308 Mr Glazebrook, On a common defect of lenses. [Jan. 29, 



when the two sources are viewed directly, 9 t its position when 

 the cell filled with water is in the way, and 2 its position when the 

 cell filled with the liquid to be examined has been substituted for 

 the water cell. The following table gives a series of observations 

 taken in this manner by Mr Sanderson. Two strengths of solution 

 were used, the values of 2 are indicated by 2 and 6' 2 . 



Substituting in the formula above we find for the ratio of the 

 strengths of the two solutions determined from the observations in 

 the blue, green, and red respectively the values 1*46, 1*43 and 140 

 respectively. The values of 0, &c. in the above table for the green 

 and red were the mean of two series which agree fairly closely, 

 more uncertainty attaches to the values in the blue. The mean 

 of the errors in the value of 6, &c. as determined from 5 or G 

 observations is about 45'. These observations would seem to shew 

 that the instrument is capable of comparing the strengths of two 

 solutions to within 4 or 5 per cent, without difficulty. 



A somewhat similar piece of apparatus has been described by 

 Hufner (Journal fur praktische Chemie, Band XVI.), but in this 

 case the light from only one source is plane polarized — it would in 

 fact correspond to my apparatus without the Nicol G. Now by 

 the reflexion at the mirror and refraction through the prisms this 

 unpolarized light becomes partially polarized as the analysing 

 Nicol is turned, therefore its intensity is altered, and an error of 

 unknown — though calculable — amount is introduced into the 

 observations. Arrangements of apparatus of the same nature have 

 been used by Crova (Comptes Rendus, 1881), and Vierordt, Die 

 Anwendung des Spectral- Apparates zur Photometrie der Absoip- 

 tionspectren. 



(3) On a common defect of lenses. By R. T. Glazebrook, M. A. 

 F.R.S. 



The author exhibited some lenses which when placed between 

 two crossed Nicol's prisms shewed strong elliptic polarization*. 



* See letter by the author in Nature, Dec. 28, 1882. 



