378 ^^6 Philippine Journal of Science isu 



produces corrosion at the point of discharge. In badly designed boilers, 

 oxygen introduced in this way together with the air left in the boiler is 

 apt to accumulate in pockets due to improper circulation and is capable of 

 causing the corrosive action, technically known as pitting because of the 

 small holes or "'pits" found in the damaged area. Pitting is the most 

 dangerous form of corrosion. The presence of carbon dioxide together with 

 the oxygen in a boiler water increases its corrosive action. Cases have been 

 known where the mixing with air of hydrogen sulphide and carbon dioxide 

 produced by the decomposition of sewage entering a water supply have 

 caused very rapid corrosion. 



The use of alum for coagulation in mechanical filtration is sometimes 

 bad, for alum may react with bicarbonates to change them into sulphates 

 and thus liberate carbon dioxide and cause corrosion of the boiler. A 

 feed water which contains considerable quantities of magnesium salts 

 when purified by sodium carbonate alone will liberate carbon dioxide 

 as represented by the following equation : 



MgSO,+MgCl=+2Na.C03+H.O^Mg(OH):MgCO.+ 

 Na.S04+2NaCl+CO=. 



Schreiber *" calls attention to the corrosive action of such a water on 

 a boiler especially in the region of the intake. In the absence of air, iron 

 is only slightly attacked by an excess of carbon dioxide." 



The rusting of iron in the presence of air is much more energetic when 

 in contact with water containing chlorides (NH.Cl, MgCL, KCl, NaCl, 

 CaCls, BaCU)." Magnesium chloride, while not a scale-forming ingredient, 

 is accepted as an exceptionally active corrosive agent. Sometimes the water 

 is red or black, which in either case is evidence that energetic corrosion is 

 in progress. It often happens that there is pitting over the entire inner 

 surface of the boiler, but again it may be local in its action. It sometimes 

 has been explained that magnesium chloride is much more injurious than 

 the other chlorides through the hydrolytic splitting off of hydrochloric acid. " 

 Magnesium chloride cannot split off acid without simultaneously forming 

 a basic magnesium compound. The injurious corrosion resulting from the 

 presence of magnesium salts is not due to reaction of free acid but to the 

 interaction of the salts themselves. Ost,*' working experimentally in the 

 absence of air and under a steam pressure of about 10 atmospheres, has 

 shown that with boiler plate in contact with pure water or waters contain- 

 ing about 5 per cent of magnesium chloride, potassium chloride, sodium 

 sulphate, potassium sulphate, calcium sulphate, and magnesium sulphate, 

 respectively, there is rusting in all cases. That is, after each trial the 

 interior of the experimental boiler was covered with an amount of ferroso- 

 ferric oxide (Fe.O.) which he attributed to the oxidation of iron through 

 the decomposition of the hot feed water." Only when magnesium salts 



'^Chem. Zeitg. (1903), 27, 327. 



" Howe, J. L., and Morrison J. L., Jourv. Am. Ckem. Soc. (1899), 21, 422. 



"Wagner, A., Dingler's polytech. Journ. (1875), 218, 70. 



** The action of chlorides on copper is quite different, for the latter 

 forms double salts with chlorides and in this way goes into solution. 



" Chem. Zeitg. (1902), 26, 819, 845. 



" It is a well-known reaction that oxygen interacts vigorously with iron 

 and similar metals when the latter is at a red heat. The slow reaction 

 which involves the decomposition of water is of even greater interest. 



