21 



stant, the interval of successive eclipses would not be altered. But in the 

 course of six months the distance of the earth from Jupiter increases by the 

 diameter of the earth's orbit, and in the next six months' changes back again. 

 When the earth is nearest to Jupiter, the eclipse reaches us in about 32 min,, 

 but when the earth is at tlie greatest distance, it takes 50 minutes to reach us. 



It is clear from this that the intervals between successive eclipses are variable, 

 being sometimes longer and sometimes shorter than the real intervals. Delambre 

 discussed 1000 of these eclipses, observed between the years 1662 and 1802, 

 from which he calculated the velocity of light to I'equire 493. 2 seconds to pass 

 over the mean distance of the Sun. If this time, then, divides 95,360,000 

 statute miles, which is the Sun's distance, deduced from the transits of Venus 

 in 1761 and 1769, the quotient 193,350 statute miles is the velocity of light in 

 one second. 



The second process which astronomy has supplied for obtaining the velocity 

 of light, requires not space but a velocity which is commensurable with the 

 velocity of light. In nature the velocity of the earth is compounded in this 

 way with the velocity of hght, and imparts to light an apparent path, differing 

 by a small angle from the true path. The angular displacement thus caused 

 between the apparent and the real places of a star is called aberration, from 

 which Bradlev explained anomalies in observation which had been until that 

 time considered accidental. The displacement of a star works contrary ways 

 at opposite seasons of the year. Half the difference between the extreme 

 places is the distance from the apparent to the true place, or the constant uf 

 aberration. This, when known as an observed fact, establishes the ratio 

 between the velocity of light and the velocity of the earth, and enables the 

 astronomer to assign the value of the one with all the accuracy which pertains 

 to his knowledge of the other. 



The result of aberration obtained by Struve is 20" "35, from which the velocity 

 of light is calculated to be 10,088 times as great as the velocity of the earth. 

 The mean velocity of the earth is known from the magnitude of its orbit, that 

 is, of the sun's distance. 



Assuming the distance derived from Enck's parallax to be the most correct, 

 the velocity of the earth in one second of solar time is 18,977 miles. This 

 multiplied by the above ratio gives 191,513 mUes for the velocity of light by 

 Bradley's principle. It appears, therefore, that the velocities by these two 

 astronomical methods differ by 1837 miles, a small quantity comparatively, 

 being only onex>e.r cent, of the whole velocity. 



The experiment on the velocity of electricity by Professor Wheatstone, 

 published in 1834, suggested the possibility of measuring in a similar way the 

 velocity of light, and to this pui-pose it was afterwards made apphcable by 

 Fizeau, from the results of which the French Academy referred the subject to 

 a scientific commission. F. Arago next made expeiiments on rapid rotation, 

 and, being aided by the refined sldll of Brig^aet, he realised velocities in the 

 mirror of 1000 turns in a second of time. These experiments have been of 

 late much improved by M. Foucault, in causing a pencil of solar light reflected 

 into a horizontal direction by a heliostat to fall upon a micrometric mark, 

 which is made the real standard of measure. The rays which traverse this 

 initial surface fall next upon a series of rotating mirrors, to which a constant 

 velocity is imparted with air supplied by a high pressure belows. 



M. Foucault's experimental results by means of this ne w apparatus, which 

 he says has been purged of uncertainty, gives the velocity of light in space as 

 298,000 kilometres in a second of mean time. This value reduced to statute 

 miles gives the velocity of light as 185,177 miles in a second — which is less by 

 6336 mUes than the velocity admitted by science as computed from aberration. 

 This difference between the residt of experiment and those of astronomical 

 observation, which come nearest to it, is three times greater than the variation 

 between the velocity deduced from aberration and that derived from eclipses. 



M. Foucault states that the mean resvdts by his experiment can be trusted 

 to the fraction of 1-500. Now the aberration of 29" '45, adopted by astronomers, 

 cannot be at fault more than one 1,800th of the whole. How, then, is the 

 velocity of light to be reconciled with the value of aberration. Is it possible 

 there can be an uncertainty of three per cent, in the velocity of the earth ? If 

 there is an error in the velocity of the earth it is an error in space, and to 

 diminish the earth's velocity sufficiently by a change of time would require an 

 increase of nearly eleven days in the length of the year. The only other way 

 of reaching the velocity of the earth is by diminishing the earth's orbit, and 

 that would change proportionately the mean radius of the orbit — that is, the 

 sun's mean distance. Can the sun's distance from the earth, then, be considered 

 uncertain to the extent of three per cent, of the whole ? 



The limits of accuracy by which the sun's distance from the earth has been 



