THE PLANET NEPTUNE. 11 



minute of space, and now the error exceeds two min- 

 utes. In order to exhibit more palpably the nature 

 of these discrepancies, I have represented them upon 

 the jfigure on the next page. If the straight line, 

 A B, be taken to represent the path of Uranus, as 

 computed from the elements of Bouvard, the broken 

 line will represent the observed orbit. The deviation 

 of these two lines indicates the discrepancy between 

 theory and observation for the dates at the top of the 

 page, the amount of the discrepancy being given on the 

 left margin. If it was doubtful whether this difference 

 in the case of the ancient observations was not due to 

 the carelessness of the observers, no such supposition 

 is admissible in the case of the observations since 

 1821. The discrepancy between Bouvard's orbit and 

 the observations is now enormous, and is increasing 

 with alarming rapidity. 



What can be the cause of these discrepancies? Do 

 they indicate errors in the computations of M. Bouvard ? 

 In order to decide this question, the illustrious Bessel, 

 about the year 1840, subjected all the observations to 

 a new calculation ; and although he detected one or 

 two errors of Bouvard, they did not materially influ- 

 ence the results. He satisfied himself that the ancient 

 and modern observations could not be reconciled by 

 any modification of the elements, and that the differ- 

 ences could not be attributed to inaccuracy of instru- 

 ments, or to methods of observation. Are these anom- 

 alies due to the attraction of some unknown disturb- 



