having a maximum size of 1/4 inch with dust removed. As shown by the picture, the original markings are still in evidence after 17 years, and the pavement is entirely serviceable, which is remarkable considering the large amount of steel tired coal trucks that have passed over it.

The proper design of the concrete requires an appreciation of the inter-relation between the size and grading of aggregates, the consistency of the concrete, the mix, and the strength. When the aggregates are clean and of durable character, the strength of the concrete depends on the water ratio.

Time of mixing must be at least one minute, and one and one-half minutes gives increased strength.

Finishing is a most important factor, for a smooth pavement has to resist only a moving load, while a pavement which is uneven has to endure also the highly destructive force of impact. Proper finishing includes cutting through and rounding all joint spaces. (Illustrations 4 and 5.)

Curing with all of its protective measures must be adhered to. Integrity of the slab as a monolith is necessary. The first tendency of concrete after being laid is to dry out and contract. This change of length always produces cracks unless the slab is free to move. Its own weight ties it down to the subgrade and the resultant friction precludes its drawing itself together. When the stress thus generated is greater than the strength of the concrete, the concrete ruptures and cracks appear. By keeping the surface of the pavement continuously wet the tendency to contraction is lessened and the slab retains its integrity. The gain in strength may be accelerated by the use of a 2 per cent solution of calcium chloride.

Complete specifications are needed to insure good results. It is far better to use a standard specification in its entirety. than to concoct one with the paste pot and shears. The specification to be submitted at this meeting of your own society will, when intelligently used, give consistently good results.

Some unsatisfactory results may be attributed to the following causes (1) Cold weather work has been followed by a string of bad results. There is no gain in strength when the air temperature is around 40 degrees Fahrenheit; there is slight gain at 50 degrees, while real strength comes at around 70 degrees. Fahrenheit. Concrete laid at 40 degrees or below will not gain

strength until the temperature rises, so if it goes over winter and traffic is allowed on it before the gain comes, the concrete

will become shattered. Where cold weather work has to be done, it can best be accomplished by heating the aggregates by using calcium chloride and by protecting from weather. Or good concrete may be obtained in cold weather conditions by using

Aluminate cements. In curing, sufficient time should be allowed for gain of strength. Better concrete can be secured in summer with seven days curing than can be had in November with 21 days curing.

Another source of trouble is the use of too much water in the concrete. One case that I investigated showed a secondary concrete of little strength directly under the slab proper, while the concrete in the slab was porous and absorbent of water, inevitably to be disrupted by expansion from moisture and frost.

Concrete around switch points where it is subject to freezing and thawing from the use of salt by the traction companies shows a marked deterioration over the rest of the pavement. This condition was very evident around the switch points of the trolley track in Homer, New York, while the rest of the pavement was in perfect condition. Vitrified brick, granite sets or granite blocks around the switch points will give better life than concrete will.. Where the upper part of the grout comes out the space can be protected by pouring the joints with a bituminous material.

Another cause of poor results is the use of dirty and unsuitable aggregates. In a village in Northern New Jersey, where a fine quality of trap rock was available, sand was specified that should be equivalent to 80 per cent of Ottawa with the proviso that it could be used with an increase of cement if it went as low as 60 per cent of Ottawa. Needless to say, a few months after the job was completed there were sections where one could kick out the coarse aggregate. It is dangerous practice to try to use a sand that is under 100 per cent of the strength of Ottawa sand unless you are designing the mix on the strength basis, fineness modulus, water ratio, etc., and have enough control to know that any change in aggregates is taken care of by redesigning the mix.

Coated stone or gravel is another source of trouble. Many jobs show distress from this cause, for when the aggregate is dirty the mortar forms a honeycomb structure around the individual pieces, which is weakly resistant of stress. Many an area can be found where the pavement is breaking up from this cause. One-half mile section on the Merrick Road, near Oak

dale, Long Island, New York, built in 1914, is a notable example of this.

Poor joint construction or no joints at all is one of the greatest sources of trouble both early and small, and large and late. Concrete, by its very nature, has the characteristic of extensibility, increasing and decreasing in length. Dr. W. K. Hatt,

Purdue University; Dr. A. T. Goldbeck, Chief of Tests, Bureau of Public Roads; Professor Alfred H. White, University of Michigan, and other scientists, have shown definitely through their researches that there is expansion (increase in length) in concrete pavements. Some studies have shown that the tendency of concrete is to hold some of each increase in length over subsequent periods of contraction, so that part of the expansion is cumulative. When no expansion joints are provided very high stress is set up; the concrete either expands laterally up to its elastic limit or ruptures along its weakest parts. The rupture may not appear the first year, but frequently shows. the third or fourth. After some experiences of this kind, which showed first in blow-ups, our construction was changed to allow more expansion and an extra joint 111⁄2 inches wide was provided every 750 feet. Work done in 1917 with this allowance is excellent today, while pavements built a year or two earlier with less expansion have shown a decided tendency to crack.

Other sources of trouble with joints are: (1) the submerged joint which on compression spalls out on top; (2) the joint which is not perpendicular to the pavement surface and which when pressure comes rides above the adjoining slab; (3) the