Pneumatics and hydraulics:
from the vacuum to the steam engine

Anne-Catherine Bernes

Just like other pieces of apparatus, pump and communication vessels were used in in classical times before the principles of physics on which they were based, were known. Pumps, of which beautiful Roman examples have been found in mines of the Iberian Peninsula, were derived from a bellows with a piston, that had adopted three fundamental elements: the piston, the cylinder and the valve. On the other hand, the use of communicating vessels and also syphons was developed by the builders of aquaducts.

There are two types of pump: the force pump, in which the piston pushes the liquid in the body of the pump into a lateral tube, and the suction pump, in which the liquid rises in the body of the pump when the piston rises by the effect of atmospheric pressure. The fire pump illustrated here (instrument 77) combines the two principles, because in each cylinder the water is alternately first sucked in by the raising of the piston, then pushed by this piston through a lateral tube to the central glass receptacle, and from there into the leather tube. Such a combination of suction and force pump ensures a constant outflow at the exit.

At the beginning of modern times it was taught that "nature abhors vacuum." Diverse phenomena were put forward to support this principle, for example, the adhesion of two perfectly polished surfaces to one another (instrument 62), or the raising of water in the suction pump. That it was impossible to make water rise up more than 10 metres in the suction pump was attributed to the imperfection of this last example.

The existence of the vacuum had previously been recognised in classical times by the atomists Democritus and Epicurus and by the Greek mechanical engineers Philo of Byzantium and Hero of Alexandria. Later, the Italian Renaissance brought renewed interest in atomic theory and the work of Hero of Alexandria, and at the beginning of the 17th century some writers reacknowledged the existence of the vacuum.

In his Discorsi (1638), Galileo had limited the resistance to the vacuum to that of a 10 metre column of air. The reading of his Discorsi provoked a group of learned Romans, brought together around Gasparo Berti, to carry out the following experiment at some time around 1640. Berti fixed a lead tube, approximately 11 metres in length, to the front of his house, sealed at the top with a large copper lid. The lower part of the tube, that contained a valve, was suspended in a tub full of water. At the beginning of the experiment, the tub was full of water and the valve closed. When the valve was opened the level of water in the tube fell to approximately 10 metres.

While supporters and opponents of the vacuum theory discussed amongst themselves the nature of the space situated above the column of water, Evangelista Torricelli (1608-1647) had the idea of replacing the water with mercury. A tube closed at one end was filled with mercury and placed over a tub also containing mercury. Immediately the level of mercury in the tube fell and always stopped at the same height (± 760mm), leaving an empty space above without mercury. Torricelli correctly attributed the cause of the phenomenon to atmospheric pressure on the surface of the mercury. In order for this idea to reach the educated world, it was necessary to wait for the work of Blaise Pascal and the decisive experiment carried out by his brother-in-law Florin-Périer. Florin-Périer repeated Torricelli's experiment three times on the same day; once at the foot of the Puy-de-Dóme mountain, then again halfway up the mountain and then at the summit. Since the column of mercury fell proportionately the further it was carried up the Puy-de-Dôme, it was proved that the result was due to atmospheric pressure so that the column of mercury was higher at low altitude, and lower at high altitude. Pascal's experiments on raising water in the body of the pump and in Torricelli's tube allowed him to develop his work on hydrostatics, explained in his two Traitez de l'équilibre des liqueurs et de la pesanteur de la masse de l'air (1663).

The vacuum experiment became tremendously popular and was reproduced all over Europe and notably by the mayor of Magdeburg, Otto von Guericke (1627-1691). In order to produce a better vacuum than that of Torricelli's experiment, he invented a pump that permitted the creation of a vacuum inside copper spheres. The hemispheres of Magdeburg, reduced replicas of which are exhibited here (instrument 69), have become famous: Guerick had succeeded in making a vacuum in the interior of two hemispheres of approximately 80 cm in diameter joined together at the base. In 1654, presented before the Diet of Regensburg, they could only be separated by the strength of eight horses. Two bigger hemispheres required the strength of twenty-four horses.

The Englishman Robert Boyle (1627-1691) perfected Guerick's pump with the help of Robert Hooke. Besides the First Law of Gases (known by the name Boyle-Mariotte, stating that the pressure of a gas is inversely proportional to its volume), the works of the learned Englishman permitted the development of the technical manipulation of gases. His study of the properties of air was the beginning of a plethora of further studies which produced some of the most important discoveries in the fields of physics and chemistry, in particular the discovery of oxygen by Joseph Priestley (1733-1804) and Antoine-Laurent Lavoisier.

Torricelli's experiment had given rise to a new instrument called a barometer (Mariotte, Essai sur la nature de l'air, 1676). Invented a little after the Puy-de-Dôme experiment, it was clearly designed for two independent uses: to forecast the weather and measure altitude. The barometer has had a varied development since then, depending on its function.

