110 



III. On the Longitude of the Armagh Observatory, given by fifteen Chronome- 

 ters of Arnold and Dent, Sfc. By the Rev. Dr. Robinson, M.R.I. A., &c. 



Read 10th December, 1838. 



IHE determination of this important element is at least as difficult as essential ; 

 and whatever be the care of the astronomer it often happens that after years of 

 observations have elapsed, the result still remains in some degree uncertain. The 

 various methods of determining arcs of longitude have each their peculiar causes 

 of error. When the methods of signals can be employed with only one inter- 

 mediate station, it is decidedly the best; but obviously the measurement of large 

 distances is in most cases impracticable, and when many stations intervene the 

 accumulated errors may attain a serious magnitude. The expense of this pro- 

 cess, and the number of assistants required, are also frequently very serious 

 objections. 



The longitudes assigned by geodetic operations depend on an assumed figure 

 of the earth, whose constants are not well known, whose very existence is proble- 

 matical; and even if correct, it will differ from the Astronomical longitude 

 whenever local attractions deflect the direction of gravity to the east or west 

 of the theoretic vertical. 



The mere observation of an occultation is the most satisfactory that can be 

 imagined in common cases ; but there is uncertainty enough in deducing from it 

 a longitude, caused by the doubtful nature of some elements that enter the calcula- 

 tion. It is affected by errors in the tabular place of the moon, which are not totally 

 corrected when the declination has been actually observed, as only one limb can 

 be taken, and that is affected by irradiation. It is influenced by the error of the 

 tabular semldlameter, and still more of the horizontal parallax, which is to a cer- 

 tain extent hypothetical, whether given by theory or deduced from observation. 

 And lastly, it depends on the assumed distance of the spectator from the earth's 

 centre, a quantity computed on the hypotheses of its spheroidal figure and given 



