On the other hand, there must not be too much clay, or, after the lime is turned into silicate, there will be a surplus of free clay left - -having in itself no hardening property, and which will decrease the strength of the resulting cement.

The proportions required to produce a cement vary, however, within tolerably wide limits (22 to 36 per cent in the raw material); as a general rule, the quicker-setting cements are produced from stones containing most clay.

Composition Of Clay

Those clays which contain a large proportion of iron and alumina cause the lime in which they occur to set with greater rapidity than do ordinary clays. "Clays deficient in iron and alumina, and in which the silica is present in the shape of finely divided quartz, are apt to form insoluble silicates of lime at a high temperature owing to the want of suitable bases to combine with the silica, which also renders them unfit for making Portland cement. Such clays also when calcined at too low a temperature yield hardly any soluble silicates, form therefore no protection to, and seem in no way to prevent the hydration of the lime, and produce a material devoid of hydraulic properties."1

The following Table sums up approximately, and in a concise form, the

Proportion of

Clay before burning.

Composition of Clay.

Degree of Calcination.

Setting Properties.

Examples of the Class.

0 to 8 p. c.

...

Very low.

Absorb carbonic acid from air.

Fat limes.

8 to 18 p. c.

Various. Those with most iron and alumina set most quickly

Moderate.

Moderately quick setting. No great strength.

Lias and other hydraulic limes

20 to 30 p. c.

Iron and alumina. Silicic acid.

Very high.

Sets slowly. Very strong.

Portland cement.

Do.

Do.

Extreme.

Become inert.

Do. over-burnt.

28 to 55 p. c.

Iron and alumina. Silicic acid.

Low.

Sets very quickly. No great strength.

Roman cement and others of that class (see p. 158).

Do.

Do.

High.

Become inert.

Do. over-burnt.

mutual relations of the proportion of clay, composition of clay, degree of calcination, and setting properties in different classes of limes and cements. We have considered the effect of clay in conferring hydraulic properties upon limes, which it does by presenting silica in a state fit for combination, but clay is not the only substance which has this effect.

1 Scott and Redgrave; M.I.C.E., vol. lxii. p. 80.

Pozzuolana, etc. - As before noticed, the presence of several other forms of soluble silica and pozzuolana will also answer the purpose in a greater or less degree.

The general nature of the reactions that take place in the setting of limes containing these substances are much the same, and produce effects similar to those already described.

It has been recommended that mortar made with substances of this kind should be allowed to remain in paste for some time before use. The reason for this is that in consequence of the clay and lime not having been burnt together, none of the silicates have been formed, as they are in ordinary hydraulic limes burnt in the kiln. Every facility should therefore be given to the silica to attack the lime through the intervention of the water (see p. 221), and thus to form silicates, before the mortar is used.

Carbonate of Magnesia, - Carbonate of magnesia is a substance very similar to carbonate of lime; it loses its carbonic acid in burning, combines with silica, etc., and behaves generally in the same way, with one important exception, viz. that the calcined magnesia will not slake on the addition of water, but combines with it gradually and quietly, and sets to some extent in doing so. When silica is present it combines with the magnesia, and with the lime, forming a double silicate of lime and magnesia, which is of greater strength than either silicate of lime or silicate of magnesia separately.

Besides this, the magnesia and lime, even without the intervention of the silica, will combine and harden under water.

The hydraulic mortar that is produced from magnesian limestones and dolomites (see p. 59) owes its properties to the different combinations above mentioned.

Several failures that have recently occurred in Portland cement after use have been attributed to an excess of magnesia in the cement.

As this has caused considerable mistrust of the material, the following remarks on the subject by Mr. Dent will be valuable : -

"When the lime is associated with magnesia, the magnesia should be regarded as to some extent taking the place of the lime, and the quantity of the lime should be proportionately diminished.

"A well prepared Portland cement, such as is made on the Thames or the Medway, should not contain any appreciable quantity of magnesia, say about 1 per cent. Although any large proportion of magnesia in Portland cement cannot be considered desirable, yet it must not be forgotten that magnesia is capable of forming hydrates of great permanence and hardness, and that some very good hydraulic cements contain as much as 8 per cent of magnesia, such, for example, as the well known Rosendale cement of the United States of America.

"There can be little doubt but that the assertions that have been frequently made as regards the tendency of cements containing magnesia to disintegrate, may sometimes have arisen from overlooking the fact that the results observed might be due to excess of basic constituents in the cement. In a recent statement put forward as to the injurious action of magnesia, the cement referred to contained 72 per cent of lime and magnesia, and it could scarcely be regarded as extraordinary that such a cement should prove a complete failure since it is well known that such a proportion as 72 per cent of lime would render Portland cement so unsafe as to cause it to be condemned." 1

It should be mentioned, moreover, that in the cases of failure which have occurred the magnesia has been found by analysis after the cement has been for some time under seawater, and that it may not have been in the cement when originally deposited, but introduced by the chemical action of the sea-water upon the uncombined lime existing in the cement. If this is so, the best safeguard against the evil would be extreme care to avoid overlimed cements, and to cool and aerate the cements thoroughly before use (see p. 176). With regard to the action of sea-water upon cement, Mr. Dent says : - "From a recent report of Professor Brazier on the cause of the failure of some cement used in the construction of a graving dock in Aberdeen harbour, it would appear that the reaction which takes place between the magnesium chloride contained in sea-water and lime may, under certain conditions, be sufficient to cause the disintegration of some descriptions of Portland cement, the lime in the cement being dissolved."

1 Dent's Cantor Lectures, 1887.

Sulphates

Lime can also be made to unite with water by the presence of a small quantity of any sulphates, and the employment of this property by a suitable process will considerably increase its setting power.

The setting of lime thus treated is essentially distinct from that produced by combination with silica, inasmuch as it depends on combination with water only (which becomes solid) and the resulting substance (which is simply hydrate of lime) is entirely soluble in water, though with more difficulty than the ordinary slake lime, owing to its superior density.

Sulphate of magnesial (commonly known as Epsom salts) is very soluble in water.