254 HISTORY OF PHYSICAL ASTRONOMY. 



dinary command of analysis,) and then by com- 

 paring these, in supposed critical cases, with the 

 Brest observations. This method has confirmed the 

 theory as far as it could do so ; but such a process 

 cannot supersede the necessity of applying the 

 proper criterion of truth in such cases, the con- 

 struction and verification of Tabes. Bernoulli's 

 theory, on the other hand, has been used for the 

 construction of Tide-tables; but these have not 

 been properly compared with experiment; and 

 when the comparison has been made, having been 

 executed for purposes of gain rather than of sci- 

 ence, it has not been published, and cannot be 

 quoted as a verification of the theory. 



Thus we have, as yet, no sufficient comparison 

 of fact with theory, for Laplace's is far from a com- 

 plete comparison. In this, as in other parts of 

 physical astronomy, our theory ought not only to 

 agree with observations selected and grouped in a 

 particular manner, but with the whole course of 

 observation, and with every part of the phenomena. 

 In this, as in other cases, the true theory should be 

 verified by its giving us the best Tables ; but Tide- 

 tables were never, I believe, calculated upon La- 

 place's theory, and thus it was never fairly brought 

 to the test. 



It is, perhaps, remarkable, considering all the 

 experience which astronomy had furnished, that 

 men should have expected to reach the completion 

 of this branch of science by improving the mathe- 



