May 12, 1923 
Effect of Decomposition on Concrete Tile 
475 
More pits were opened, and it was discovered that the entire lines 
were badly disintegrated, decay being greatest at the upper end and 
farthest away from the pump house. It was at first thought that the 
tile might have broken down, owing to bad workmanship, but this was 
disproved when it was found that at the extreme upper end of the tile 
lines, where rotted muck had flowed in and partially filled the tile, the 
tile had been protected, and had not decayed. It still bore the marks 
made by the mold, and, upon breaking, the interior of the tile wall appeared 
to be in perfect condition. The surface of the tile, both inside and outside, 
was rough, due to the grains of sand and gravel falling away. Inside the 
disintegrating layer, the tile was discolored for about of an inch, 
but appeared to be dense, while a tliin layer in the center of the 
wall was unaltered. This last fact shows that the tile had not broken 
down due to the passage of waters through the walls, but to solvent 
action upon the walls themselves. Disintegration was greatest on the 
outside of the tile and at the ends and much greater on fiie top than on 
the bottom. Disintegration on the inside of the tile was greater at the 
top than on the bottom, and was not greater at the water line, except 
at the joints, where it met heavy disintegration from the outside. 
Since the tile showed signs of good workmanship, it was assumed that 
the disintegration must have been due to the solvent action of some 
substance present in the soil. 
Correspondence was begun and the literature searched to secure 
information which would throw some light on the reason for the dis¬ 
integration. The most probable solution of the difficulty was first pre¬ 
sented by Prof. H. B. Roe, of the University of Minnesota, who cited the 
destruction of concrete tile in his State by the action of sulphate salts. 
This possibility was further investigated and talked over with chemists. 
The conclusion tentatively arrived at was that the disintegration was 
caused by sulphates produced by the decomposition of proteins in the 
peat. This solution seemed satisfactory. 
Work on the marsh was discontinued at this point because of cold 
weather. The compilation of data could not go ^ead until more field 
work was done and opportunity was taken to test out the assumed 
presence of sulphates in the marsh and their effect on new tile. 
PART II. PRELIMINARY STUDIES 
LABORATORY INVESTIGATIONS 
With the advice and assistance of Prof. E. Truog and E. J. Graul of the 
Soils Department, laboratory studies were made, beginning in February, 
1920. A cement tile company furnished new, uncontaminated tile for 
the tests. These tile were machine made, very dense, steam cured, and 
three months old at the time of the tests. 
MARSH WATERS 
No sulphur in the form of sulphates could be detected in water from the 
marsh by the addition of barium chlorid to a hot solution slightly acid 
with hydrochloric acid. The assumption that the disintegration was due 
to the presence of sulphate salts was therefore unfounded, and the whole 
cause of the destruction still remained obscure. 
The next step was to find actually what was in the marsh waters and if 
those marsh waters would dissolve the tile. Iron and aluminum were 
