This is one of the most important of physical tests, for it indicates whether there probably exist elements which will eventually disintegrate the mass. The action of pats should be carefully observed; and if warping, cracking, or blotches appear, the material should not be used. Form on small pieces of clean glass four pats of cement mixed with only so much water as will turn the color of the cement. The pats are to be about 1/2-inch thick in the center, and 3 inches in diameter, tapering to the edge, which will be as thin as possible. After a day under a damp cloth, put two of the pats in water for a week, and leave two in air, and examine for symptoms referred to above.

Should it be possible to test the material chemically, the following points should be observed:

The lime should be slightly over 3 times the silica. Thus, if there is 20 per cent silica, the lime should run from 60 to 62 per cent; this is a good balance. The silica should not fall below 19 per cent, nor rise above 22 per cent, which is about the range of good cements on the market.

More than 2 per cent of sulphur should never be allowed. The reason for this is that the "setting" of cement is a process in which the various materials combined unite chemically to form one solid mass. Sulphur in its various forms is an ingredient which does not work in harmony with the other materials; and while it does not at once prevent the satisfactory combination and setting of the other materials, yet, under conditions that are very likely to prevail, it sets up a chemical action tending to disintegrate the mass in time.

Magnesia can generally be considered an ingredient which has no value either as sand or cement, being what is known as "inert" material. While 4 per cent may not injure masonry, wearing surfaces should not have over 3 per cent.

The manufacturers of Portland cement have tried to eliminate both of these ingredients - but so far without success - at a cost which would make it possible to sell the cement at a profit.

The specific gravity should not fall below 3.10, as a low specific gravity indicates poor burning of the clinker. These later tests, however, are delicate laboratory operations, and only experienced chemists are capable of getting satisfactory results.