264 E. S. Larsen — Relation between the Refractive Index 



ponent present. "We can apply this formula to the elements 

 making up a compound as well as to a simple mixture or solu- 

 tion. An identical relation holds for the formula of Lorentz 

 and Lorenz : 



Since the appearance of the papers of Gladstone and Dale, 

 much worl has been clone to test their formula and several 

 other formulae have been proposed, but only those of Glad- 

 stone and Dale and of Lorentz and Lorenz have theoretical 

 significance or general application. Both formulae, together 

 with the corresponding formulae for .dispersion and for the 

 additive relation for solutions and the elements of a compound, 

 have been carefully tested on many organic liquids and salt 

 solutions. In general, the two formulae hold almost equally 

 well for varying temperatures and concentrations, the first 

 holding better in one case and the second in another. Lorenz,* 

 Bruhle,f and others have shown that when applied to a sub- 

 stance in different states, — to a gas and a liquid, for example — 

 the simpler formula gives errors as great as thirty per cent, 

 while the n 2 formula gives lower though still considerable 

 errors. Landolt,J Conrady,§ and other investigators showed 

 that in organic liquids the refractive constant depends upon the 

 atomic constitution as well as upon the chemical composition, 

 and that some elements, such as oxygen f must be given differ- 

 ent values for k, depending upon the structure of the molecule. 



An excellent monograph, including a review of former work 

 on this subject and a complete bibliography, has recently been 

 published by Cheneveau.|| 



At the present time, but few data are available for the 

 application of these formulae to silicate glasses and to minerals 

 of simple, known composition. The present investigation was 

 undertaken to secure such data and to test the application of 

 the formulae to several different silicate glasses and corre- 

 sponding minerals. The two series selected for study were the 

 soda-lime feldspar series and the three component system 

 CaO-MgO-Si0 2 , with especial emphasis on the metasilicates. 



The glasses were prepared by mixing the pure components 

 and melting, then grinding and remelting several times to 

 secure homogeneity. The probable error in composition 

 should not exceed 0"3 per cent.*[ Glasses rich in CaO and MgO 

 could not be prepared on account of their strong tendency to 

 crystallize. The crystals were prepared in the course of the 



* Wied. Ann., xi, 70, 1880. f Zeitschr. f. phys. Chem., vii, 1, 1891. 



JPogg. Ann., cxvii, 353, 1862. § Zeitschr. f. phys. Chem., iii, 210, 1889. 

 || Ann. d. Chem. et Phys., xii, 145, 289, 1907. 

 T Allen, E. T., and W. P. White, this Journal (4), xxvii, 2, 1909. 



