56 GENERAL DIS0VS8I0N OF CLAT8. 



rational eomposition has changed, there being a mark- 

 ed increase in the amount of feldspar. 



If there hed been much difference in the size of the 

 clay-particles of Nos. 3 and 4 or Nos. 1 and 2, the 

 shrinkage in each case would probably have been dif- 

 ferent. 



The degree of fineness of the clay-particles, and per- 

 haps their shape also, probably exert more influence 

 on the shrinkage than has been imagined, but just how 

 far this makes itself felt is still undetermined. 



As an illustration of the practical use of the rational 

 analysis we may take the following : 



Suppose that we are using for the manufacture of 

 porcelain or fire-brick a kaolin which has 67.82 per 

 cent, of clay-substance, 30.93 of quartz, and 1.25 of 

 feldspar, and that to 100 parts of this is added 50 

 parts of feldspar. This would give us a mixture of 

 45.21 per cent, of clay substance, 20.62 of quartz, and 

 34.17 of feldspar. 



If now for the clay we had been using, we substitu- 

 ted one with 66.33 per cent, of clay-substance, 15.61 

 of quartz, and 18.91 of feldspar, and made no other 

 changes, the mixture would then contain 44.22 per 

 cent, of clay-substance, 10.41 of quartz and 45.98 of 

 feldspar. 



This last mixture shows such an increase in feldspar 

 that it must give mucli greater shrinkage and fusibil- 



