a palace belonging to the king of France, between Versailles and St Germain; seated in a valley, near a village and forest of the same name. It is noted for its fine gardens and water-works, there being a curious machine on the river Seine, which not only supplies them with water, but also those of Versailles. It is 10 miles north-west of Paris. E. Long. 2. 11. N. Lat. 48. 52.

Machine at. When Lewis the Great had fixed upon a favourite situation in the forest of Marly, where he intended to erect a splendid castle, he found that it wanted nothing either in point of beauty or convenience but a fountain of water; and he immediately determined to supply by the assistance of art what nature had denied it. An ingenious and self-taught carpenter from Liege, named Rannequin, undertook to conduct from the Seine a copious supply of water, and for this purpose contrived and erected the celebrated and complicated machine which we are now to describe.

The machinery is driven by 14 underflot water wheels of 36 feet diameter, reckoning from the ends of the floatboards, disposed in three rows. In the first row there are seven wheels, in the second fix, and in the third only one. By these wheels the water is raised through pumps into the first reservoir about 160 feet above the level of the river, then to a second reservoir 346 feet high, and from this to the summit of a tower about 533 feet above the Seine.

The two extremities of the axle of each wheel extend beyond the gudgeons on which they rest, and are bent into a crank so as to form a lever two feet long. The crank which is towards the mountain drives the water of the river into the first reservoir, and the other crank gives motion to the balances.

An engine of eight pumps is wrought by one of the cranks of each of the six wheels in the first row. These engines consist of a balance, at each end of which hangs a square piece of wood that supports and directs four pistons. This balance is moved by a beam in the form of a T, the horizontal part of which is connected at one end with the balance by the intervention of a vertical regulator or beam, and at the other with the crank of the wheel by means of a horizontal iron rod.

One of the cranks of each of the six wheels of the first row, (excepting that which is next the mountain), and two of the cranks of the 14th wheel, or that in the last row, give motion to the pumps in the river, and carry the water into the first reservoir. This motion is communicated from the cranks by means of an iron rod which is fixed to the lower end of a vertical balance. A horizontal regulator or beam is fixed to each end of this balance, and to these regulators are fastened chains which follow the declivity of the mountain till they reach the superior reservoirs. When the wheel is revolving, therefore, one of these chains will be dragged towards the river, and the other towards the mountain. In order to produce this alternate motion, the chains are supported and kept at equal distances by a number of vertical balances, placed along the mountain at every three toises, and moving upon a centre supported by a frame lying between the two chains and equidistant from them. When these chains reach the first reservoir they are fixed to vertical regulators, which carry frames, to which are adapted the pistons of the sucking pumps. These regulators therefore will be drawn one after another by their corresponding chains; and when one regulator is drawn by its chain, the piston of the pumps which it carries will be raised, and the water will follow them: At the same time the pistons of the other regulator are descending to form a vacuum; and these in their turn ascend with their load of water when the others are in the act of descending. In the pumps formerly mentioned which work in the river, an effect is produced upon the pistons both when they ascend and descend, because they are moved by stiff iron rods; but in the present case the pistons descend merely by their own weight, as the motion is transmitted only by a chain. By these pumps the water is conveyed to the upper reservoir by two conduit pipes of eight inches, and three others of six inches diameter.

The sixth wheel of the first row, which is the first towards the dam, moves a long chain which works the pumps of one of the wells of the upper reservoir. The seventh wheel gives motion to a chain which goes to the first cistern.

By means similar to these already described, the six wheels of the second row move by each of their cranks a chain that goes to the second reservoir, and eight of these chains work 16 pumps behind it, to bring back into the reservoir the water which is lost out of the six pipes that go to the tower. These chains go over one of the first cisterns, and five of them at the same time give motion to the pistons of thirty pumps, whilst the other chains go on straight to the great reservoir. These 30 pumps convey their water through two pipes of 8 inches diameter into the upper reservoir. The five chains, after working these 30 pumps, give motion to the pistons of 82 pumps in the second reservoir which raise the water from it to the tower.

The basis of the tower which receives the water raised from the river is 610 fathoms distant from it; and the water runs from this basin along an aqueduct of 36 arches by its own weight. From this aqueduct the water is distributed into great reservoirs, from which it is conveyed to the gardens and shrubberies around the castle.

The quantity of water raised by this machine amounts at a mean rate to 30,000 or 40,000 gallons per hour; though in favourable circumstances it raises more than 60,000 gallons per hour. But while the Seine either overflows its banks, or is frozen, or when the water is very low, the machine is scarcely capable of performing any work.

The yearly expense of the machine at Marly including the salaries of the superintendents and the expenses of repairs, amounts to about 3300l. sterling, or 6l. per day, which makes the expense of 90 gallons of water one farthing. But if we take into the account the interest interest of 333,000l. the original expence of the machine, 90 gallons will cost three halfpence, or 15 gallons one farthing.

Notwithstanding the magnificence of this great machine, and the ingenuity which is displayed in its construction, every person who examines it with care, will perceive innumerable defects, whether he examines it as a whole, or attends to the parts of which it is composed. In several positions the moving forces act with great obliquity, and therefore occasion an immense waste of power; and in order to give an alternate motion to a number of chains of balances extending to a distance of 3804 feet, more than nine-tenths of the impelling power are destroyed.

By a few changes upon the construction of the machine, the water might have been raised from the river to the tower without any intermediate reservoirs. This appears from two experiments made upon the machine in 1738 and 1775. In 1738 M. Camus attempted to raise the water to the tower at once. He was able, however, only to bring it to the bottom of the tower which was considerably higher than the second reservoir. By this experiment the machine was so much strained that several parts required chains to secure it. In 1775 the water was elevated to the second reservoir at one jet at different times, but from the age and infirmity of the pipes several of them burst during the experiment. Hence it is obvious that if the pipes had been made stronger, the first reservoir and the machinery connected with it might have been dispensed with; and it is very probable that if the machine had been constructed with more judgment, the water might have been conducted at once from the river to the tower.