128 THE ROYAL SOCIETY OF CANADA 



It was not necessary, however, to weigh the liquid at each ob- 

 servation as shown by the following. Dipping into the liquid was a 

 finely drawn out glass tube, supported by an upright iron stand. 

 The lower end of the tube was bent so as to be parallel to the surface 

 of the water. Held vertically by another support was a graduated 

 1 c. c. glass tube, joined to the other by heavy walled rubber tubing 

 and a tap. To the upper end of the vertical graduated tube was 

 attached a three-way tap leading to an exhaust pump. Hence to obtain 

 the absolute change of pressure required to produce good contact 

 it was only necessary to open the taps, let the water flow down into 

 the beaker on the scale-pan noting the positions of the water column 

 in the graduated tube at the moments when the spot of light on the 

 scale indicated an infinite resistance and a zero resistance or good 

 contact. The difference between those positions of the water column 

 is then the absolute pressure required to produce good contact. To 

 decrease the contact pressure the water was drawn up by the exhaust 

 pump, observations of the water level being made again as before. 

 Thus the effects due both to evaporation and to changes in the 

 weight on the hinged contact jaw for the wires of various sizes 

 and lengths were eliminated for each observation. 



(e) Devices to prevent "bumping" together of contacts. 



As already stated in a previous part of this paper, the lower 

 end of the tube dipping into the beaker was bent so that it was parallel 

 to the surface of the water. Hence there was no vertical force com- 

 ponent due to the inflow or withdrawal of the water. 



Furthermore, the two taps made it possible to regulate the rate 

 of change of pressure. In the actual experimental work the rate of 

 change was from 10 to 20 mgms. per minute. 



(f) Means of producing steadier deflections. 



A further noteworthy feature of the form of the apparatus used 

 was that the force arms of the contact could be varied in length. 

 It was convenient to use arms 6-3 and 3- 15 cms. long, the former be- 

 ing the force arm, the latter the resistance arm. By this device the 

 change of pressure as recorded on the graduated 1 c.c. tube was les- 

 sened, while at the same time the distance, the contact wire moved for 

 a given change of pressure, was diminished. The effects due to the 

 smaller external vibrations were thus removed ; it was easier to obtain 

 steady readings, and the experimenter was enabled to obtain points 

 on the pressure-resistance curve between an infinite resistance and good 

 contact. 



