SOUND-RANGING 81 



the gun. It is, of course, this latter portion of the record that 

 is used to calculate the position of the gun. The time in- 

 tervals between S and G serve to identify the caliber of the gun. 



The listening instruments were grids of very fine wire elec- 

 trically heated and mounted in the narrow neck of bottle- 

 shaped containers. When the sound from a gun arrived at 

 the container, air was forced in and out by the pressure changes 

 existing in the passing sound wave ; the air rushing in and out 

 cooled the hot wire mounted in the neck of the bottle and this 

 cooling disturbed the flow of the electric current used to heat 

 the wire, and the variation in the flow of current was what 

 actuated the moving part of the galvanometer at the " central " 

 and caused the twitch in the shadow recorded on the moving 

 photographic paper. The listeners were rendered purposely 

 insensitive to loud but high-pitched noises like rifle fire, etc., 

 but purposely very sensitive to grave and sometimes almost 

 inaudible sounds like heavy caliber artillery fire; in fact, for 

 the purpose for which they are designed the listeners were 

 superior to the human ear and were able to pick up German 

 guns as far in the rear as guns were likely to be placed. Very 

 often a gun the report of which had not been audible was 

 found on the same record with a nearer and audible gun. 



The timing device at the " central " station was run continu- 

 ously day and night but the remainder of the apparatus was run 

 only when firing was taking place ; for this reason the apparatus 

 was electrically controlled by observers stationed near the front 

 line trenches; these observers had certain groups of enemy 

 artillery assigned to them for surveillance and they were 

 instructed to start the recording mechanism whenever they 

 heard firing from their assigned areas. There were generally 

 two or more forward observation stations (marked O. P. on 

 Figure 4) to each sound-ranging section, so chosen with refer- 

 ence to the lay of the land, that no enemy-firing on the five mile 

 front could take place without attracting the attention of at least 

 one of the groups of forward observers. 



A typical record not only contained data from which the 



