Aqueous Fusion of Glass. 107 



water at 185° is found to increase regularly from '000077 

 near the beginning of the experiment, to over '000221, or 

 more than three times its initial value on the same isothermal 

 (185°). This result is wholly unexpected, since without 

 exception the effect of solution is a decrease of the compressi- 

 bility of the solvent, in proportion as more body is dissolved. 

 Silicated water in the present experiment shows the reverse 

 effect. Now although the hydration increases the volume of 

 the glass, the gradual choking of the capillary canal goes on 

 uniformly from top to bottom of the thread of water. Hence, 

 since the bore is diminished at the same rate throughout the 

 wetted tube, the observations for compressibility would remain 

 to the same degree unchanged, ccet. par. Supposing that fine 

 particles of the glass were broken off* and gradually accumu- 

 lated on the mercury meniscus near the bottom of the thread 

 of water, it would be possible to account for the data for /3 

 cited, in consideration of the gradual constriction of the 

 thread near the bottom. In such a case, however, the com- 

 pressing thread of mercury would not have advanced and 

 retreated through this debris with the observed regularity. 

 In general mere stoppage and clogging would have been 

 noticed in duplicate experiments f. I also made correlative 

 experiments with saturated solutions of zinc sulphate in water 

 and naphthalene in alcohol. In both cases markedly increased 

 compressibilities were possible in a turbid column, and due to 

 the precipitation of part of the dissolved salt, isothermally, 

 by pressure. During compression a part of the dissolved body 

 is changed from the liquid to the solid state by pressure, and 

 hence the apparent increase of compressibility. 



From this point of view I have endeavoured to account 

 preliminarily for the observed regular increase of j3 given in 

 the table, though I confess some reluctance to this explana- 

 tion : I have supposed that the dissolved silicate is precipitated 

 out of solution by pressure and redissolved on removing 

 pressure, thus producing accentuated compressibility ; that 

 this effect increases as more silicate is taken up in solution, 

 until finally the whole thread becomes too viscous for further 

 observation. However this may be, the fact of a regularly 

 and enormously increased compressibility remains as colla- 

 teral evidence of the stage of progress of the reaction. 



2. There is a final result to be obtained from this experi- 

 ment, and it is to this that my remarks chiefly apply. The 

 reaction of the water on the glass must be along the surface 



* This does not occur. See below. 



t Thus for instance the thickness of the thread of mercury seen in the 

 cathetometer did not seem to diminish. 



