52 
BULLETIN 126, U. S. DEPARTMENT OF AGRICULTURE. 
cipal reason for it arises from the reaction between the various alkali 
salts and the calcium hydroxid of the cement. The new compounds 
formed have a greater volume than the replaced hydroxid and their 
formation weakens or destroys the concrete by forcing apart the 
particles of cement. In order for this action to occur it is necessary 
for the water containing alkali to percolate into or through the con- 
crete. Under field conditions 1 i ' these reactions referred to are much 
Fig. 1. — Briquets showing repellent action of oil-cement concrete on alkali water. (From tests by 
the Office of Public Roads, United States Department of Agriculture.) Top row contained 10 per 
cent of semiasphaltic oil; middle row contained 5 per cent of semiasphaltic oil; bottom row con- 
tained no oil. Briquets were immersed one year in a 10 per cent solution of sodium sulphate. 
retarded if not entirely suspended in most cases, due probably to 
the carbonization of the lime of the cement near the surface or the 
formation of an impervious skin or protective coating by saline 
deposits." 
It is also doubtless true that 2 "wetting and drying or freezing and 
thawing will hasten the destruction of the cement by extending the 
cracks already started." 
i U. S. Dept. Com., Bur. Standards Technol. Paper 12 (1912). 
2 Montana Sta. Bui. 81 (1910). 
