Atkins — Factors affecting Hydrogen Ion Concentration of Soil. 389 



From what has been said already it is evident that the snpply of bases 

 capable of neutralizing plant remains is the primar'y factor \\hich regulates 

 soil reaction. In an interesting series of papers, a list of which may be found 

 in the bibliography, from 1916 onwards Wherry studied the relations of the 

 strata to the soil and of the latter to plant distribution. Thus granitic and 

 quartzite rocks have no bases free to act in neutralizing acidity. Slaty rocks 

 are in much the same position, and so are many sandstones. Probably some 

 slates contain a minute amount of carbonate, and sandstones, especially when 

 porous and underlying limestone, may contain notable amounts. The reac- 

 tion that deeper strata are capable of giving may be indicated by springs, as 

 shown further on. 



The following observations may serve to illustrate the relationship. The 

 determinations were made by the colorimetric method, using about a gram 

 of soil to lOc.c. of water, and testing the approximately clear supernatant 

 liquid. If turbid, the device was adopted of viewing the standards through 

 tubes containing the turbid extract without indicator, the extract of the 

 sample containing indicator being viewed through water to compensate. 

 As explained by Clark, dilution in many cases has not a large effect upon 

 pH, and this was tested with a fine calcareous silt by using only one-fifth 

 the usual quantity of soil and obtaining an indistinguishable colour. The 

 advantage of thus obtaining a clearer solution was obvious. A slight turbi- 

 dity often increases the buffer action of the solution and lessens errors. 

 With acid soils it appears that larger quantities of solid are necessary to 

 avoid dilution effects. It has sometimes been found convenient to mix the 

 soil with a slight excess of water in a watch-glass, and to test a drop draining 

 away against a white background. As explained by the writer in describing 

 the determination of the acidity of plant tissues, use is made of drops of the 

 standard solutions, or even of the latter rendered turbid by the soil extract. 

 Since the extract has relatively slight buffer action, and that of the standard 

 is intentionally very considerable, the latter is not seriously altered by this ; 

 and if identical in pH value with the sample, the error is obviously zero. 

 As a rule, two, or even three, indicators were used in fixing the pH of a 

 sample. 



In addition to the indicators given in their final list by Clark and Lubs, 

 it has been found that di-ethyl red — also tested by these workers — is very 

 useful for a region just belovs' that of methyl red. Di-ethyl red changes 

 from yellow to red between pH 6'0 and 5'8, and is increasing in colour just 

 where methyl red is yellow, and where brom cresol purple is turning yellow 

 also. A red colour is very easily observed in a turbid solution, so the 

 value of di-ethyl red in checking the results given by other indicators is 



