On the time at which Nearchus left the Indus. 229 



concerned myself at all with Columella's risings and settings, 1 '' 

 still, if " ill luck" had not turned the inquiry into that point, 

 his investigations would probably have taken a different di- 

 rection. This discussion so completely supplies all the neces- 

 sary calculations, that we should have had to regret any thing 

 which had curtailed it : but he was a scholar as well as a ma- 

 thematician ; and if, without any previous bias, he had gone 

 to the original text of Strabo, he most probably would have 

 consulted, as he ought, the Greek astronomers for an inter- 

 pretation of it. He would have found that Geminus would 

 have led him to the truth ; and instead of the confusion, which 

 made him think that he had discovered something yet unob- 

 served about the acronychal risings and settings, he would 

 have found that Euctemon's was the only date to which Stra- 

 ta's words could be properly applied. It may be remarked, 

 likewise, that there is an inversion of regular and legitimate 

 method in the course of his reasoning. The coincidence of 

 date which he arrives at must certainly leave a strong impres- 

 sion of the accuracy of the conjecture which leads to it ; but 

 still it cannot be denied that in this we shall rather have esta- 

 blished our premises from our conclusion, than have proved 

 the point in question by a sound course of reasoning. We 

 have arrived at what will afford a very probable solution of 

 our difficulty ; but there is something still wanting to show that 

 the proposed solution is really the true one. This is what 

 now remains to be more particularly considered. 



When any precision is required, the observations of the 

 heavenly bodies on the horizon are the very worst which can 

 be made. The effects of refraction at such very low altitudes 

 are great, and liable to so many variations, that allowances 

 cannot always be made for them with certainty, even in the 

 present state of our knowledge : and when we consider that 

 the appearance and disappearance of stars, as they are affected 

 by the sun's light, will vary (even in the same place) with 

 their different degrees of brilliancy, the state of the atmo- 

 sphere, the powers of perception in the eye of the observer, it is 

 evident that great latitude must be allowed for the times 

 which are fixed for these phenomena. It is only astonishing 

 that these epochs come so near to the truth ; and this can only 



