pretty well cut up with the telephone company's work, what you can see of the pavement is in very fine condition. There has been no marking along the curbs due to parking. It has a concrete foundation, and I think we counted two faint cracks clear across the street, which are undoubtedly base cracks. The mixture showed from 24 to 27 per cent of material passing the 200-mesh screen, with the ordinary sand grading which is considered proper for sheet asphalt work.

(Returns to manuscript.) With these dense, high filler mixtures, temperatures must be kept up so that the material will retain sufficient heat during the laying operations. This means that the mixture should reach the street with a temperature between 325 and 400 degrees F. It must be spread and rolled. as quickly as possible, and initial compression must be effected by means of heavy rollers. The 3-wheel type of roller is much to be preferred for the initial rolling. The mixture must be deposited far enough away from its final position to insure turning. Raking must be thorough and deep. Long-toothed rakes or forks for rapid combing ahead of the regular rakers have proved very satisfactory. All fine mixture asphalt pavements should be luted just behind the rakers, using the ordinary lute, something like those used for luting sand cushion for brick pavement, but using a longer handle so that it can be operated from the curb.

The same methods of design that are described here with respect to sheet asphalt mixtures are equally applicable to asphaltic concrete and stone-filled sheet mixtures. In order to show how accurately mixtures of this character may be designed, two instances during the present season are recalled when pavements for especial traffic conditions were so designed, and upon determining the formula to be used, it was found that as little as 3 or 4 pounds more of bitumen in a 2,000-lb. box affected the mixture on the street. One of these projects was a short length of Jackson Boulevard between Canal and Clinton Streets, adjacent to the new Chicago Union Station. As might be expected, this pavement was designed for very dense heavy traffic. A typical analysis of the top mixture is:

rollers, one of which was of the 3-wheel type, were used in laying the pavement. The coarse aggregate was Wisconsin granite and the filler local limestone dust, 90% of which passed a 200-mesh sieve.

Quite a number of investigators have been, and still are, giving considerable time to research work with asphalt paving mixtures. Some of this work has proceeded along the line of search for fundamental principles, while others have sought to develop empirical tests that will rate mixtures as to stability, and still others have combined the two. Such tests will fall short of the real answer if they neglect basic knowledge of the mechanics of the mixture. Nevertheless, any serious work of this nature is not be be discouraged, since at least it must lead to a better understanding of the problem. A curious thing about most of the tests so far made public is the fact that the authors stress shear strength as the most important factor, yet some of the tests described either fail to develop shear, or combine shear with so many other forces that the results are difficult of interpretation or wholly misleading.

Compressive strength of mixture is of no particular importance after the pavement is laid. Tensil strength, and more especially shear strength, are highly important, since the traffic forces which develop distortion, place the mixture structure in tension and in shear. Tests that will correctly measure these stresses direct and free of compression, viscosity and friction against mold forms, will certainly give us much more definite knowledge of our product than tests which develop intricately combined stresses. In the investigation work which has been in progress in our own laboratory for the past few years, we have used both tensil and shear tests, measuring only these forces. Some very interesting data along these lines is still in the process of making.

One of the most interesting investigations that has been recently reported is by Captain S. B. Moore, and Father F. A. Tondorf, Seismologist of Georgetown University, in "The Military Engineer" for March and April, 1925. This work meas

ured the intensity of vibration in surfaced and unsurfaced P. C. concrete slabs under moving loads. It is to be hoped that these investigations will be given wider application.

Some facts are so self-evident that they require no investigation to demonstrate them. For example, it is useless to attempt to stabilize mixtures by the use of a bitumen a few points lower in penetration than is common in practice. About the only material effect this will have on the mixture is to increase its stiffness during the laying operations somewhat. Such distortion as does occasionally appear in asphalt pavements, occurs during hot weather, and is due primarily to an excess of bitumen; the consistency of the bitumen has almost nothing to do with this phenomenon. Tests which endeavor to measure resistance of mixture to distortion must then deal with summer pavement temperatures, say, 140° to 150° F. There will be no need to waste time investigating mixtures which contain more bitumen than they have room for at this temperature as they are immediately doomed to failure. Tests dealing with old standard mixtures are of no value unless the scope of the investigation is extended to include later practice so that a comparison of values may be made.

Asphalt pavements as a general proposition are by no means a failure. Such failures as do occur now and then represent only a minute fraction of the total yardage of pavement. But, it is the failures rather than the successes that quickly and repeatedly come to our notice. There is a continuous race between traffic and pavement design, so that we must be always on the alert for improving our design, and before we can do this intelligently, we must have a thorough understanding of the fundamentals. Therefore, these various investigators should be given all the encouragement possible so that we may be able to reduce asphalt mixture design to the exact science that it really is. (End of paper.)

