SOILS OF THE ARID AND HUMID REGIONS. 385 



usual course of soil analysis is generally referred to this min- 

 eral, which contains silica and alumina nearly in the proportion 

 of 46% to 40%. 



In very many cases, however, the reference of these two in- 

 gredients to clay is manifestly unjustified. This is clearly so 

 when (as not unfrequently happens) the amount of alumina 

 found exceeds that which would form clay with the ascertained 

 percentage of soluble silica ; it is almost as certainly so when, 

 in addition to the alumina, other bases (notably potash, lime 

 and magnesia), are found in proportions which preclude their 

 being in combination with any other acidic compounds pres- 

 ent. The only possible inference in such cases is that these 

 bases, together with at least a portion of the alumina, are pres- 

 ent in the form of hydrated, and therefore easily decomposable 

 silicates or zeolites. 



The subjoined analysis by R. H. Loughridge, of a clay obtained in 

 the usual process of mechanical soil analysis (by precipitating with com- 

 mon salt the turbid water remaining after 24 hours subsidence in a 

 column of 200 millimeters) from a very generalized soil of northern 

 Mississippi, shows one of the many cases in which the numerical ratios 

 of the several ingredients are incompatible with the assumption that 

 silica and alumina are present in combination as clay (kaolinite) only : 



ANALYSIS OF COLLOIDAL CLAY. 



Insoluble matter 1 5.96 



Soluble silica 33- IQ 



Potash KaO) 1.47 



Soda (Na'O) 1.70 



Lime (CaO) 09 



Magnesia (MgO) 1.33 



Br. ox. of Manganese (Mn 3 O4) 30 



Peroxid of iron (Fe^Ch 18.76 



Alumina ( A1 3 O 3 ^ 18-19 



Phosphoric acid (PsOs) .18 



Sulf uric acid (SO 3 ) 06 



Carbonic acid (CO a ) oo 



Water and organic matter 9.00 



Total 100.14 



If in this case we assign all alumina to silica, as required for the 

 composition of kaolinite or pure clay, there yet remains a trifle over 

 twelve (12.17) per cent of silica to be allotted to the other bases present. 

 Deducting from this the ascertained amount of silica soluble in sodic 

 carbonate, pre-existing in the raw material (.38 per cent), we come to 

 11.79 P er cent - as tne amount of silica which must have been in com- 

 25 



