March 17, 1305.] 



SCIENCE. 



•419 



the measures published by Plummer and 

 Poor were not made on the original nega- 

 tives, and can not, therefore, be properly 

 called measures. The only real measures 

 so far published, to my knowledge, have 

 been those of the Eros plates taken with 

 the Crossley reflector, and measured at Co- 

 lumbia and Lick. A few of these measures 

 discussed by Hinks showed distortions 

 giving anomalous results near the edge of 

 the plate, but these anomalies are matched 

 on the plates taken at Algiers with the 

 standard photographic refractor, and no- 

 ticed on the following page of Hinks's 

 paper. It should also be stated that the 

 aperture ratio of the Crossley was large, 

 about 1 to 6. 



For this work a diaphragm twelve inches 

 in diameter has been used on the 24-inch 

 reflector, and as the focal length is 93 

 inches, the ratio is a little greater than 1 to 

 8. Allowing for the area cut out by the 

 flat, the clear aperture of the mirror is 

 equivalent to lOJ inches. The exposures 

 have been timed to give good measurable 

 images of all the stars on Hagen's charts 

 which extend to twelfth or thirteenth 

 magnitude ; in good seeing this requires 

 ten minutes with ordinary plates and fif- 

 teen minutes on the isochromatic plates. 

 The magnitudes have been deduced by 

 measurements of disk diameters, the in- 

 crease per magnitude being nearly uniform 

 and amounting to about 0.025 mm. As the 

 diameters are measurable with a probable 

 error of 0.001 mm., corresponding to 0.04 

 of a magnitude, the results are comparable 

 with the best visual measures. 



This Avork has some similarity to the 

 spectral photometry of the Draper cata- 

 logue, each taking account of the intensity 

 curve of the spectrum; but differs from it 

 in two respects : It is not confined to the 

 bright stars, but can reach to the faintest 

 visible ; also, taking account of the entire 



spectrum, its results will harmonize with 

 visual magnitudes. 



The Quadruple System of Alpha Gemi- 



norum : Heber D. Curtis. 



The well-known binary star a Geminorum 

 was pronounced by Sir John Herschel to 

 be the largest and finest of the double stars 

 in the northern portion of the sky. Meas- 

 ures, of a very rude character, were made 

 of this pair as early as 1718 by Bradley 

 and Pond, so that this system has been un- 

 der observation for nearly two hundred 

 years. In spite of this fact some of the 

 elements of the orbit are still quite uncer- 

 tain, particularly the eccentricity and the 

 period. Vah;es of the eccentricity have 

 been derived, ranging from 0.32 to 0.80, 

 with corresponding periods of 1,001 to 232 

 years. In recent years the distance be- 

 tween the two components has commenced 

 to decrease, with the result that the ele- 

 ments have become rather more deter- 

 minate, and Doberck (A. N., 3970) has re- 

 cently expressed the hope that through this 

 decrease in the distance it will be possible 

 to fix the orbit with considerable accuracy 

 within the next ten or twenty years. 



Doberck has derived the following sets 

 of elements, of which he regards the sec- 

 ond as the most probable and most in agree- 

 ment with recent measures. 





Elements 



of Castor. 





o 



29° 29' 



33° 56' 



42° 34' 



X 



84 44 



82 26 



118 11 



?' 



73 3 



63 37 



61 56 



e 



0.7513 



0.4409 



0.2321 



Period 



268 years 



347 years 



502 years 



T 



1,936.65 



1,969.82 



1,963.30 



a 



7".326 



5".756 



6".467 



Retrograde. 



In January, 1896, Belopolsky at Pulkova 

 discovered that the fainter of the two stars 

 forming this system is itself a rapid 

 spectroscopic binary.* The period of this 



* Bull. Acad. 8t. Petersburg, December, 1896. 

 Astrophysical Journal, January, 1897. Mem. 

 Acad. St. Petersburg, XI., 4, January. 1900. 