I would just like to talk very briefly on one or two things covered by certain curves in pamphlets which you will find on the chairs. I have included a chart which will be found on the second page of the reprint, which has to do with the effect of various fillers upon a given sand in reducing voids. Our work was started because our first results which made this thing quite interesting to us were obtained with high filler mixtures.

We started out to determine the value of various fillers that were on the market in the way of reducing the voids in the mixture. And that chart shows a number of materials, including Portland Cement, limestone dust, silica dust, slate dust, hydrated lime and Wilkinite, a semi-colloidal clay obtained in the Wyoming district. We found that so far as total reduction of voids is concerned ordinary limestone dust, about 85 per cent of which passed the 200-mesh screen, gave us the best results; that is, it brought the voids down to a lower point. Hydated lime, however, on account of its extreme fineness, and also difference in volume weight, reduced the voids much more rapidly, and then it quickly breaks, also, the curve starts up very sharply. Portland cement is almost as good a filler as far as void elimination is concerned; and silica dust is between the two. For the lower percentages silica dust reduces the voids more rapidly. I think that might be due possibly to the fact that the silica was a little finer, but not so well graded, had higher percentages passing the 200-mesh screen, but in the final apportioning of the filler it was not so well graded as the limestone dust. And after all the voidage of the filler itself and its grading have a very material effect upon its reduction of voids in sand.

One of the men in our laboratory, on investigation work, had a rather happy thought last winter-or rather two years ago, as to how to show very briefly and very effectively the untenability of the surface-area theory of mixtures. His thought was to take Richardson's ideal heavy traffic mixture, make pat stains at a certain temperature-all this work being done under very careful control-and then to note the change, if any, in the mixture by increasing the filler content, screening out everything except the 200-mesh and simply increase the 200-mesh by definite percentage each time to see what effect it had upon the pat stains. This photograph here-there is a smaller reproduction of it on the last page of the pamphlet, not so clear, possibly, and I don't know whether this can be seen very well from the back or not. These pat stains were all made at 300 degrees F., all with the same pressure on the pat, the usual way. We found the only practical way to make pat stains is for a man to throw his whole weight on a board-usually a 2x4, for instance-and rock back and forth a certain definite num

ber of times. We used ten back and forth pressures.

They were all made in that way by the same man, all at the same temperature. The first is Richardson's ideal heavy traffic mixture, containing 10.5 per cent bitumen, and regular heavy traffic sand grading so as to exactly comply with this reqiurement, and 13 per cent of filler passing the 200-mesh screen. The next was identically the same mixture, except with the filler raised to 16 per cent. In the next one the filler was raised to 19 per cent-three per cent more. Then the last one, pat stain No. 4, it was raised to 22 per cent, and we could hardly get the paper off of the sample. It didn't leave a stain; it just left a solid smear of asphalt. The next ones after that-we could not make any more because it was just mushy and would stand no pressure. It effectively shows that the addition of that small amount of filler each time was crowding the asphalt out of the voids. In the first instance we got a comparatively faint stain, or what we call ordinarily a pretty good stain. The bitumen was approximating the voidage. It was a little bit under the correct amount, this one here (indicating). The second stain is probably about all that the mixture would carry. (Applause.)

MR. SKIDMORE: Mr. Chairman, I just had another thought. Mr. Hubbard has a most interesting paper that deals very much with the same subject, and I wonder if we can not delay the discussion until after Mr. Hubbard's paper is read, and then we can discuss both papers.

CHAIRMAN BLANCHARD: We certainly can, especially since you make the suggestion, sir.

MR. SKIDMORE: It deals with almost the same subject, and Mr. Hubbard says that is agreeable to him.

CHAIRMAN BLANCHARD: All right. The next paper is entitled "Researches on Asphalt Paving Mixtures," and is a joint paper by Prevost Hubbard and F. C. Field. Mr. Hubbard is Chemical Engineer, and Mr. Field is Chemist of The Asphalt Association. It is understood that Mr. Hubbard will present the paper. Mr. Hubbard. (Applause.)

MR. PREVOST HUBBARD: Mr. Chairman and Gentlemen: Mr. Skidmore has already introduced to you the subject of stability of asphalt pavements, so that I will dispense with a good many preliminary remarks that I might otherwise have made. I want