220 THE ROYAL SOCIETY OF CANADA 



(c) Envelopment hy precipitation of the hydroxides of alumi- 

 nium, zinc and cadmium, and by colloidal arsenic sulphide. 



(d) Cataphorcsis. 



The general result was that the light beam in watei and in aqueous 

 alcohol — the two liquids most thoroughly investigated — consists of 

 two parts; a part which is removed by each of the methods of puri- 

 fication and a part which is not removed by any of the methods and 

 which is constant in intensity irrespective of the method of 'purification 

 used. This permanent part is faint and is plane-polarized, and can 

 be seen only if the room is dark, the light beam very intense, and the 

 vessel clean and free from striae. 



The intensity of the permanent opalescence of water was not 

 affected by change of temperature. It was also unaltered by the 

 addition of hydrofluoric acid and is therefore not due to silica. 



A sample of water, purified by method (a), was cooled in a tube 

 2 mm. in diameter to -26°C before it froze. Ordinary distilled water 

 under the same conditions froze at -11° C. {Winter 1912-1913). 



