2 8o SCIENCE PROGRESS 



indices of the elements, which may tell us how they differ from 

 each other ; the comparison of the indices of compounds with 

 that of their constituents, which may throw light on the nature 

 of chemical combination ; and the dispersion of light by elements 

 or compounds, which is confidently looked to for a clue to the 

 actual configuration of the atom, and in particular to the part the 

 electron plays in its composition. 



Of these the first is the most fundamental, and it was therefore 

 natural that the early workers should begin by examining those 

 elements which lay nearest their hands. Biot 1 and Arago 

 measured the refractivities of hydrogen, oxygen, and nitrogen. 

 Dulong repeated their work and added chlorine. But here 

 all further progress was checked by want of material. Bromine 

 had only just been discovered, and all the other elements 

 required so large a rise of temperature in order to vaporise 

 them that they were outside the sphere of practical work at the 

 time. Nor did the results obtained in the case of these elements 

 seem to foreshadow the existence of any simple relation between 

 the indices of the elements, such as might have encouraged the 

 pioneers to further exertions. The refractivity 2 of hydrogen is 

 139, of oxygen 270, and of nitrogen 297. At that time Newton's 

 emission theory of light held the field, and it was sought to 

 connect the acceleration of light as it was then thought to be 

 with the density of the medium. But neither when these figures 

 were regarded simply, in which case they represented the relative 

 retardations of equal numbers of atoms, nor when they were 

 divided by the atomic weights, when they were proportional to 

 the relative retardations caused by equal weights of different 

 elements, was any clear rule visible. By the first calculation 

 hydrogen retards light less than the others ; but for unit weight 

 its refractive power is the greatest known. Nitrogen, with a 

 lower atomic weight than oxygen, retards light slightly more, 

 and chlorine, with higher atomic weight, had a refractivity of 

 768, very much above the others. 



1 Biot calculated his results for air to ten places of decimals, though his 

 experimental results only agree to six, and when this was pointed out he replied 

 sarcastically that if the first figures were wrong, perhaps the last would be right. 



2 The term " refractivity " is used to mean (ji— 1) io 6 and is proportional to the 

 absolute amount of retardation of light passing through a gas of standard density. 

 Thus for air at o° and 760 mm., /x = 1 "000292 and the refractivity is 292. In other 

 words, light is retarded nearly mfo-o of its velocity in vacuo while passing through 

 air under these conditions. 



