82 



SCIENCE. 



[N. S. Vol. XVI. No. 394. 



velocity in the medium within the observ- 

 ing telescope. This is the only method 

 thus far used which gives the absolute 

 velocity of light. 



The uncertainty in the constant of aber- 

 ration and the errors in the observations of 

 Jupiter's satellites render these methods 

 unsuited for the comparison of the tvFO 

 v'eloeities. We owe perhaps to Arago more 

 than any other person a solution of this 

 problem of the velocity of light. In 1838 

 he submitted to the French Academy the 

 plan of an experiment for solving directly 

 and definitely the question— which was be- 

 ing much debated— whether light was a 

 corpuscular emission from radiant bodies 

 or simply the result of the vibrations of a 

 very rare medium. The method was sim- 

 ply an adaptation of the rotating mirror in- 

 geniously devised by Wheatstone for deter- 

 mining the velocity of propagation of elec- 

 tricity in a wire. It was not till 1850, 

 hoAvever, that this method, in the hands of 

 Foucavilt, was actually put to the test of 

 determining this constant. In the preced- 

 ing year, however, Fizeau had published 

 the results of his experiments by means of 

 a toothed wheel. These were the first ob- 

 servations made on the velocity of light at 

 the earth's surface. The first idea of this 

 method seems to have originated with the 

 Abbe Laborde, who communicated it to 

 Arago some years before. Upon these 

 two principles is based our entire knowl- 

 edge by terrestrial methods of this constant. 

 On the one side we have the refinement and 

 modifications of the toothed wheel as used 

 by Cornu in France and Young and Forbes 

 in England and on the other the very ac- 

 curate results of Michelson in his classic 

 experiments and those of Newcomb with 

 his highly refined phototachometer in this 

 country. It may well be questioned 

 w'hether much gTeater certainty in data is 

 attainable than those which the late la- 



mented French savant and illustrious 

 physicist Cornu has left as one of his most 

 brilliant legacies to science. Nor can we 

 hope for any considerable refinement in 

 the determinations by the other method as 

 used in this country, and which have al- 

 ready given the most surprising agreement 

 and warrants us in taking the value ob- 

 tained as the most accurate of all. Not- 

 withstanding the elaborate execution of 

 these experiments serious discrepancies 

 exist between the results of the two 

 methods. The latest discussion by Cornu 

 of his own results gives as the most prob- 

 able value of this constant 300,130 meters, 

 while the values of Michelson and of New- 

 comb are 299,853 and 299,810, respectively. 

 This makes the eri'or of the results by the 

 two methods ten times the error between 

 those of the same method. This difference 

 has given rise to the impression that in the 

 one or the other or in both methods a 

 systematic error exists and the discussions 

 and corrections made by different critics 

 have left the problem in a somewhat uncer- 

 tain state. The relation of these determin- 

 ations to the absolute velocity in air at the 

 earth, and to the absolute terrestrial velocity 

 in vacuo and the possible difference from 

 the velocity in space, renders the problem 

 of great interest and importance in this 

 aspect alone. The famous theory of Weber 

 of moving charges to explain the action of 

 electric currents, while incompatible with 

 the principle of the conservation of en- 

 ergy, has done much to enlarge our views 

 of the origin of electrodynamic phenomena 

 and to establish a comprehensive theory of 

 present phenomena. The brilliant experi- 

 ment of Rowland as a natural sequence of 

 Weber's theory demonstrated the electro- 

 magnetic reaction of a moving charge and 

 showed further that if the velocity were 

 that of light the mechanical reaction would 

 be approximately that calculated from 



