The principal test for soundness is that of boiling, and it may be carried out with quite elementary appliances. A circular pat of cement is made upon glass, slate, or some other smooth substance, about 3 inches diameter, and 1/2 inch thick in middle, tapering out to the edges. If made in the evening this could be ready by morning for further treatment, care being taken to keep the temperature of the place in which it is deposited well above freezing-point, and preferably as near as possible to 60°. It is convenient to make such a pat upon what is known as a lantern-slide cover-glass, this being 3 1/4 by 3 1/4 inches In the morning, by which time the pat has set, it should be put into a small saucepan or kettle just as it is, glass and all, covered with water, and brought to the boil over a fire or oil stove; and it should be kept boiling for about six hours, fresh water being added as that in which it is immersed evaporates. An unsound cement will not stand this test at all, but will turn to soup in the boiling process, while a good cement will remain adherent to the glass and perfectly sound in all respects. An overcool cement will probably shrink, curling up at the edges and becoming detached from the glass, while a slightly hot cement may show hair cracks on the pat without absolutely disintegrating. For most purposes any cement may be accepted which comes out of the test as a homogeneous pat, though judgment must, of course, be exercised. The effect of the boiling is to accelerate the changes which take place in use, and of these expansion is that which is most to be dreaded. It is undue expansion which breaks up an unsound pat. During boiling this has taken place in the course of a very few hours, but under ordinary circumstances in work the same thing might not occur for several months, and then result in some considerable misfortune, such as the blowing off of the collars of a series of drain-pipes.

This test, like the others mentioned, can be more scientifically applied if desired, by means of an apparatus somewhat resembling an incubator, but, generally speaking, the rough test described above will suffice.

Sand is not a difficult material to test; it is usually specified to be sharp pit or river sand, free from loam. This implies that there should be no softness or smoothness in its feel when rubbed between the hands, and that, if a small quantity be stirred in a tumblerful of water, the water should remain clear and unclouded; but builders are particularly fond of using a loamy sand if they can get it passed, as it is generally cheaper and easier to work up. River sand is sometimes rounded in its grit, but in spite of this may generally be passed. Pit sand, on the other hand, may be loamy, and may even be salt; this can be detected by tasting a little in the mouth. If only loamy sand be obtainable in any given district, and it is specified as above, it is necessary to have it washed, and it is the Clerk of Works' duty to see that this is done thoroughly by placing it in a shallow trough with water passing over it in a continuous stream, while it is turned over occasionally. A few hours' washing will generally remove all trace of loam and even of salt, and render the sand suitable for use. Even sea-sand may be utilised with safety if well treated in this way; but, of course, watchfulness is necessary to ensure that the washing is sufficient. The objection to the presence of salt is that it attracts moisture, and so is likely to produce a damp building.

Gravel and other aggregates for concrete are, in respect of loam and salt, similar to sand. Everything in connection with these, however, depends upon the particular specification. Sometimes a gravel or ballast is specified which is expected to contain a certain proportion of small stuff or sand, while in other cases it is intended that it should be screened, and only the coarser particles used, the sand being afterwards added in definite proportion, with the result that occasionally the sand which is sifted out from the aggregate is again returned to it to make the proportions right. If sifting is specified, however, it ought, as a matter of course, to be enforced.

Lime is not often tested, but judged by inspection; but if there is any suspicion that a pure lime is being supplied when a moderately hydraulic lime is intended, detection is easy on the application of water, as a pure lime slakes much more rapidly than does the other. Possibly the greatest trouble is experienced with lias lime. If the maker be not definitely stated it is possible, and legally permissible, for something else than a lime from the lias formation to be supplied; but the Clerk of Works should insist, so far as he is able, upon having the real article from one of the well-known Somersetshire or Warwickshire burners.

The presence of impurities, such as sand or colouring matter, if not detected by inspection or by the sense of touch, may be ascertained readily by placing a small quantity of the lime in a saucer and covering it with hydrochloric acid. Effervescence will result, and when this subsides it will be found that the lime has been "digested," any residue left in the saucer being of some foreign substance.

This is also the test to apply to mixed mortar, if there be any doubt as to its proportions. It must be allowed to dry before the test is applied, and weighed before and after the test, the proportion by weight of the residue showing the proportion of sand that has been used.

Stones are generally specified as being free from sandholes, vents, and all other defects. It is rarely that any test is stated in a specification; but if so it is usually one for absorption, either by percentage or by weight of water taken up in a given time by a submerged piece of stone of a given size. This is an easy test to apply, proportion being generally ascertained by weight, needing that the stone shall be weighed before and after the test, and the excess of weight due to the presence of water over that of the dry stone indicating the amount which has been taken up. If the test specified be one of bulk it is necessary to cover the stone with a definite measured amount of water, and afterwards to measure that which is left, the difference, of course, representing the amount absorbed. The only essential precaution to take is that the test be carried out in a vessel which itself does not absorb moisture. A dry wooden tub, for instance, would not be permissible, while an iron bucket might be trusted.