RATIONAL ANALYSIS OF CLAY. 55 



From this table a number of interesting conclusions 

 may be drawn. Columns 1 and 2 represent two clays 

 which agree very closely in their ultimate composi- 

 tion ; but in the rational analysis there is a difference 

 of 6 per cent, in the clay-substance, 12 per cent, in 

 quartz, and nearly 19 per cent, in the feldspar. Nos. 

 3 and 5 and 10 and 12 also illustrate this point. 



In Nos. 6 and 7, one a German, and thie other a 

 North Carolina kaolin, the ultimate analyses are very 

 closely alike, and the rational analyses also agree very 

 well. This is frequently the case when the clay-sub- 

 stance is very high, between 9.6 and 100 per cent., as in 

 Nos. 9 and 11. 



A third case would be presented if the rational an- 

 alyses agreed, but the ultimates did not. Such in- 

 stances, however, seem to be much less common. 



The practical value of the rational analysis bears 

 chiefly upon those branches of the clay-working in- 

 dustry, such as manufacture of porcelain, white earth- 

 enware, fire-brick and glasspots, which use materials 

 with comparatively few fusible impurities ( iron, lime, 

 magnesia) . 



There is much concerning clays which sitll remains 

 unexplained, but it seems probable that, other things 

 being equal, two clays having the same rational com- 

 position will behave alike. 



We can illustrate this point by the foUownig tests 

 made on washed kaolins from the vicinity of Senne- 

 witz, near Halle, Germany. From the figures given 

 below, it will be noticed that in the case of Nos. 1 and 

 2 there is a close agreement in the shrinkage, which 

 amounted to about 10 per cent, when the clay was 

 heated up to the temperature of a hard-porceclain 

 kiln. In Nos. 3 and 4 the shrinkage is very nearly the 

 same, but greater than in Nos. 1 and 2, because the 



