288 ON THE CHANGE IN THE ADOPTED UNIT OF TIME. [37 



Mr Stone's error appears to arise from his equating two things which 

 are really different, and which are inconsistent with each other, viz. Bessel's 

 and Le Verrier's determinations of the Sun's mean motion in longitude in 

 the same interval of time. 



Major-General Tennant is wrong in supposing that solar observations 

 are no longer employed in Observatories for the determination of mean solar 

 time. If this were the case, it would only shew that the Observatories 

 had taken a very retrograde step, since the final test whether the mean 

 solar times have been correctly found can only be supplied by solar obser- 

 vations. Whenever the mean solar times are deduced from the observed 

 sidereal times, it is tacitly assumed that the tabular mean longitudes of 

 the Sun which have been employed are correct ; and if this is not the 

 case, the mean solar times deduced will require a corresponding correction, 

 which can only be found by solar observations. 



Thus mean solar time may be determined with reference to a natural 

 phenomenon, viz. the transit of the true Sun over the meridian of a given 

 place ; and the mean solar day is the average of all the apparent solar 

 days defined as the intervals between two successive transits, and therefore 

 has nothing arbitrary about it. To speak of Besselian mean time and Le 

 Verrian mean time, or of the Besselian mean solar day and the Le Verrian 

 mean solar day, can produce nothing but confusion in our ideas of the 

 measure of time. 



