Surface construc tion is cheaper and seems to be much more common than trench construction. Surface construction is the better, since the depth of the gravel at different distances from the center is approx imately proportional to the amount of traffic; while in the trench construction, if the graveled portion is wide the sides are liable not to be much used, and if the graveled portion is narrow passing vehicles are forced upon the earth shoulders. Therefore it appears that surface construction is best for roads having a large amount of traffic. In park drives and streets, the whole width of the road way is excavated and filled with gravel.
Trench construction is a little more economical of gravel, and is therefore most suitable where gravel is expensive.
It is sometimes advocated that there should be two tracks, an earth road for sum mer travel and a graveled way for winter use. This plan has some advantages and also some disadvantages. When the earth track is dry, it is preferred by most teamsters to the hard gravel road; and the use of the earth roadway decreases the wear on the gravel, —which is clearly an advantage, for a gravel road like most other things will wear out. On the other hand, if the summer track is immediately adjacent to the hardened way, the earth of the former will become mixed with the gravel of the latter, much to the detriment of the gravel. The chief source of expense in the maintenance of gravel roads is due to the damage done by the mixing of earth from the side of the road with the gravel, thus forming a mixture that will hold water and cause the road to cut up. It has been suggested that the objection to the two tracks could be obviated by constructing a ditch, or sodding a narrow space between the two; but this is impracticable. The two tracks require a wider right of way, and therefore for this reason are fre quently impossible.
The gravel usually contains many stones too large to be used in or near the wearing surface, and therefore it is economy to screen the material and lay the larger pebbles in the bottom. Some writers object to using pebbles larger than 1 or 11 inches in diameter for the bottom course, on the ground that the heaving effect of frost and the vibration due to the pass ing wheels will cause the larger stones to rise to the surface and the smaller ones to descend—" like the materials in a shaken sieve."
Unquestionably, if a vehicle is driven over a layer of loose stones of all sizes, the larger ones will tilt up when the weight comes upon them and the smaller ones will roll down into the space made va cant by such tipping; and by a repetition of this process, the large stones will gradually reach the surface. The heaving action of the frost acts in a similar way. But it does not therefore follow that a layer of coarse stones at the bottom of a gravel road will thus work to the top when the interstices of the gravel above are filled with binding material and all is compacted by traffic or by rolling. Experience has shown that if 2 to 4 inches of the top dressing has suitable binding material, it is extremely improbable that pebbles 2 to 23- inches in diameter in the bottom course will ever work to the surface.
Other materials than coarse pebbles may be used for the lower course. In many localities there are large quantities of coal slack, which is useless as fuel and is too friable for the wearing surface of a road, but which can be used for the bottom course of a gravel road. Coal slack has thus been successfully employed, and is often cheaper than gravel. Blast-furnace slag has also been used for this purpose. Sometimes broken stone is used for a bot tom course; but on account of the expense of breaking, only a stone found already broken in the quarry is suitable for this pur pose. A " flake " stone or quarry chips are the forms generally used. The celebrated gravel roads of Central Park, New York City, have a " rubble foundation "—not a Telford foundation (§ 302). The rubble layer is 10 to 12 inches thick, and the gravel 4 to 6 inches after being thoroughly compacted. The stones, none of which exceeded 9 inches in greatest dimensions, were dumped upon the subgrade from carts and "evenly adjusted by a little labor of the hand." *