38 CELESTIAL MECHANICS: LEVSCHNER 



(94) AURORA 



Discovered by Watson 1 , 1867, September 6, at Ann Arbor. 



Preliminary elements were computed by Tietjen 23 , the second set 

 given below as Elements A. 



With Elements A, H. Leppig 4 formed eight normal places over 

 162 days, and determined Elements B. An accurate ephemeris includ- 

 ing special perturbations by Encke's method was computed for 1868, 

 December 15 to 1869, January 31. Observations by Vogel, January 

 15, 17, 18, 1869, gave residuals, Aa 21 S .6, AS +53".5. 



Thereafter, the B. J. gives elements by Leppig from 1872 to 1915, 

 (but with a misprint of 4 in o> from 1887 to 1897 5 ), Bauschinger 

 gives Leppig's Elements C in the "Tabellen" 6 for the epoch 1883, July 

 12.0. The various elements in the B. J. to 1915 are probably brought 

 up from Leppig's Elements B, with special perturbations. The char- 

 acter of the perturbations is not given. Nor is any reference made 

 to arbitrary corrections. 



In Kleine Planeten, 1916, the elements are changed by estimating 

 the perturbations 1883-1910 and roughly determining, M and /*,, Ele- 

 ments D. 



In Kleine Planeten, 1921, the last elements are again corrected by 

 the computation of Jupiter's perturbations and a representation of 

 observed positions, 1884-1918, within l m .5. Osculating elements, 

 1921, April 24, not available. 



From 1884 to 1899 the planet was practically lost, mainly on ac- 

 count of the misprint in to. Coddington 7 computed a place with the 

 elements of B. J. 1901, and found the planet Aa +5 m .2, AS 20', 

 1899. For the computation of the perturbations and tables of the 

 Watson asteroids, Leuschner made a collection of Leppig's elements in 

 the B. J. including Elements C and derived average Elements E from 

 those in B. J. 1871, 1873, 1874, 1875, 1877, 1878, 1881, 1898. From 

 these, approximate mean elements were derived. The perturbations 

 were developed and a preliminary correction of the mean motion 

 and mean anomaly from observations in 1867 and 1899 was at- 

 tempted. A mistake in one of the main terms of the perturbations 

 was discovered and corrected. Nevertheless, large discrepancies be- 

 tween observation and computation remained ( 5 in a, 1867). These 

 differences were used for a preliminary correction of the mean motion 

 and of the mean anomaly. With the new value of the mean motion 

 the perturbations were corrected, the residuals re-determined, and 

 further corrections made to the mean motion and to the mean 

 anomaly. A least square solution of 12 places from 1867 to 1899 

 was then made including the corrected perturbations. The residuals 



