pressed on the street brings the stability up to 1070 pounds, which more nearly represents the stability of the mixture and is closer to the laboratory samples. This brings us to perhaps the most important place of the entire problem, upon which the authors next propose to concentrate, i. e., the correlation of the laboratory test with service. results. Until this is accomplished any deductions drawn from the test must of necessity be more or less theoretical in character. Many investigators hesitate to publish their data until an investigation is completed, with the result that much of value. is never revealed to the engineering profession for the reason that one line of research leads to another and it is seldom that an entire field is completely covered. The authors are frank to admit that much of the value of their test will depend upon whether or not it may be used to determine minimum stability values necessary to prevent displacement of a pavement under given climatic and traffic conditions, they, therefore, intend to next extend their work into the examination of samples of pavements which have actually shoved under as accurately known conditions as it is possible to find. In the meantime it is believed that the results already obtained are of sufficient interest to bring to the attention of those interested in asphalt pavement construction. DISCUSSION CHAIRMAN BLANCHARD: Gentlemen, both of these interesting papers on experimental investigation pertaining to asphalt mixtures are now before you for discussion. MR. SKIDMORE: Mr. Chairman, there is just one thing that I wish to again refer to in Mr. Hubbard's paper, and its relation to our own work. The 15 per cent hydrated lime which he has just been talking about, he got his highest results with a little bit less filler than gives the minimum of voids. I just wonder if possibly that might not be due to the fact that with the higher percentage of hydrated lime he did not get the mixture compressed quite as well. In other words, the increased amount of hydrated lime made a mixture which was more resistant to compression. So that his results with the 20 per cent mixture and 15 per cent mixture were not directly comparable because they were not in the same degree of denseness in the sample. And also the temperatures, possibly if he had increased the temperature with a higher percentage of filler he might have found his curve up a little higher for the 20 per cent value. MR. HUBBARD: I think there is a possibility of that being so. At the same time the results with hydrated lime that I showed you are merely illustrative. We have obtained exactly the same indications with limestone dust, both commercial and screened through a 200-mesh sieve. It may be that we are not using as high pressure in forming our test samples as we should. At the same time, we have worked up to the maximum initial compression that we feel safe in adopting, because we have found that if we further increase the pressure we begin to crush down the grains of the sand we are using, and this sand is one that is very widely used in the construction of asphalt pavements in the metropolitan area in and around New York City. BEHAVIOR OF CONCRETE PAVEMENTS UNDER SERVICE CONDITIONS By H. Eltinge Breed, Consulting Highway Engineer, Just 60 years ago the first concrete pavement of modern times was built in Inverness, Scotland; other concrete pavements built within that decade are still giving service in Edinburgh. The earliest recorded one in the United States is at Bellefontaine, Ohio, laid in 1892. (Illustration No. 1.) The following year other streets were laid, all of these are still in active service. From this time up until 1909, 770,022 square yards had been laid in streets. The graph illustrates the number of square yards of city streets laid each succeeding year since 1909. (Illustration No. 2.) Fourteen cities of over 100,000 population have 30 miles or more of concrete pavements 30 feet wide, while four of them have over 100 miles each, Los Angeles topping the list with over 300 miles. I am leaving out of this consideration 354, 433,187 square yards of roads outside of cities and 17,098,157 square yards of alleys. In all of this work some of the pavements, like that at Bellefontaine, Ohio, have been entirely satisfactory, some have not. Our problem is to ascertain what produces the satisfactory result so that we may build 100 cents worth of the pavement for the dollar. Conversely, what do we know about the causes of failure so that we may guard against them? Where the pavements are giving satisfactory results, we discover that attention has been given to certain fundamentals, the repetition of which would seem monotonous were it not proven necessary by the negligence one still sometimes sees. Foundations must have uniform density and a smooth even surface. Where they are of soil which makes a large volume change with the addition of moisture, they should be rolled very little; where subject to capillarity or generally unstable, they should be strengthened by a blanket coat of granular material such as cinders, slag screenings, coarse sand, gravel, etc. The Bureau of Public Roads, in their publication "Public Roads," have been reporting on researches that they have been making in regard to subsoil conditions, which are important for any engineers who are doing foundation work. Drainage from overhead water is generally assured in city work, for the roofs, walks, gutters and the pavement itself, are the umbrella which protects the whole. Underground waters must be cut off and taken away, especially on side hill work. The design of the pavement slab must be adequate. Mr. Older will discuss that in his paper. I have just one point I should like to make: The pavement can be constructed thinner near the curb and gutter on wide streets and where the sides are used for parking than would be safe for an area continuously subjected to rapidly moving loads. Selection of aggregates requires a knowledge of the properties of the sand and the stone. In concrete, greater appreciation of this has been indicated in the last ten years and we are securing better results. It is often possible where the local aggregates are not of suitable character for the surface to build in two courses simultaneously, using the poorer aggregates in the bottom and making the top surface of either cement and screenings or of cement sand and stone of better quality. An example of two course work is Chapel Street, New Haven (Illustration 3) This pavement was of the Blome Granitoid type built in 1908, over 3,000 feet long on the main thoroughfare from the East, and probably the heaviest traveled street in the city. The pavement is 634 inches thick, the first five inches being a 1-3-4 concrete covered by a top 134 inches, composed of one part cement to 11⁄2 parts of clean crushed trap rock |