44 



observation nor theory having yet detected in it the least varia- 

 tion." By '* uniform" is meant that the successive transits of tlie 

 same star over the same meridian of the earth's surface happen 

 after always precisely equal intervals of time ; but it must be ob- 

 vious from the slightest consideration, that if the solar days are 

 not uniform, neither can the sidereal days be so; for the meridians 

 over which the sun's and the star's transits are made are the 

 same, or at least corresponding meridians of the same surface of 

 the same revolving body, and consequently liable to the same 

 variations and disturbances in all respects. If, therefore, the 

 *' uniformity" of the sidereal day be an observed and positively 

 ascertained fact, some other cause of the variations of the solar 

 day than those assigned by astronomers must be sought for ; 

 and that cause is most readily to be found in the elliptlcity of the 

 earth's orbit, and the sun's eccentric place within it. The direct 

 result of these circumstances is that the earth's radius-vector al- 

 most constantly subtends a continually varying angle to the orbit, 

 sometimes approaching it on one side and sometimes receding. 

 As the earth revolves from her aphelion point, the radius-vector 

 diverges from the perpendicular, and gradually approaches the 

 west side of the orbit, forming an acute angle with its vertex to 

 the north, and consequently meeting the earth's meridians pro- 

 portionally sooner every day. Soon, however, it begins to re- 

 cede, and open out from the orbit, and consequently to make the 

 earth's meridians take a proportionally longer time to reach it in 

 their rotation. The same process is repeated four times, in con- 

 verse order however, during the whole orbital revolution, and 

 sufficiently well accounts for the difference in respect of '' uni- 

 formity" between the sidereal and the solar days ; for the stars 

 outside, in respect of relative position, are quite unaffected by 

 the variations of the radius-vector. Having no means of making 

 accurate observations for myself, I can only reason on the data 

 supplied by others. Taking, therefore, the common tables for 

 the equation of time, I find them to agree remarkably well with 

 the cause which I have now supposed. From this it will follow 

 that mean and apparent solar time will coincide or equally show 

 true time, only at those points where the radius-vector is perpen- 

 dicular to the orbit. In a perfect orbit or circle, the radius- 

 vector would be invariably perpendicular to the orbit, and con- 

 sequently would always be the true index and measure of the 

 planet's rotations, while the star outside would fall short of that 

 true measure by more than three minutes every day. 



In a popular treatise on science I find it stated that " any 

 meridian on the earth will revolve from a fixed star to that star 

 again in exactly the same time as if the earth had only a diurnal 

 motion, and remained always in the same part of its orbit." 



