Atkins — Factors affecting Hydrogen Ion Concentration of Soil. 381 



The relation of these carbonates to marine plants will be discussed else- 

 where, but it may be added that the photosynthetic activity of both marine 

 and fresh-water algae or phanerogams is a very efficient process for removing 

 dissolved carbon dioxide and biearbonates (Angelstein (1911), Nathansohn 

 (1907)), so that a point at or close to the maximum pH value for the carbo- 

 nate in the solution may be attained provided this does not prove fatal to 

 the assimilating plant. It may, however, be the means whereby one plant 

 destroys another, as has been found to happen with Ceramium rubrum in 

 presence of actively assimilating Ulva laduca. 



It is not, of course, suggested that a soil containing appreciable amounts 

 of magnesium carbonate is at all as alkaline as pH 10, but this upper limit 

 is higher than that of calcium carbonate. It is possible that the widespread 

 dislike among agriculturists (see Hall, 1910; Eussell, 1912; Aston, 1916) 

 towards using limestone rich in magnesium carbonate may have some 

 connexion with this. It has been shown by Hardy (1921) that dolomite 

 retards nitrite and nitrate bacteria, which grow normally with calcium 

 carbonate. 



When other salts are present in addition to calcium carbonate the 

 complex mixture may have appreciable effects upon the less soluble 

 constituents. Thus when Merk's pure precipitated calcium sulphate was 

 dissolved in cold water a solution of pH 7'9 was obtained; on boiling to 

 decompose bicarbonate pH 8"3 was found. Thus the traces of carbonate 

 present were insufficient to produce the maximum alkalinity pH 9'0. How- 

 ever, when equal volumes of calcium sulphate solution at pH 7'9 and 

 calcium carbonate (bicarbonate) at pH 7"2 were mixed and boiled, the result- 

 ing reaction was over pH 8'8, which suggests a lessening of the solubility of 

 the carbonate through the presence of calcium ions derived from the sulphate.. 

 This is conclusively proved to be the correct explanation by boiling water 

 containing both solids in excess. The resulting solution when quickly cooled 

 and tested was at pH S'O approximately. 



Similarly magnesium sulphate, a good commercial sample, was found to 

 give pH 8-0, rising to pH 8'5 on boiling. Kahlbaum's salt was, however, 

 neutral in solution. On performing a qualitative test with the sulphate and 

 carbonate after boiling pH 9-3 was not surpassed, and with the chloride 

 pH 9'2, the differences doubtless being due to the amounts of sulphate and 

 chloride having been added in small indefinite quantities. Magnesium 

 nitrate was found to give pH 96; thus it must have contained an appreciable 

 amount of carbonate. Again, magnesium phosphate in the cold gave pH 8'3, 

 but on boiling this value decreased to pH 7'0, which, after an hour, only 

 altered to pH 6'8 ; boiling for an extra half hour changed it to pH 6'5. A 



