- 

 odic terms of 14 and 12 months, respec- 

 tively. Each observed latitude, therefore, 

 gives an equation of this form 



X sin N -}- y cos iV + s sin o -f- m cos -|- Ef-\- T/J. 



+ A^ + ^„ = <I, 



where A' is the 14-month term, Ep the cor- 

 rection to the aberration and T^ secular 

 change in the latitude. • 



The evening and morning observations 

 furnish 1,744 and 1,052 equations of 

 this form, respectively. These were solved 

 retaining all terms. Then another solu- 

 tion was made excluding the annual 

 term, as it is obvious that this can not be 

 separated from the 14-month term in a 

 series embracing a period of less than 16 

 months. The resulting value of the con- 

 stant of aberration from both the prelimi- 

 nary and revised solution is as follows : 



Preliminary. Revised Solution. 



1st solution, 20'''.552 ± .0095 20". 551 ± .0092 

 2d " 20 .555 ±.0093 20 .552 ± .0090 



It thus appears that the suspected error 

 in the preliminary reduction was a vanish- 

 ing quantity. 



Two other series of observations made 

 at the Sayre Observatory have been em- 

 ployed for a similar investigation. The 

 first, from December 1, 1889, to Decem- 

 ber 13, 1890, embracing 1,344 latitude de- 

 terminations, was treated in a manner sim- 

 ilar to that above described except that only 

 one solution was made involving periodic 

 term of 14 months. The resulting aber- 

 ration constant is 20".448 ± .014. It is 

 doubtful whether this result is entitled to 

 much confidence, the most serious objection 

 being that the series does not cover a full 

 term of 14 months, which is assumed for 

 the period of latitude variation. The other 

 series referred to extends from January 19, 



1894, to August 19, 1895, the arrange- 

 ment being that sometimes called the 

 polygon method by which the aberration is 

 obtained independently of the latitude 

 variation. The result is 20". 537. Other 

 recent determinations are given for com- 

 parison : 



FloiDcr Observatory. 



1896-98 20*.580 



1898-99 20 .542 



1900-01 20 .560 



J. W. J. A. Stein, Leiden — Zenith Telescope. 

 1899-1900 20*541 ± .016 



Alhrechi — International Latitude. 

 20*515 



Chandler frorm Pond' s Observations. 

 1825-36 20*512 ±.019 



Tlie Period of Delta Equulei: W. J. 



IIUSSEY. 



Delta Equulei enjoys the distinction of 

 having a period shorter by nearly half than 

 that of any other visual double star and 

 in being at the same time a spectroscopic 

 binary. The latter characteristic is due 

 to the visible components. It therefore 

 forms a connection between the visual and 

 spectroscopic double stars. This pair was 

 discovered in 1852 by Otto Strvive and ob- 

 served by him occasionally for thirty years. 

 Burnham's observations from 1880 to 1883, 

 inclusive, appeared to indicate a period of 

 10.8 years; the elements subsequently de- 

 rived by Wroublewsky and by See gave 

 nearly 11.5 years. Between 1899.85 and 

 1900.65 the components rapidly approached 

 each other and the apparent distance was 

 extremely small at the latter date. This 

 rapid change in distance is not explainable 

 on the hypothesis that the period is nearly 

 11.5 years, but is entirely in accord with 

 one of about half this length. The ele- 

 ments derived by the writer a year ago 

 have a period of 5.7 years. From these 

 elements he predicted that the components 

 should separate to a measurable distance 



