58 



BRIDGMAN. 



all the individual substances for which the data are now at hand will 

 then be presented, and finally, the significance of the data and their 

 suggestions as to a possible mechanism will be discussed. 



Methods. 



The apparatus was the same as that which has already been de- 

 scribed and most of the measurements were made at the same time 

 as those of preceding papers. The only measurements made with this 

 subject alone in view were a few on potassium chlorate. To every 

 equilibrium point of the preceding papers there may con-espond two 

 velocity curves, giving the rate of reaction with rising and falling 

 pressure. But in many cases it was not practicable to obtain these 

 two curves, especially if the reaction were a rapid one or one accom- 

 panied by a small change of volume, so that the data given here are 

 by no means so complete as the equilibrium data. 



The following brief description of method applies to readings all 

 at the same temperature. For definiteness we -will suppose that we 

 are measuring the rate at wliich the low pressure phase (I) changes to 

 the phase stable at higher pressures (II). Pressure on I is first slowly 

 increased until it has been carried so far into the region of stability 

 of II that nuclei of II are formed spontaneously. After the formation 

 of nuclei, the phase II grows at the expense of I by an advance of 

 the surface of separation. The formation of II is accompanied by 

 dropping of the pressure back toward the equilibrium line. The 

 reason for this fall of pressure is, of course, that the phase stable at the 

 higher pressure necessarily has the smaller volume. Since the fall 

 of pressure is proportional to the change of volume, and therefore to 

 the amount of I changed into II, the reaction velocity may be simply 

 measured by measuring the rate at which pressure drops back. It is 

 obvious that a precisely similar procedure gives the rate at which II 

 changes to I, the pressure now rising as the reaction runs. These 

 two sets of readings with rising and falling pressure are to be repeated 

 at different temperatures. 



For the measm'ements of time an ordinary watch was used, and 

 the pressure was determined in the usual way by measuring with a 

 Carey Foster bridge the resistance of a coil of manganin wire sub- 

 jected to the pressure. The resistance was given b;}- the position of 

 the slider on the bridge wire, the setting being made for no galvanom- 

 eter deflection. It can be well understood that this measurement 



