162 THE ROYAL SOCIETY OF CANADA 



the experience gained in the previous determination, soon obtained 

 by the aid of Dr. King's graphical method. 1 

 The values adopted were: 



Period 140 -70 days 



0-7 

 55° 

 K .... 26-25 km 

 T . . ..J. D. 2,418,054-80 

 ..+42-21 km 



From these preliminary elements a least squares solution using 

 Schlesinger's convenient method 2 was carried through, resulting in the 

 following values of the elements: 



Period P = 140-70 days. 



Eccentricity e = 0-717± -022. 



Longitude of apse co = 54° • 16± 4° -35. 



Semi-amplitude K = 27-12 ± 1 -44 km. 



Periastron passage T = J.D. 2,418,054 -723± 0-520 



Velocity of system 7 = +42-59 km. 



Maximum velocity =81-10 km. 



Minimum velocity =26-85 km. 



Projected length semi-axis major a sin. i = 37,471,000 km. 

 These elements may be considered final for 2 f>w is only re- 

 duced from 75-04 to 72-87 and the differences between the values 

 obtained by substituting in the observation equations and those 

 obtained from the ephemeris from these elements are very small. 



The comparatively high values of the probable errors of the 

 elements are due principally to the abnormal deviations, between phase 

 70 and 134, from the velocity curve drawn from these elements and 

 shown in the full line in the accompanying figure. These deviations, 

 which will be more fully discussed later, make the probable error of a 

 normal place of unit weight and consequently the probable errors of 

 the elements nearly double what they would otherwise be. 



From a carefully drawn curve on a large scale, the residuals from 

 the observations were obtained and are given in the last column but 

 one of the table of observations. From these residuals the probable 

 error of a single Ottawa plate comes out as ±3-6 km per second. 



It will be of interest to compare the probable errors for the 

 different dispersions used and these are given herewith. 



'Astrophysical Journal XXVII p. 125. 

 2 Pub. Allegheny Observatory I p. 33. 



