78 BRIDGMAN. 



exhibited by all substances, but for most substances the rate of advance 

 of the surface in the region of nucleus formation is so rapid that it 

 was not possible to make any measurements. The existence of the 

 effect was established for AgNOa, KCIO3, Acetamide, and Carbamide. 

 The effect was in most cases shown only on the rising branch, on which 

 the velocity was less, thus making it easier to observe. Carbamide, 

 however, showed two examples on the falling branch, and acetamide 

 one. 



The question of the extent of the region in which no nuclei are 

 formed is of interest. As has been said, no fixed limits can be assigned 

 to this region, but the formation of a nucleus is a matter of chance. 

 It is to be expected, therefore, that occasionally nuclei will be formed 

 outside of the usual region. The effect of such a freshly formed 

 nucleus will be to add more surfaces of separation, and so to increase 

 the measured rate. Such effects were found for two substances, 

 AgNOs, and KCIO3. Several of the rate curves for AgNOs with rising 

 pressure showed a secondary maximum or a region of arrest, to be 

 explained in this way. The effect was so pronounced with KCIO3 

 that these secondary maxima appeared persistently throughout the 

 entire region of observation with rising pressure at 0°. It is evident, 

 therefore, that different substances differ greatly in the sharpness of 

 the boundaries of the region of nucleus formation. For most sub- 

 stances the boundary is so sharp that no chance nuclei were ever 

 formed in the region of observation. Furthermore with AgNOs 

 and KCIO3 the boundary of the region at pressures above equilibrium 

 was very much sharper than at pressures below; no secondary maxima 

 or regions of arrest were observed with falling pressure. In the case 

 of x^gNOs the boundary was sufficiently sharp so that the extra velo- 

 city due to additional nuclei did not affect the limiting acceleration; 

 the extra surfaces diie to the extra nuclei having merged into the other 

 surface before the final reading. This was shown by the regularity 

 of the limiting accelerations, which lie on a fairly smooth curve, irre- 

 spective of whether there had been a secondary maximum or not. 

 The disturbing eft'ect with KCIO3, however, extended throughout the 

 entire region of observation at 0°, so that it was not possible to estab- 

 lish any limiting acceleration. At higher temperatures the boundary 

 for KCIO3 became much sharper and the curves became normal 

 in every respect, as shown by Figure 4. 



The question of the sharpness of the boundary of the region of 

 nucleus formation is an interesting one for investigation, both experi- 

 mental and theoretical. It has received little attention, but would 



