92 



SCIENCE PROGRESS 



point on the hyperbola which has these stations, E and F, for 

 foci and is such that EP — FP = the observed time difference. 



If, now, a third station be set up and its time reading be 

 taken in conjunction with one of those already used, a second 

 hyperbola can be constructed upon which the gun must also 

 lie. The weapon must, therefore, be at the intersection of the 

 two curves, and so its position is determined. 



In practice six such stations were employed, and were 

 arranged in general on the arc of a circle just behind the front 

 lines. They were situated about a mile apart, so that sound 

 would take something like four and a half seconds to travel 

 directly from one to another. 



The accuracy of the results does not depend on having 

 a very large number of observing stations close together, but 

 on having a few stations separated as widely as possible, so 

 as to give a large time difference. The accuracy of the whole 

 system depends, indeed, on the correct measurement of these 

 time differences. Mechanical means have to be employed, 

 for human agency cannot be relied upon. It is found in prac- 

 tice that from one-twentieth to one-fifth of a second elapses 

 between the time a man hears a sound and the time he presses 

 a button to record it, and, moreover, that this quantity varies 

 from man to man, so that no satisfactory record can thus be 

 made when the time differences are required accurately to the 

 order of about one-hundredth of a second. This difficulty 

 was overcome in the British Army by the invention of the 

 Tucker microphone, an instrument that records the arrival 

 of the sound instantaneously. The details of the constructing 

 of this instrument cannot be published, the officers and 



men of the sound-ranging 

 sections having been pledged 

 to secrecy. 



The record made by this 

 instrument took the per- 

 manent form of a photo- 

 graph of the shadows of the 

 strings of a very sensitive 

 galvanometer, the arrival of 

 the sound being indicated 

 by a " kick " on the cinema 

 film which was used for this 

 purpose. 



The film is graduated in 

 hundredths of a second of 

 time (tenths are shown in Fig. i), and therefore can be read 

 to one-thousandth part of a second, the graduation marks 

 being made by the flickering of a light caused by the rotation 



7 second 



/^ 



^V 



Fig. I. 



