Coupe et vue générale de la première machine de Marly

Cross section and general view of the first Marly machine.

The Niépce Brothers participate in the Competition to replace the Marly Machine

In 1807, the imperial government opened a competition to receive projects for hydraulic machines to replace the one in Marly, used to deliver water to the Palace of Versailles. Built in 1684, the original machine — located in Bougival, on the Seine river — pumped up water over one kilometer and a vertical drop of 150 meters.

For the competition, the Niépce brothers imagined a new principle for the machine:
“The theory of this machine, which we call hydrostatic pump, is based on a simple principle that consists in the equilibrium of two water columns, the heights of which are in opposite ratio to their diameters. One of these columns represents power, and the other resistance. To transmit the action of the first one, we do not need to resort to wheels and levers. The power force acts in a direct manner. It is independant of the water motion: the loss of this liquid is only used to determine the succession of effects. Two pump bodies, two pistons, three valves, and a ram; this is all that composes the machine mecanism. {…}.
This machine has only been finished for a few days: it is our intention to present it to the Institute.”
from a letter to L. Carnot, December 24th, 1807

La Machine de Marly par Pierre-Denis Martin, 1723.

The Marly Machine by Pierre-Denis Martin, 1723.

“As you use no wheels, and no levers, this presents a great advantage”, Carnot admits {…} “I advise you to realize precise experiments on the produce of this machine before you present it.”
Carnot’s answer, December 31st, 1807

Two years later, on December 8th, 1809, the Niépce brothers have considerably improved their machine:
“The machine has undergone a lot of changes in many of its parts. The mecanism in the system is more elaborated: its pistons join to the advantage of being more precise, another one that is to create far less resistance. We tested it many times, and the result was that with a drop of 4 feet 4 inches, it lifts to 11 feet the 7 /24 of the water it loses.”

Carnot answered on December 1809. They had waited too long because the Emperor took himself the decision to ask the engineer Perier (1742-1818) to build a fire machine also known as a steam engine, to operate the pumps at Marly. Comparing it with other projects, Carnot appreciated the Niépce brothers’ machine. It seemed to him “very good” and simpler than the rest. In conclusion, he wrote: “It also seems to me that your machine is not overly complicated and under all these aspects, it can confirm the idea of your talents that were already given with your fire engine.”