in the Velocity of Propagation of Light in Water. 271 



ment which a ray of light experiences in passing- through a known 

 thickness of a horizontal liquid layer. This experiment is per- 

 formed by the observation of a mark at the bottom of a vessel 

 before and after it is filled with a liquid. 



Although it has been repeatedly shown that this method of 

 determination is defective, and that a small error in the observa- 

 tion of the displacement gives rise to great differences, because 

 great magnitudes are determined by small ones, yet of late re- 

 course has been frequently had to various modifications of this 

 plan*. Its value was stated to consist especially in the fact 

 that the luminous ray can at once enter the exposed surface of 

 the liquid from the air; while in investigating by means of 

 prisms, alterations of the density may occur just at the important 

 terminal surfaces. According to Poisson's view of capillarity, 

 such a condensation must certainly take place at the surface of 

 contact between the liquid and the glass ; but such condensation 

 would undoubtedly take place in layers parallel to the confining 

 glass surface. It would vanish at a short distance from the 

 glass, and the liquid would soon assume its normal density, so 

 that the angle of refraction could not be affected by it. If, 

 however, such an action of condensation were to take place 

 within the range of observation, the image refracted (through a 

 narrow hollow prism) of a long homogeneous straight line of 

 light would necessarily appear convex ; also a condensation of 

 the liquid would take place towards the walls of the tubes. But 

 I have never yet been able to detect anything of the kind, even 

 on the passage of light through narrow metal tubes. 



The method of refraction which has been chiefly used in the 

 actual determinations of the index of refraction, and by means 

 of which the excellent measurements of Fraunhoferf and Baden 

 Powell J were made, as well as those of Becquerel and Cahours§, 

 Deville||, DelnVj|, Dutirou**, Beer and Kremerstt* Hoekjj, 

 Landolt§§, Handel, Ad. Weiss and E. Weiss ||||, Gladstone and 

 Dalelffl, Forthomme***, and finally also my own, includes a 



* " Memoire sur la determination des indices de refraction," Comptes 

 Rendus, vol. xxxix. pp. 27-29; and " Neue Methode den Brechungsindex 

 von Fliissigkeiten zu messen, von Ch. Montigny/' Pogg. Ann. vol. cxxiii. 



f Denkschriften der Munchener Akademie, vol. v. pp. 1812-1815. 



% Ann. de Chim. et de Phys. S. 3. vol. v. p. 129. Pogg. Ann. vol. lxix. 



§ Comptes Rendus, vol. vi. p. 867. Also Pogg. Ann. vol. li. p. 267. 



|| Ann. de Chim. et de Phys. S. 3. vol. v. p. 129. 



*f[ Pogg. Ann. vol. lxxxi. p. 470. 



** Ann. de Chim. et de Phys. S. 3. vol. xxviii. p. 176. 



tt Pogg. Ann, vol. ci. %% Ibid. vol. cxii. p. 347. 



§§ Ibid. vol. cxxiii. 



|| || Wiener Afcademieberichte, vols. xxv. xxxiii. 



WI Phil. Trans. 1858, pp. 887 &c Also Pogg. Ann. vol. cviii.p. 632 &c. ; 

 and Phil. Trans. 1863, p. 317. 



*** Ann. de Chim. et de Phys. vol. lx. p, 307- 



