284 The Philippine Journal of Science i9i6 



with the effect of the sulphide on cement and, accordingly, an 

 effort was made to determine its cause. The briquettes, in a 

 measure, retained the green color even until the time of breaking, 

 but is was noticed that the color was more intense in the interior 

 than on the surface. 



The distribution of the colored areas is shown in Plate I. 

 Broken briquettes were selected at random, ground square on an 

 emery wheel, and photographed while wet. The section shown 

 is at the point of fracture. The briquettes were broken at the 

 end of twenty-eight days, and were gauged with solution No. 1, 

 containing 9.70 grams sulphide per liter. The upper two are 

 neat briquettes, and the lower two, mortar briquettes. Nos. 

 1 and 3 contain no sulphide; they were photographed simply 

 for comparison. It is interesting to note the irregular distribu- 

 tion of the color in the mortar and the fairly uniform distribu- 

 tion in the neat cement. 



The boiling pats lost the color on the upper surface, which 

 was exposed to the steam, but the surface which remained next 

 the glass plate and the interior retained the color. In order to 

 demonstrate that the color was not produced by the action of the 

 sulphide on the steel trowels, or on some foreign substance about 

 the mixing table, some of the cement was mixed with a sulphide 

 solution, where the mixtures came in contact with nothing but 

 a glass plate and a porcelain spatula. The color appeared as 

 before. Since iron is the only element present, in any quantity, 

 which is known to give a dark-colored sulphur compound, it 

 was believed to be involved in the reaction. This idea was 

 strengthened by the fact that de Koninck and Ledant ^* report 

 the formation of a bright green on treating iron salts with sodium 

 or potassium polysulphide. The first effort to determine the 

 factors in the reaction was as follows: 



Brand E, which is one of the most sensitive of the cements 

 under consideration to the influence of sulphide, as shown in 

 both the setting and the tensile-strength test, was chosen for 

 the preliminary work. Accurately weighed 5-gram duplicate 

 portions of the cement were placed in beakers. Ten cubic centi- 

 meters of a sulphide solution, the analysis of which was known, 

 were added; after thoroughly mixing, hot water was added and 

 the mixture was stirred. After settling, the cement was sepa- 

 rated from the supernatant liquid and washed until the wash 

 water showed no further tests for sulphide. The filtrates and 



" De Koninck, L. L., and Ledant, M., Zeitschr. f. angew. Chem. (1891), 

 202. 



