VELOCITY OF POLYMORPHIC CHANGES BETWEEN SOLIDS. 63 



two values, therefore, for the hmiting transition acceleration, those 

 with rising and falling pressure, and these two values may be plotted 

 as points in another diagram. 



It is evident from the diagrams that there is very little possible 

 ambiguity as to the way in which the curves should be extrapolated 

 to the axis. The range of velocities covered by the readings is very 

 wide, and it Avas possible to approach very close to the axis in nearly 

 every case. In se\eral cases the lowest observed velocity was 5000 

 times less than the greatest. That the character of the velocity curves 

 is such as to allow this extrapolation is an important fact that will be 

 referred to again in the discussion. One consequence of this is, how- 

 ever, to be insisted on here. The points at which the extrapolated 

 curves for rising and falling pressure at the same temperature cross 

 the axis are not the same. This is particularly evident in Figures 3 

 and 4 for phosphorus and KCIO3. This means that there is a pressure 

 range within which the reaction will not run at all, even when the two 

 phases are in contact with each other. This range may be called the 

 " region of indifference." Now as a matter of fact the probability is 

 that this statement is not rigorously true; in at least one case it was 

 possible to just detect the progress of the reaction within the limits of 

 the "region of indifference." In another case just as careful searcli 

 failed to detect any progress of the transition within the region. This 

 was on the I-II curve of TINO3 at 3100 kgm. In any event the velo- 

 city within this region must be so small as to belong to phenomena of 

 ciuite a different order. Thus in the best marked case, that of car- 

 bamid, the curve plotted for a range of transition velocities of 1000 

 fold was of exactly the character of those shown above, indicating 

 unambiguously an extrapolated pressure of zero velocity, but below the 

 lowest measurable point the curve apparently turns and runs along 

 nearly parallel to the axis. This means that if Figures 2-5 were 

 drawn on a scale one meter long, the curves might turn abruptly and 

 run along the axis at less than 1 mm. distance. x\lthough it was not 

 possible to measure the rate of transition within the region it was 

 possible to state that it was of the order of 0.00002 parts per minute. 

 The reaction would not, by its own progress, carry the pressure into 

 this region in practical limits of time. Pressure had to be artificially 

 shifted into this region and the subsequent reaction observed. Various 

 special precautions were necessary in these readings because of the 

 extreme minuteness of the effect. In view of the great difference in 

 order of magnitude, it seems justifiable to suppose that the mechanism 

 involved on the two parts of the curve is different. In the later dis- 



