274 Dr. Y. H. Veley on Initial 



cooling o£ the liquid. Fourteen yeitrs ago the same pheno- 

 menon was observed in other cases in which gases were 

 evolved, as a product of chemical change, from homogeneous 

 liquids, and^ further, that the phenomenon could also be pro- 

 duced temporarily by suddenly increasing the superincumbent 

 pressure and its converse by reducing the pressure ; also that 

 the addition of finely- divided particles causes an increase in 

 the rate of evolution of gas. 



These phenomena resemble in general character those ob- 

 served by Oersted, Schonbein, Gernez, Schroder, Tomlinson, 

 &c., in the case of supersaturated solutions of gases, especially 

 carbon dioxide^ whether in artificial preparations such as so- 

 called soda-water, or in natural waters such as those of Spa 

 and Langen-Schwalbach. 



In all cases the method of observation was that known as 

 the twin U-tube method, which has previously been described ; 

 in those in which the evolved gas is practically insoluble in 

 water the U-tubes were inverted over it, but in others in 

 which the gas is more soluble, then either a liquid other than 

 water was used, or the gas was made to displace its own 

 volume of water which had previously been saturated with 

 it. This latter method of observation is not, however, so 

 convenient, as the observations cannot be made quite con- 

 tinuously. The advantage of this method is that constant 

 units of chemical change can be measured. 



Example I. The evolution of carbonic oxide from formic 

 acid by sulphuric acid. It was found that after the period of 

 maximum constancy the carbonic oxide, taken as a measure 

 of the decomposition, is evolved in accordance with the 

 bimolecular law 



dOldT=-kO'' (6) 



but previously to that period the rate of evolution was not in 

 accordance with that law. If then x is the amount of carbonic 

 oxide formed after an interval T from the commencement, 

 and C the total amount w^hich can be formed, then 



Q-x={C-^ + kT)~\ ..... (7) 

 further, if a portion of the carbonic oxide is detained in the 

 liquid (while a supersaturated solution is being formed) in con- 

 tinually diminishing quantity ^, and y is the amount evolved 

 after an interval T, then, during an unit interval, dz is detained 

 while dy is evolved, and dz is a continually diminishing 

 fraction of dy. The hypothesis which Esson suggests is 

 expressed by the relation 



dzldT = nbex^. {-ny)dy/dT, . . . . (8) 

 which gives z = h{l — ex^.( — ny)} = h(l-m-y), . , (9) 

 z being when 2/ = and approaching a limit h as j/ increases. 



