170 



Mr Glazebrooh, On the effect of moisture in [Oct. 27, 



latter have shewn that with most substances between certain 

 values of the angle of incidence the reflected beam is elliptically 

 polarised. 



In all such experiments one great difficulty meets us. The 

 effects we are looking for do not seem exactly the same from day 

 to day. Modifications of some nature appear to go on at the 

 surface and the means from results of observations made on 

 different occasions differ by quantities which are large compared 

 with the differences between anyone of the means and the in- 

 dividual observations of the set from which it is deduced. 



Thus the following readings represent observations on the 

 position of the plane of polarisation after plane polarised light 

 polarised at an angle of 45° to the plane of incidence has been 

 allowed to pass through a wedge of glass of small angle. These 

 observations were made on six occasions during August last 

 year. 



Each is the mean of five or six observations no two of which 

 probably differed by more than 3'. 



The object of the experiments described in the present paper 

 was to determine some of the conditions which produce this 

 variation. 



Numerous observations seemed to indicate some connexion 

 between the effect and temperature. Thus readings of the 

 analyser, a Laurent's plate and a Nicol's prism, were found 

 generally to be higher on warm days than on cold. 



The following set of observations shew the change on July 19 

 as the temperature of a thermometer placed almost in contact 

 with the glass rose from 20° to 22°. 



Now it is improbable that the change is due to inequalities 

 of temperature in the glass which is thin and carefully shaded 



