Molecular Constitution of Water. 471 



Under one atmo we found p 2 = \375 at 0°, so that at this 

 temperature dp 2 /df= '024/149 = -00016, in fair agreement, as 

 it ought of necessity to be, with the value found from com- 

 pressibilities, namely, '00017. 



In order to disentangle the dissociation terms of (C) from 

 the expansion terms, we must make the dissociation produced 

 by pressure at the different temperatures run a similar course 

 to that in Table IV. A little arithmetic leads to 



MrtM . '087724 -05398 1Aonc nm 



•° 77M ?= 1 + -00837ft + 1-001077. - 10298 ' (19) 



•46128 _ -62186 _ 



l-*(r252# + * 1 ) ~ 1--001077* " ibUDb ' 



In one gramme of water the amount of 2 (trihydrol) at 

 any temperature being '351 (1/2 + q), we have the following 

 amounts of 2 under 150 atmos for comparison with those 

 under 1 atmo. 



Table VI. 



t 0° 20° 40° 60° 80° 100° 



10 3 jo 2 (150 atmos) 351 300 264 237 217 203 



10 3 p 2 (l atmo) 375 321 284 255 234 217 



10 3 Ap 2 24 21 20 18 17 14 



Smoothing these last differences and dividing them by 

 150, we get: — 



Table VII. 



10 3 Ai? 2 24 22 20 18 16 14 



-Wdpjdf 160 147 133 120 107 93 



If the values just found for dp 2 /df hold up to pressures at 

 which the whole of 2 is dissociated, it appears that at 

 •375/-00016 or about 2300 atmos the whole of the 2 (tri- 

 hydrol) will be dissociated at 0° ; at 100° the pressure of 

 complete dissociation would be '217/-000093, also about 

 2300 atmos. These results agree well with Amagat's con- 

 clusion that at high pressures, even below 3000 atmos, the 

 irregularities of water disappear; water becomes an ordinary 

 liquid, that is to say, the mixture of 2 and 1 has been con- 

 verted into pure compressed 1. 



The effect of pressure in dissociating 2 has an important 

 bearing on the composition of the surface-film of water. If 

 pressure causes dissociation, tension may be expected to pro- 

 duce association, and therefore we may expect the proportion 



