986 
EEY. T. E. EOBINSON ON SPECTEA OE ELECTEIC LIGHT. 
their own different from that of their elements. In even a simple mixture like an-, I 
have shown that its spectrum, with a given metal, cannot certainly be deduced from 
those of its parts, and it is probable that this rule may be widely extended. 
The bearing of this on electro-spectral analysis is obvious ; for if the presence of one 
substance can be shown in any instance to disguise or transform the spectrum of another, 
or if the state of density, solution, alloyage, &c. have influence, it becomes necessary to 
eliminate such eficcts before we yield implicit confldence to this powerful guide. This 
implies a wide range of experiment and of cautious study ; in fact, a complete system of 
spectral research through the whole range of our chemical elements and their compoimds, 
conducted with strict inductive logic, and with the highest appliances of chemistiy and 
optics. Whatever shall be so effected will be “ an everlasting gift ” to science, because, 
taking nothing for granted, it will be a real fact. 
