18 THE FOKMS OF WATER IN 



our inference by employing our opaque iodine filter 

 Placing it on the path of the beam, the light is entirely 

 stopped, but the water boils exactly as it did when the 

 full beam fell upon it. 



51. The truth of the statement made in paragraph (34) 

 is thus demonstrated. 



52. And now with regard to the melting of ice. On 

 the surface of a flask containing a freezing mixture we 

 obtain a thick fur of hoar-frost (Par. 14). Sending the 

 beam through a water-cell, its luminous waves are con- 

 centrated upon the surface of the flask. Not a spicula 

 of the frost is dissolved. We now remove the water-cell, 

 and in a moment a patch of the frozen fur as large as 

 half-a-crown is melted. Hence, inasmuch as the full 

 beam produces this effect, and the luminous part of the 

 beam does not produce it, we fix upon the dark portion 

 the melting of the frost. 



53. As before, we clench this inference by concentra- 

 ting the dark waves alone upon the flask. The frost is 

 dissipated exactly as it was by the full beam. 



54. These effects are rendered strikingly visible by 

 darkening with ink the freezing mixture within the 

 flask. When the hoar frost is removed, the blackness 

 of the surface from which it had been melted comes 

 out in strong contrast with the adjacent snowy white- 

 ness. When the flask itself, instead of the freezing 

 mixture, is blackened, the purely luminous waves being 

 absorbed by the glass, warm it ; the glass reacts upon 



