30 The Philippine Journal of Science isis 



and that the finished construction may be protected from them. 

 Although it does not follow that the effect of a substance on 

 cement as shown in the laboratory will necessarily be duplicated 

 with concrete in practice, nevertheless such research is useful 

 in indicating what substances are likely to cause trouble. 



The literature since 1889, when Chandlot * made the important 

 observation that certain substances affect the setting of cement, 

 contains many papers on the subject. Some of these are simply 

 reports of observed failures of concrete, while others are the 

 result of much careful labor. At first it appears that the subject 

 must be fairly well covered, but a study of the papers reveals 

 that there is not much agreement among the various inves- 

 tigators. Table I shows an alphabetical list of electrolytes that 

 have been studied relative to their effect on cement. When the 

 same one is mentioned in more than one paper, the references 

 have been arranged chronologically. 



Table I. — Effect of electrolytes on cement, as reported by a 7nimber of 

 investigators. 



Aluminium chloride. Dobrzynski, Journ. Soc. Chem. Ind. (1892), 11, 525. 

 A cement was gauged with water and with ammonium chloride solu- 

 tions, ranging from 1 to 6 per cent. The lower concentration resulted 

 in an increased tensile strength for 7-day briquettes, while the re- 

 verse was true with the higher concentrations. All the solutions 

 lowered the strength of 28-day briquettes. 



Aluminium chloride. P. Rohland, Ber. d. deutsch. chem. Ges. (1901), 33, 

 2831. 

 The set was accelerated; no details are given. 



Aluminium chloride. P. Rohland, Zeitschr. f. angew. Chem. (1903), 16, 

 1049. 

 The original setting time of a cement was five hours and eight minutes. 

 It was accelerated to one hour and eight minutes by using a 5.5 per 

 cent aluminium chloride solution. With a 9 per cent solution the 

 setting time was four hours and fifty-nine minutes. 



Ammonium sulphate. L. Perin, Journ. Soc. Chem. Ind. (1906), 25, 812. 

 Ammonium sulphate (0.86 per cent) has a greater influence on the set 

 of cement than an equivalent amount of calcium sulphate (on the 

 basis of the sulphuric anhydride content) . 



Barium chloride. Dobrzynski, Journ. Soc. Chem. Ind. (1892), 11, 525. 



A cement was gauged with a solution of barium chloride, ranging from 

 1 to 6 per cent. The tensile strength was considerably increased in 

 every case. 



Boric acid. P. Rohland, Ber. d. deutsch. chem. Ges. (1901), 33, 2831. 

 The set was retarded; no details are given. 



'Chandlot, Journ. Soc. Chem. Ind. (1889), 8, 543. 



