xv, i Witt: Portland Cement Raw Mixture 113 
sample may be directly calculated from the known content of the 
standard by proportion. A table for making these calculations 
covering the usual range of results for both the standard and 
the sample may be easily prepared. Errors, such as those due 
to absorption of carbon dioxide in the hydrochloric acid, lack of 
accuracy in either thermometer or barometer, and the like, are 
thus automatically corrected. Everything considered, the labor 
required by this procedure is no greater than that required by 
the other, and the results are more satisfactory. When the 
sample contains considerably more' silicate than the standard 
there will be an error in the result, the reason for which is ex- 
plained in this paper. It has been proposed to correct this by 
preparing a series of standards ranging from the lowest to 
the highest calcium content likely to be found in a sample, and 
then, for a given determination, using the standard nearest 
that sample in composition. This would be an Improvement, 
but it is doubtful if the increased labor would be justified. 
In the acid-alkali titration method, the sample is placed in a 
300-cubic centimeter Erlenmeyer flask, fitted with a rubber 
stopper bearing a vertical glass tube as a condenser. A measured 
amount of 0.4 N hydrochloric acid is added, and brought just 
to a boil over a Bunsen burner. The solution is cooled, diluted 
to from 100 to 150 cubic centimeters, and titrated with standard 
sodium hydroxide of the same concentration, using methyl orange 
as the indicator. Of the several concentrations of acid tried, 
0.4 N was the lowest that would dissolve nearly all the sample 
without the application of heat. However, that solution was 
not quite complete without warming. Methyl orange is pre- 
ferred to phenolphthalein, although the end point is more dif- 
ficult to see. It is not necessary to remove the carbonates in 
preparing the standard sodium hydroxide solution (as should be 
done if phenolphthalein is to be used), and it is not necessary 
to boil the acid after the sample is dissolved, to remove the 
carbon dioxide. 
The method for determining acid-insoluble matter is the same 
as the one given by Meade, except that hydrochloric acid having 
a density of 1.125 is used. The sample is treated with this 
acid, and the mixture is boiled several minutes. The insoluble 
matter is then collected, washed, ignited, and weighed. 
The raw mixture with which this paper is concerned is made 
from three materials; namely, a hard limestone, and two sili- 
ceous materials that are designated silicate 1 and silicate 2. 
The analyses of these materials are shown in Table I. 
166316—8 
