166

HISTORY OF ENGINEERING.

Book I.

when they selected a spring, on which they considered the health of mankind so muchdepended. Before an open or running stream was laid on to a town, they examined theinhabitants of the neighbourhood; if strongly formed, fresh coloured, sound legs, andwithout blear eyes, they considered the water good. By throwing a drop of water into aclean brass vessel, if it left no stain, it was thought pure. And after boiling, if no sedi-ment was deposited, it was equally so. If in the spring it appeared limpid and transparent,and no moss or reeds generated near it, the water was considered light and wholesome.

Instruments used by the Romans for Levelling. The libra aquaria and the dioptera werenot for this purpose considered so correct as an instrument called the chorabates, which wasa rod or plank, about 20 feet in length, mounted on a leg at each of its extremities, both ofequal length. The rod and the legs were fastened or secured by diagonal cross pieces orbraces, on which were marked correctly vertical lines. A plumb line attached at eachextremity and acting over these diagonal braces indicated whether the instrument was level.When the wind prevented the plumb bobs from remaining stationary, a channel cut in theupper edge of the horizontal rod was filled with water, and if the water touched equallyboth extremities, the level was supposed to be correct, and then the observation of thedescent or elevation of the ground was made with accuracy. Vitruvius observes thatalthough Archimedes asserts that water is not level, but takes the form of a spheroid whosecentre is that of the earth, yet the two ends of the channel on the rod will neverthelesssustain an equal height of water. If it be inclined towards one side, that end which ishighest will not suffer the water to reach to the edge of the channel on the rod. So thatthough water poured in may have a swelling and curve in the middle, yet at its extremitiesit will be level. When this instrument was used, if the ground was very unequal, the feetwere propped up, and supported till they were brought level. The other two instruments,as the water-level and the dioptera, are not very accurately described.

Conducting of Water. This was effected in various ways by the Romans, either in channelsbuilt to convey it, or in earthen or in leaden pipes. When channels or aqueducts wereadopted, they were solidly and substantially executed, with a fall of not less than six inchesin a hundred feet, and arched over at top, to prevent the sun from affecting the water.When the water arrived at the walls of the city, a reservoir or castellum was built, whichcontained three cisterns to receive it. In the reservoir were three pipes of equal diameter,so connected, that when the water overflowed at the extremities, it was discharged into themiddle one, which supplied the pipes for the fountains; a second pipe supplied the baths,and a third the private houses. The water for public use was never deficient, nor could itbe diverted if the mains or pipes were properly constructed. The private houses had atax levied upon them, which was expended in keeping the aqueduct in repair.

When hills intervened between the spring head and the city, tunnels were driven underground, with a fall of one in two hundred; when the material cut through was stone,a channel was formed in it; when of earth or gravel, side walls were built, and an archturned over, through which the water was conducted. Wherever these tunnels wereformed, perpendicular shafts were sunk every 120 feet distance. When the water wasbrought in leaden pipes, a reservoir was made near the spring, and other pipes of sufficientdiameter conducted it to the city reservoir; these were made in lengths of not less than tenfeet, out of sheet lead of different widths and weights; hence a sheet of 100 inches widewould make a pipe weighing 1200 pounds; 80 inches wide, 960 pounds; 50 inches wide,600 pounds; 40 inches, 480 pounds; 30 inches, 360 pounds; 20 inches, 240 pounds; 15inches, 180 pounds; 10 inches, 120 pounds; 8 inches, 96 pounds; and 5 inches 60 pounds.

When pipes of this kind were used, the lead varied in weight, though it was generallyat the rate of fifteen pounds per superficial foot; if there was a uniform and proper fall, anylittle impediment was made up by means of substructions, or by taking a circuitous course,provided it was not too far about; the pipes were all laid with a regular and proper current.

If the valleys were long, they took the slope of the hill, and when the water arrived at thebottom, it was carried across the valley by a low substruction at as great a distance as possible;a venter here prevented the w r ater on its arriving on the opposite side or acclivity frombursting or destroying the joints of the pipes. Over the venter were placed lofty uprightpipes, by which the violence of the air escaped. Tlius when water w*as conducted throughleaden pipes, due attention was paid to the fall of its descent, its circuit, its vent, and thecompression of the air. It was, however, always found expedient, after the fall from thespring was obtained, to build reservoirs, at distances of 20,000 feet, to allow of repairs.These reservoirs were not made on any descent, nor on the venter, nor on a rise, nor invalleys, but only on plains.

When economy was considered, earthen tubes were substituted, not less than two inchesin thickness, so made, that one end fitted into the other. Tty ? joints w’ere then coated witha mixture of quick-lime and oil, and on the elbows made by the level part of the venter,instead of the pipe, was placed a block of red stone, so perforated, that the last length ofinclined pipe, and the first length of the level part, were received into it. On the oppositesides, where the acclivity commenced, a block of red stone received the last length of the