238 Prof. L. P. Wheeler on the Reflexion of 



By means o£ this equation and equation (11), we can 

 compare the results of experiment with thoory. Thus from 

 the latter we see that if n± is less than n , then A will be 

 positive and consequently by equation (14) n' should be less 

 than n . This, then, is the result we should expect in case 

 there is a gaseous film between the mirror and the liquid ; 

 or, providing that is eliminated, it is the result to be expected 

 in the event that a real transition layer (belonging to the 

 liquid alone, as we have seen), exists. Similarly, if n x is 

 greater than w , we see from equation (11) that A will be 

 negative, and consequently from equation (14) that n' should 

 be greater than n . Such a result would mean a film of 

 greater index than that of the liquid, and hence must be 

 attributed either to films of surface contamination or to a 

 real transition layer, no gaseous film being possible under 

 these conditions. Finally, we see that if A — 0, n' = ?i , and 

 this is the result to be expected if no films of any kind are 

 present. 



Thus if we can make sure that there is no gaseous film 

 nor any film of surface contamination present, a value of n! 

 less than n would indicate that at its surface the liquid has 

 an index of refraction less than in the interior ; while a value 

 of n' greater than n would be evidence that the index of the 

 liquid at the surface is greater than in the interior. 



Experimental Arrangements. 

 In repeating these experiments which have been performed 

 so many times and by such competent observers, it was 

 sought to improve upon their results in two directions : first 

 in securing a mirror which should be more free from surface 

 contamination than those used before, and second in reducing 

 the possibility of a gaseous film between the mirror and the 

 liquid to a minimum. On account of the difficulties in the 

 way of certainly getting rid of films of surface contamination 

 in solid mirrors, it was decided to attempt to attain the first 

 object by using a mercury mirror. The advantages of such 

 a mirror are, first, that it admits of obtaining an extra- 

 ordinarily clean surface much more easily than any other 

 metal ; second, exactly similar surfaces can be easily and 

 quickly reproduced ; and third, the degree of cleanliness 

 attained can be estimated better than with mirrors of solid 

 metal. The method of cleaning the surface by stretching 

 the dirt to the edges (first suggested, in the case of water, 

 by Lord Kayleigh *) was adopted in the form recommended 

 by Rbntgen f. This is described below. 



* Phil. Mag. [5] xxx. 1890, p. 392. 



f Wied. Ann. xlvi. 1892, p. 152. 



