Films of Sodium and Potassium. 103 



There seems to be ho doubt about the fact that there are- 

 numerous examples of colour production which cannot be 

 explained by any of the known principles of interference and 

 diffraction. 



The results which I shall now describe lead me to believe 

 that there is some peculiar action of deep pits and crevasses 

 upon light-waves, which it is of great importance to investi- 

 gate theoretically. 



Prof. Mallock, in a recent conversation, told me that he 

 also had come to the conclusion that cavities or pits gave 

 rise to optical phenomena of unique type. 



The nature of the granular deposit of sodium, which we 

 shall take up first, depends somewhat upon the condition of 

 the glass surface upon which the condensation takes place. 

 If the glass is perfectly clean the thin deposit which first 

 forms is sky-blue in colour by transmitted light. If, how- 

 ever, the glass surface has a trace of any oil or an adsorbed 

 layer of a hydrocarbon, the deposit is pink tending towards 

 purple. 



The bulbs used were about 3 cm. in diameter, and were 

 freshly prepared for each experiment. A fragment of 

 sodium, about 1 mm. 3 , cut from a dry lump, i. e. one free 

 from hydrocarbons, was introduced into one stem of the bulb, 

 which was quickly sealed. The other stem was connected 

 with a mercury pump, the bulb being brushed with a bunsen 

 flame during the exhaustion. The stem containing the 

 metal was then cautiously heated until a quantity passed 

 over into the bulb sufficient to coat the interior with a deep 

 blue film. This film was then driven about from one side of 

 the bulb to the other by means of the bunsen flame, the pump 

 working continuously. This operation gradually removes 

 the hydrogen which is condensed on the wall along with the 

 sodium. A few repetitions of the process are sufficient in 

 the present case, after which the bulb is sealed off from 

 the pump. 



A uniform film must now be formed on one side of the 

 bulb by means of a pointed gas flame about 1 cm. in length 

 burning at the tip of a glass tube drawn down to a capillary. 

 One hemisphere is first freed from all traces of deposit by 

 brushing the exterior with the flame. The opposite side is 

 then carefully heated, the flame being first applied at the 

 centre of the deposit and worked gradually towards the edge. 

 If the conditions are right a film will form which shows In- 

 interior reflexion a colour or colours as pure and brilliant as 

 those shown by a burnished silver film varnished with a 

 coloured lacquer. The colour shown by a thin deposit should 



