CHAPTER III. 



DISCUSSION OF THE OBSERVATIONS OF NEPTUNE. 



§ 21. During the four years following the discovery of Neptune, observations 

 of this planet, both meridian and extra-meridian, were very numerous. If the 

 results of all these observations were free from constant errors, and, therefore, 

 strictly comparable both with themselves and with subsequent observations, their 

 combination would give very accurate positions of the planet. Unfortunately, 

 however, we cannot assume that observations of different kinds, made at different 

 observatories, are strictly comparable, nor have we, in many cases, the data for 

 reducing them to a common standard. 



Let us consider, for instance, the meridian observations. Under the title of 

 " Meridian Observations of Neptune," we find in astronomical periodicals series of 

 observed Right Ascensions and Declinations. But right ascensions and declinations 

 can never be really observed with any instrument. Only times of transit, and the 

 readings of micrometers and other instruments, are really observed. The right 

 ascensions and declinations of the planet are concluded from the observations, by 

 the aid of a great number of subsidiary data, some relating to the stars, others 

 to the instrument. Respecting these data we have, in most cases, absolutely no 

 information whatever. But a knowledge of some of them, at least, is indispen- 

 sable. Even if we grant that the instrumental errors are in all cases perfectly 

 known for every observation, we still do not know either the names or the assumed 

 right ascensions of the stars used in determining clock errors. Hence we cannot 

 use the results, because the right ascensions given in standard catalogues not 

 unfrequently differ by a second of space. 



The declinations of the planet are sometimes determined by comparison with 

 standard stars, sometimes by measures of nadir distance, combined with the lati- 

 tude of the observatory. The Paris observations are reduced by the former 

 method ; those of most otlier observatories, by the latter. Using the latter method, 

 it would naturally be supposed that the declinations from the observations of all 

 observatories of which the latitudes are well determined ought to agree. But 

 such is far from being the case. Compare, for instance, the declinations of funda- 

 mental stars concluded from observations with the great transit circle at Green- 

 wich with those in the Tabula3 Reductionum of Wolfers, and we shall find that 

 for stars more than 45° from the pole, the Greenwich positions are systematically 

 nearly a second south of Wolfers', an amount greater than the probable error of 

 a single isolated observation. We cannot impeach either authority. Wolfers' 

 positions depend on such authorities as Pond, Struve, Argelander, Henderson, 

 Airy, and Bessel. The conscientious care bestowed on the reduction of the 

 Greenwich observations would seem to render their results unimpeachable. 

 Besides, from a comparison of Winnecke's observations of his "Mars Stars" in 



