Water in Hot Glass. 477 



I may here refer to the apparent expansion on solidifica- 

 tion, and the eventual breakage of tubes spontaneously. In a 

 general way, this means an effectively different coefficient 

 of expansion fur the igneous glass without, from that for the 

 water-glass within. Such a result could be brought about, 

 for instance, if latent heat is set free by the water-glass on 

 solidification : for the water-glass is nearly incompressible 

 when solid, and the case would therefore be far different 

 from the solidification of such relatively compressible bodies 

 as paraffin in capillary tubes. 



The characteristic formation of bubbles is evidence against 

 expansion on solidification. Indeed I am not sure whether 

 even an excessive contraction of the core may not start an 

 internal crack in the water-glass which thereafter spreads into 

 the igneous glass. Cracked water-glass cores with a sound 

 envelope of igneous glass were a frequent occurrence. Finally, 

 the congealed water-glass in the cold may tend to resume its 

 volume before the volume-contraction was impressed by 

 spontaneous decomposition. The definite decision can be 

 made only when water-glass has been obtained on a larger 

 scale. 



12. Tubes 4, 5, 6 show conclusively that the action of 

 water on glass ceases after the solution has reached a certain 

 degree of concentration ; compared with tubes 1, 2, 3, they 

 show that the rate of reaction increases in marked degree with 

 rise of temperature. These observations are in favour of the 

 occurrence of a definite limit of saturation * or even of a 

 definite chemical compound. Reaction, cat. par., ceases with 

 its formation. The marked contraction of bulk as the reaction 

 proceeds may be adduced in favour of the chemical view, since 

 in colloidal solution the volume phenomena are usually 

 secondary. 



It is to be noted that, apart from saturation, a gradual 

 subsidence of the rate of reaction would be the result of 

 ordinary colloidal solution, if the concentration gradient at 

 the surface of contact of the water-glass core within and the 

 igneous glass envelope without decreases as more glass enters 

 into solution. The retention of the original composition of 

 the glass with water added, the bodily rejection of colouring 

 material like cobaltic oxide, is further to the point. In how 

 far increased compressibility may be expected in a liquid 

 undergoing colloidal solution I cannot say, even when such 

 solution is accompanied by decreasing volume ; but as the 

 phenomenon of solution terminates with a datum very near 



* Observing that saturation also occurs in colloids below their 

 melting-point in excess of solvent. 



