112 
IOWA ACADEMY OP SCIENCE 
But the microscope shows clearly that the theory is true, at least in part. 
The theory of crystallization states practically that each tiny particle of Port- 
land Cement is incased in a thin envelope of water and passes into solution in 
part. Then it crystallizes to its neighboring particle or to the grains of grog 
or sand. 
If this theory be correct then the obvious explanation of the phenomena of 
reaching maximum strength and later decreasing in strength can be explained 
at least in part, by the slow completion of this process of crystallization; dur- 
ing this process there being a steady loss of a small part of the cement, removed 
in solution by percolating waters, this process continuing steadily after the 
process of crystallization is completed. 
In continuation of our work we sent pieces of well cured cement construction 
to Dr. Knight at Cornell College. These were treated with distilled water and 
a considerable percentage of solids was dissolved. Analysis of these solids re- 
vealed the presence of silica, calcium and aluminum in about their ratio in Port- 
land Cement. This strongly indicated that we had secured solution of the Port- 
land Cement as a whole and that we were not securing merely free lime. By 
these experiments and others conducted at the University of Illinois we justified 
ourselves in the belief that a very appreciable part of Portland Cement is 
soluble in pure water. Before the Clay Workers of the state we presented our 
conclusions that the thin wall construction of Portland Cement drain tile, being 
very porous and subject to continuous action of water, would be short lived and 
to all practical intents and purposes would entirely dissolve away. 
Since that time we have conducted further experiments at Cornell College 
and the University of Illinois, the results of which we wish to give you now. 
Mr. L. R. Ernest, University of Illinois, finds that one gallon of water will 
dissolve about one gram of the material of a Portland Cement drain tile. (We 
use Portland Cement drain tile and sewer pipe in our investigation, as in these 
uses the cement is subjected to very severe conditions and disintegration 
will work most injury, as it is out of sight and would not be observed until the 
damage done compelled investigation.) Mr. Ernest further treated a weighed 
amount of the material with distilled water and the alkalinity was titrated 
with standard acid. The per cent of dissolved lime after one day was 0.9 per 
cent. After two weeks 1.9 per cent. After one month 2.1 per cent. It would 
appear from this that the saturate solution would be reached somewhere be- 
tween two and three per cent. 
Mr. C. J. Derrick of the University of Illinois has done work practically 
duplicating this and the additional work with the Soxhlet extractor. A small 
piece of tile was broken in two pieces, which would fit in a large size extractor. 
The dust v/as carefully removed from the two pieces in order that only a piece of 
the undisturbed tile would be acted upon. The idea was to duplicate the condi- 
tions of nature as far as possible where the water comes in contact with the un- 
broken tile. The Soxhlet extractor was employed in order that pure, fresh 
water would constantly wash the tile. Of course the difference here was that 
the water was pure, being condensed from steam, while it is not pure in nature, 
but contains organic acids. The tile was subjected to the action of water in this 
way for 72 hours with the following results: Weight of piece No. 1, 24.9185 
grams; weight of material extracted, 0.5097; weight of piece No. 2, 18.2227; 
weight of material extracted, 0.4281; per cent of material extracted from No. 1 
was 2.08 per cent; from No. 2 was 2.34 per cent. This tile being one-fifth Port- 
