56. The amount of money wasted in continually opening up the streets, digging, bracing, and refilling, is a considerable item, not counting the interference with travel and business, and would be sufficient to cover the interest on the cost of the subways. The waste, being distributed through many companies, is not sufficiently felt to cause a reconstruction. The streets of New York were opened 27,088 times in 1890 by the gas, steam-supply, and other companies.

57. Under the best municipal administrations of Europe neither corporations nor individuals are permitted to disturb the pavements. All removals and restorations are done by the city's own employés, upon the deposit, by the parties who require the streets to be opened, of a sufficient sum to cover the expense for each piece of paving done, at a fixed price per yard according to the kind of pavement.

Moreover, interference with the pavements is of rare occurrence, for the companies having pipes underground are required to thoroughly examine and reinstate their mains and services concurrently with the paving of a street, of the execution of which due notice is given them by the city. Such regulations are quite practical, and there can be no hardship in requiring American companies to pay for like work.

It is stated that in Victoria Street, one of London's busiest thoroughfares, not a single stone has been disturbed from the carriage-way in twenty-five years. This street, as well as many others, has a subway in which are contained the gas and water pipes and upwards of six conduits for telegraph and electric wires.

CHAPTER II.

MATERIALS EMPLOYED IN THE CONSTRUCTION OF

PAVEMENTS.

58. Selection of Paving Material.-The materials most commonly used for pavements are stone in the form of blocks and broken fragments: wood in the form of blocks and plank, asphalt in two forms,-sheet and block, and clay in the form of brick.

59. In considering the relative fitness of the various materials, the following physical and chemical qualities must be sought for: (1) Hardness, or that disposition of a solid which renders it difficult to displace its parts among themselves.

(2) Toughness, or that quality by which it will endure light but rapid blows without breaking.

(3) Ability to withstand the destructive action of the weather, and probably some organic acids produced by the decomposition of excretal matters, always present upon roadways in use.

(4) The porosity, or water-absorbing capacity, is of considerable importance. There is perhaps no more potent disintegrator in nature than frost, and it may be accepted as fact that of two rocks which are to be exposed to frost, the one most absorbent of water will be the least durable.

60. Breaking and Crushing Tests possess no definite value in determining upon the fitness of a material for paving purposes. It is an elementary fact in mechanics that a body may bear an enormous crushing strain gradually applied and yet be readily broken by a smart blow from a light hammer. Taking the ascertained breaking and crushing strains as lying between 3 and 7 tons per square inch, it may be safely said that no such strains are ever brought to bear upon any single inch of roadway in practice, not

even during the passage of a ten-ton roller. The direct pressure or strain as applied in a testing-machine has no resemblance to the quick blows of horses' hoofs, much less to the abrading action of wheels.

61. Methods of Testing Durability.-The only true test of the fitness of any material for paving is by an experimental trial upon a certain length of roadway under a unit of traffic. The "Rattler" tests now much employed to test the quality of bricks for paving do not fairly represent the condition of the materials in the pavement; in the latter the material is supported on all sides but one, and is subjected to pressure and percussion on this side, while in the "Rattler" tests the materials are thrown into violent collision with large pieces of iron weighing anywhere from five to fifteen pounds. It is evident that under this treatment the corners of the material will readily succumb, and the wear in consequence will be much greater and of a different nature than it would be under actual conditions. The methods adopted for testing any material should represent as nearly as possible the requirements of practical use.

62. The following plan of testing the comparative value of paving-stones is adopted at the Paris Laboratory for Testing Materials. While it may be questioned whether this method is superior to the "Rattler" test, it indicates foreign appreciation of the fact that the "Rattler" test is not what it should be. The stone or other samples are clamped to a horizontal plate revolving round a vertical spindle and brought to bear with equal pressure against a similarly disposed revolving plate of cast-iron. Along with the samples to be tested is placed a specimen of the standard material, which is Yvette sandstone. The coefficient of wear is the proportion between the volumes worn, which is ascertained by weighing the specimens and determining the volume from this weight. The coefficient for first-class materials is from 1 to 1.40, and for secondclass materials from 1.40 to 2.40. If the wear is greater than that represented by the coefficients, the material is rejected.

