﻿756 Dr. T. J. Baker on Breath Figures. 



the possibility that ions derived from the flame may be the 

 cause of these figures ; but the extraordinary permanence of 

 the effects and the fact that actual contact with the flame 

 gases is not essential render the hypothesis untenable. 



Permanence of breath figures. 

 The peculiar condition of the surface of glass which causes 

 it to reveal the flame-track when it is breathed upon is 

 singularly permanent, and persists for many months. As 

 stated hj Lord Rayleigh and Mr. Aitken, a breath figure may 

 be removed by rubbing with soap and water, but the author 

 has occasionally experienced great difficulty in getting rid of 

 the last traces of the effect. Rubbing with a dry cloth 

 weakens the figures, but does not destroy them. 



Experiments ivith chemically cleaned glass. 



Up to this point the glass plates had been cleaned as 

 described on page 754, and it now appeared necessary to 

 examine the behaviour of glass which had been subjected to 

 chemical cleansing processes such as are employed preparatory 

 to silvering. After the final washing in distilled water the 

 plates were supported on glass rods in a desiccator, and were 

 left there until dry. 



When these chemically clean plates were breathed upon 

 the condensation was almost entirely of the " black" kind — or, 

 in other words, the glass was covered with a continuous film 

 of water. When a flame was made to traverse a chemically 

 clean plate and was afterwards breathed upon, no breath 

 figure, or at most a very imperfect one, appeared, thus sug- 

 gesting that a film of some contaminating material must be 

 present upon the glass before it is exposed to the action of 

 the flame. 



Glass plates which have been cleaned as described on 

 page 754 are certainly covered with a contaminating film, and 

 in this connexion the work of W. B. Hardy and J. K. Hardy 

 (Phil. Mag. July 1919) is significant. These investigators 

 found that truly clean glass surfaces will nut slide over each 

 other, but seize owing to cohesion. A very small amount of 

 contamination lowers the resistance to relative motion, and 

 sliding becomes possible. The author has applied this method 

 to test the condition of the surfaces of glass plates used in 

 obtaining breath figures, and the following details are repre- 

 sentative of the results obtained. 



A chemically clean watch-glass was placed on an equally 

 clean sheet of plate glass as in Hardy's experiments, and it 

 was found that a horizontal pull of 8 grams was needed to 



