destruction oe concrete by alkali 
5 
ted the presence of a large amount of water, also of considerable quantities 
of sulfuric and carbonic acids. The decomposed mass was washed with 
distilled water to remove as much sand as possible. Examination of the 
wash water showed that it had taken sulfuric acid, lime and soda into sol¬ 
ution. The portion soluble in distilled water may have been partly derived 
from other sources but probably came, for the most part, from the altered 
cement. The wash water, therefore, was evaporated to dryness and the 
residue added to the insoluble portion and the whole dried at bo^C. The 
material thus prepared was submitted to analysis with the following re¬ 
sults; with which is also given an analysis of a sample of the same brand 
of cement. 
PORTLAND CEMENT DECOMPOSED BY ACTION OF ALKALI 
WATER 
Original Cement 
Decomposition 
Same calculated on 
Product* 
water free basis 
SiO, . 
22.55 
10.98 
14.16 
SO3 . 
1.29 
15.56 
20.07 
COo . 
1.07 
9-39 
12.16 
Ch . 
not det’d 
0.51 
0.65 
P.O. . 
trace 
CaO . 
59.86 
29.14 
37.58 
MgO. 
1-75 
5.79 
7.46 
ALO3 . 
5.88 
2.08 
2.68 
Fe ,03 . 
4.83 
2.51 
3.23 
M11304 . 
0.21 
trace 
Na ,0 . 
0.76 
1.27 
1.63 
K.,6. 
0.63 
0.29 
0.38 
HoO Ignit. less CO3 
0.72 
(22.48) 
99-55 
100.00 
100.00 
The loss on ignition in the analysis of the decomposition product agreed 
fairly well with that obtained by difference, but owing to the presence of 
alkalies, carbonates and sand the results of direct ignition are not to be 
relied upon except to show that the difference indicates the amount of 
water present with reasonable accuracy. 
The composition of the water acting on this tiling may safely be assum¬ 
ed to be represented by the following analysis of a sample of Arkansas 
river water, essentially a return water, taken at Rocky Ford. 
The summation of the analysis presents a rather large excess over one 
hundred which gives an excess of about one grain in the total amount of 
salts present. The excess of bases shown in the analysis as combined is 
not necessarily due to bad analytical work. The waters of our mountain 
streams show an excess of acids over the bases, but the alkalies forming 
efflorescences on our soils and alkaline waters, frequently show an excess of 
the bases, but seldom so large as in this case. 
*6.01 per cent, sand deducted. 