For the first type of instrument, the work concentrated in making the information given by the height of the column of mercury more readable. Robert Hooke was the first to combine Torricelli's barometer with a mechanical system that permitted the reading of the details on a quadrant. This type of quadrant barometer, with a few modifications, was the most widespread for two hundred years. An example is on display here (instrument 78).

The second type of instrument has undergone a series of improvements aimed at resolving the difficulties connected with its transportation. In these travelling barometers (instument 89) the main modifications were as follows: the tube was constructed in a manner which minimized the risk of breakage, the container was carefully shaped so as to prevent air entering during transportation, the top of the tube was sealed during movement and the barometer was fixed on a tripod that served as a protective case when it was closed.


The aeolipile, "ball of Eolo", is a hollow, pear shaped metal container open at its narrowest part. Filled with water and heated it permits a strong current vapour to escape. Seen here (instrument 80) is an example mounted on a vehicle which it moves forward through the phenomenon of reaction (this phenomenon was not understood in classical antiquity). In Vitruvius (1st BC) the aeolipile was no more than an apparatus that served to demonstrate the origin of the winds. Blowing as much air as water vapour it was used as bellows throughout the classical times and the middle ages.

Hero of Alexandria also described in his Pneumaticas a little sphere perforated by two nozzles bent at right angles that rotated by means of the escaping steam supplied by a boiler. This machine, inappropriately called an aeolipile, was frequently wrongly considered as the ancestor of the steam turbine. It still entertained natural philosophy audiences in the 18th century. As can be seen here (instrument 82), it was sometimes attempted to transfer the movement of the sphere to a mechanism that would do work; in this case to pick up a body.

Rediscovered and popularized in the 16th century, the work of Hero attracted the attention of inventors towards the exploitation of water vapour. In 1629, Giovanni Branca demonstrated a system in which a bronze boiler, well heated, emitted a jet of steam that hit the spokes of a wheel with sufficient force to make it turn. This principle of the steam turbine only gained practical application in the 19th century.

In 1601 Gianbattista della Porta (1535-1615) described an apparatus where the pressure of water vapour would raise a column of water; on the other hand, the condensation of the vapour created a depression that in its turn sucked in water. This principle was to be rediscovered one hundred years later by Thomas Savery, the inventor of a steam machine destined to extract water out of mines.

Denis Papin (1647-ca. 1712), a contemporary of Thomas Savery, worked on another principle of steam machines; the machine called atmospheric. His work is closely connected to the experiments with the vacuum carried out by Torricelli, Pascal, Otto von Guerick, Huygens and Boyle (who collaborated with Guerick and Huygens). In order to produce this type of machine it was necessary to be aware of the enormous forces exercised by air on surfaces, as well as the pressure exerted by water vapour (demonstrated in his "digester"-[instrument 81]) and the vacuum created by condensation of this water vapour..

The honour of having perfected the first operational steam machine belongs to the Englishman Thomas Newcomen (1663-1729), in 1712. Like that of Papin, which was never exploited, it is an atmospheric machine. It is composed of a vertical cylinder which encloses a piston, whose shaft is joined by a chain to a rocker arm. The shaft of the piston is fixed at the other end of the rocker arm, also by a chain. There is a large copper container of boiling water over the cylinder of the steam machine. Under this boiler there is a furnace.

Under the weight of the piston of the pump, the rocker arm remains inclined towards it. In this position the steam originating from the boiler is introduced into the cylinder under the piston. When the piston reaches the top of the cylinder the admission of vapour is interrupted while a jet of cold water is automatically directed at the cylinder. The cold water causes the rapid condensation of the steam contained in the cylinder. The steam now transformed into water occupies less volume in the cylinder, thus creating a vacuum. The atmospheric pressure that is exerted on the external face of the piston ceases to encounter resistance and pushes the piston towards the bottom of the cylinder. Once it reaches the bottom, the piston starts to rise again, driven by the pump. It then starts all over again.

The Englishman James Watt (1736-1819) made decisive improvements to Newcomen's machine. In his machine, instead of the condensation of steam occurring inside the cylinder, it is condensed in a separate condenser. In order to keep the cylinder as hot as possible, it is surrounded by a casing full of steam, consequently the atmospheric pressure which served as the driving force in Newcomen's machine now plays no part in Watt's machine: the descent of the piston, following the evacuation of steam in the condenser, is caused by the pressure of the steam on the upper face of the piston. The alternate movement of the rocker arm was transformed into a rotating movement by a system of piston rod and crankshaft, setting a flywheel in motion. Thanks to these improvements the steam machine was able to be adapted to all types of work.

One of the applications of the steam machine was to move a vehicle. It was to this end that Richard Trevithick built the first high pressure machine, using steam at a higher pressure than atmospheric pressure (instrument 83). To achieve this he created a cylindrical boiler inside which he placed a furnace. He eliminated the condenser. The cylinder was encased at one end of the boiler in such a manner as to keep it hot. This machine was, moreover, one of the first types of machine in which the rocker arm is substituted by a beam that moves up and down over a stream of water, carrying a piston rod that causes the rotation of a shaft by means of a crank.