﻿Dr. T. J. Baker on Breath Figures. 759 



A flamed plate which had been used for certain expe- 

 riments was by chance placed with its flamed surface down- 

 wards on another glass plate which had not been used. Next 

 day the plates were separated, and it was observed that both 

 plates gave breath figures, one being an exact copy of the 

 other. Moreover, the original showed no diminution of 

 intensity. This accidental observation was many times 

 confirmed, and it became clear that closer investigation was 

 necessary. 



The transference of the effect from one plate to another 

 lying upon it suggested that some volatile material was 

 concerned in the process, and that the escape of this material 

 might be assisted by reduction of pressure or by increase of 

 temperature. A flamed glass was therefore put face to face 

 with a " cleaned " sheet of glass and the two plates, clipped 

 together, were placed in the receiver of an air-pump, and the 

 pressure was reduced to a few cms. of mercury. A few 

 minutes later the plates were withdrawn, separated, and 

 breathed upon, when a perfect copy of the original was 

 obtained on tha "cleaned " plate, whilst the flamed plate still 

 retained its power of producing a breath figure with unim- 

 paired intensity. 



Thus transference occurs as effectively in a few minutes 

 under reduced pressure as it would in the course of hours 

 under atmospheric pressure. 



Experiments were then made with plates separated about 

 4 mm., and a clear transfer was obtained in about 15 minutes. 

 The separation was increased to about 2 mm., and again a 

 transfer occurred, but much diminished in intensity. With 

 a separation of 1 cm. it was not possible to detect with any 

 certainty that transfer had occurred. 



The flames of hydrogen and carbon monoxide also pro- 

 duced transferable figures ; and, since neither of these gases 

 yields solid products on burning, it appears unlikely that the 

 volatile material causing the transfer can have been provided 

 by the flame, It was also found that the transferred figure 

 resembles the original in offering marked resistance to the 

 steady sliding of a watch-glass across the glass plate, although 

 the effect, as might be expected, is weaker. Whatever may 

 be the explanation of the phenomenon of transfer, this fact 

 indicates that the transferred figure represents a portion of the 

 g/</.<s ichicli has been partially cleared of the contaminating 

 film. 



At this point it was thought desirable to discover whether 

 the peculiar properties of a flamed plate are modified by 

 breathing upon it. One half of a flame track was covered up 



