Experience in concrete work has shown that its true place is in heavy foundations, retaining walls, and such like, and then perfectly independent of other material.  Arches, thin walls, and such like are very questionable structures in continuous concrete, and are on record rather as failures than otherwise.  This may to a certain degree be due to the high coefficient of expansion Portland cement concrete has by heat.  This was found by Cunningham to be 0.000005 of its bulk for one degree Fahrenheit.  It is a matter which any intelligent observer may remark, the invariable breakage of continuous concrete sidewalks, while those made in small sections remain good.  This may be traced to expansion and contraction by heat, together with friction on the lower side.

In foundations, according to the same authority above quoted, properly made Portland cement concrete may be trusted with a safe load of 25 tons per square foot.

In large masses concrete should be worked continuously, while in small masses it should be moulded in small sections, which should be independent of each other and simply form artificial stones.

The facility with which concrete can be used in founding under water renders it particularly suitable for subaqueous structures.  The method of dropping it from hopper barges in masses of 100 tons at a time, inclosed in a bag of coarse stuff, has been successfully employed by Dyce Cay and others.  This can be carried on till the concrete appears above water, when the ordinary method of boxing can be employed to complete the work.  This method was employed in the north pier breakwater at Aberdeen, the breakwater being founded on the sand, with a very broad base.  The advantage of bags is apparent in the leveling off of an uneven foundation.  In breakwater works on the Tay, in Scotland, where the writer was engaged, large blocks perforated vertically were employed.  These were constructed below high water mark, and an air tight cover placed over them.  They were lifted by pontoons as the tide rose, and conveyed to and deposited in place, the hollows being filled with air, serving to give buoyancy to the mass.  After placing in position the vertical hollows were filled with concrete, so binding the whole together—­they being placed vertically over each other.

As mentioned before, continuous stretches of concrete in small sections should be guarded against, owing to expansion by heat; but the fact of a few cracks appearing in heavy masses of concrete should not cause apprehension.  These occur from unequal settlement and other causes.  They should continue to be carefully grouted and faced until settlement is complete.