63. At St. Louis, Mo., some years ago, strips of different pavements 22 inches wide and 8 feet long were laid down as a test, and a two-wheeled cart with tires 2 inches wide, and loaded to two tons, or 800 pounds per inch width of tire, was rolled back and forth by machinery. The heaviest traffic at that time in St. Louis was 75

tons per day per foot of width, and the average for business streets was 35 tons. Estimating the effect of horses' shoes at one third of this amount, 50 tons per foot were taken as a standard. The samples were weighed before and after testing, and were subjected to an amount of travel by the above cart equivalent to eight and one half years on the street.

The total abrasion of the fire-brick pavement was 9%, or a depth of 3 of an inch, but about one half of the bricks were broken. Asphaltum blocks under the same test wore 14%, and but one was broken. Broken stone lost 1% under a traffic of 12.7 tons per foot of width. Broken stone and sand lost 1% under 16 tons per foot. Limestone blocks lost 1% under 4400 tons per foot of width. Wood blocks lost 1% under 12,900 tons per foot, and the granite blocks lost 1% under 70,000 tons.

The action of the elements was not taken into consideration; it would undoubtedly increase the wear of the several materials.

64. Absorptive Power of Stones, etc.-All materials absorb water to a greater or less extent, and their durability is much affected by their absorbing capacity. This capacity depends largely on the density, a dense stone absorbing less than a light and more porous one. The absorbing capacity is a matter of much importance, especially in cold climates. The water absorbed, on freezing, tends by its expansion to disintegrate the stone. It has been said that the act of freezing is equivalent to the blow of a ten-ton hammer on every square inch of surface. Whether this be so or not, the continued expansion and contraction of a porous stone is quite sufficient to disintegrate it, and this disintegration will be the greater the more water the stone contains.

Stones that have already begun to decompose absorb a much larger quantity of water than those fresh from the quarry, and decay will be more rapid. Other things being equal, the less the absorption the better the stone or brick.

65. Description of Materials.-Granite is an unstratified or Igneous rock, composed of silica or quartz, feldspar, and mica. In ddition to these essential constituents, one or more accessory minerals may be present; the more commonly occurring are hornblende, pyroxene, epidote, garnet, tourmaline, magnetite, pyrite, and graphite. And the character of the rock is often determined by the presence of these accessory constituents in quantity.

Granite varies in texture from very fine and homogeneous to coarse porphyritic rocks in which the individual grains are an inch or more in length. The color is also dependent upon the minerals present: if the feldspar is the orthoclase (potash-spar), it communicates a red color; the soda-spar produces gray. The mica. also plays an important part in the modification of color: if it is the white muscovite, it produces no change; but if the black biotite mica be present, it modifies the color accordingly. Hornblende gives a dark mottled appearance; pyroxene as augite also gives a darker appearance; epidote communicates a green color.

The durability of the granites is closely related to their mineralogical composition. The presence or absence of certain species influences the hardness and homogeneous nature of the stone. Although popularly regarded as the most durable stone, there are some notable exceptions.

The quartzose, feldspathic, and micaceous granites are unsuitable for paving purposes. The quartzose are too brittle, the feldspathic are too easily decomposed. When the feldspar is in excess the granite rapidly decays and disintegrates in consequence of the action of air and water on the feldspar, the potash of which seems to be removed, and the residue falls into a white powder composed chiefly of silica aud alumina. The micaceous are too easily laminated.

The term "granite" as popularly used is not restricted to rock species of this name in geological nomenclature, but includes what are known as gneisses (foliated and bedded granites, syenite, gabbro, and other crystalline rocks whose uses are the same); in fact, the similar adaptability and use have brought these latter