The above method of finding the voids is subject to a consider able error, since the water will not expel the air from the sand; and therefore it is better to consider the above only preliminary, and proceed as follows: The sand being dry, fill the pail with the mix ture, tamping it as it is put in; and then empty the pail upon a table, taking care not to lose any of the material. Next put into the pail a little less water than is required to fill the voids, and then grad ually pour in the mixture previously emptied upon the table, tamp ing it as it is added. If necessary to keep the mixture wet, add a little more water as the tamping proceeds; and when the material is all in, add water until the pail is level full. Then the total weight of water in the mixture divided by the weight of water re quired to fill the empty pail is the per cent of voids in the rammed mixture.
The amount of limestone dust ordinarily used varies from 5 to 15 per cent of the sand (see § 625); and the fineness is usually such that all will pass a sieve having 30 meshes per linear inch and at least 75 per cent will pass a sieve having 100 meshes per linear inch.
Table 42, page 413, shows the average fineness of the sand and limestone dust used in the asphalt pavements laid in Washington City by the two paving contractors doing work there, during the year ending June 30, 1897, and also the fineness recommended by Hr. A. W. Dow, Inspector of Asphalt. When available, other asphalt paving contractors use sand and limestone dust of substan tially the fineness stated in Table 42.
To illustrate the relation between the fineness of the min eral matter and the softness of the asphaltic cement, Mr. Dow cites It will be noticed that the sand used in 1894 was considerably coarser than that employed in 1897. The asphaltic cement used in 1894 was 20° of penetration (Bowen apparatus) harder than that employed in 1897; but nevertheless the 1894 pavements were as soft, and in some cases softer, than the 1897 pavements, as indicated by the tracks left by the traffic.
The Neuchatel rock asphalt 595) is much used in Europe, and has been found to make a hard and durable pavement. This so calkd rock asphalt is a limestone powder cemented together by an asphalt "so soft that its flow is perceptible at a temperature of 75° F., which is about three times softer than any asphaltic cement used in the Washington pavements." All asphaltic cements grow harder with age, and are likely to crack in cold weather; and therefore it is desirable to use as soft a cement as possible, a condition which seems to indicate that the sand should be fine. A fine smooth sand requires a considerably harder cement than a sharp angular sand. If the cement is soft and there is a large per cent of voids in the sand, there is a liability that in hot weather the sand will work to the bottom and the asphalt to the top, where it will chip and scale off during cold weather.