Oibers's Essai/ on Comets. 149 



been completely re-calculated, as an appendix to Di. Olbers's 

 Essay, together with a table of the anomaly in a very eccen- 

 tric ellipsis, computed by a formula of Laplace. — Preface of 

 the Editor, Mr. von Zach.] 



■ § 53. 

 Before I conclude this section, I must observe in the last 

 place, that Mr. Schulze has proposed, in the Memoirs of the 

 Academy of Berlin for 1782, a method which somewhat re- 

 sembles mine in the principles on which it depends, and in the 

 general progress of the calculation. But the computation is 

 much more circuitous, and less convenient ; principally because 

 the chord of the earth's orbit is not supposed to be divided in 

 proportion to the times, and because instead of the curtate 

 distance from the earth, the distance of the comet from the 

 sun in the first observation is employed as the principal un- 

 known quantity to be determined. There is also a slight over- 

 sight with respect to the choice of observations. Mr. Sciiulze 

 observes that Lambert has demonstrated that when the in- 

 tervals are nearly equal, the revolving radius in the middle 

 observation divides the chord of the comet's orbit very nearly 

 in proportion to the times : " pounm qu'on emploie des observa- 

 tions assez distantes entrelles." This might have been supposed 

 to be an eri'or of the press ; but in applying his method to the 

 comet of 1779, he actually selects- the remotest observations 

 which he possessed, making an interval of more than 80 days, 

 and very naturally obtains elements of the comet from these, 

 which differ from the truth in an unusual degree. 



Section IV. 



Correction of the Elements of an Orbit ivhich has been 

 determined. 



% 54. 



Some correction of the method explained in the preceding 

 section will always be required : partly on account of its own 

 imperfection, and partly on account of the errors of observa- 

 tion, which must have so much the more effect on the elements 

 at they must necessarily be talK'u at short intervals of time. 



