MEASUREMENTS 



267 



or by pressing once or twice 

 on the button (fig. 154). 



When the experiment lasts 

 more than ten minutes, clocks 

 are used with a loud tick, 

 which produce a definite sound 

 at the end of each minute. 

 With care an ordinary watch 

 can be used. The error here, 

 is at the most two or three 

 seconds per hour, or a re- 

 lative average of y^no- For 

 instance, a man working hard 

 would eliminate 1,500 litres 

 of gas, and the relative error 

 would be 1 litre per hour. 



198. Experiments concern- 

 ing the measurement of very 

 rapid occurences, especially 

 nervous phenomena, and the 

 phenomena of shock, necess- 



F|G - u4 :^ itate the use of chronographs 



Stop Watch. j. ,, -, 



reading to the T tjo> an d 



even the xoVo f a second. Electrically driven tuning forks 

 can be used. The vibrations 

 of the tuning-fork can be 

 registered directly, by the in- 

 termediary of an electric in- 

 dicator, of which the best 

 known is that of Marcel 

 Deprez. 



In the former case, the two 

 prongs of the fork have be- 

 tween them an electro-magnet, 

 connected to a battery, P, as 

 shown (fig. 155). The circuit 

 being closed the prong B is attracted by the magnet and breaks 

 contact with a platinum point p. The circuit is then opened and 

 the fork returns to make contact again with the point p. 



A tuning fork can thus be made to vibrate continuously at the 

 rate corresponding to its pitch. By attaching a very light style 

 to one of the prongs, intervals of time corresponding to the period 

 of the fork can be registered. 



In the latter case, Deprez 's indicator is introduced in the cir- 

 cuit of the battery. The indicator comprises a very small electro- 

 magnet, E (fig. 156), which receives current from the same battery 



FIG. 155. 

 Electro-tuning Fork. 



