Paris, January 28, 1887. On the Champ-de-Mars a row of workers digs the foundations of four great feet of stone and concrete. Above them, over the next two years, will rise 18,038 pieces of puddled iron held together by two and a half million rivets. The client is the Republic, which wants a monumental entrance for the 1889 Universal Exhibition, the centenary of the Revolution. But among Paris intellectuals the tower is not liked: in February, before the first platform is even finished, a group of artists (including Guy de Maupassant, Charles Garnier, Charles Gounod) signs an open letter against that «odious column of bolted sheet metal». The man who wanted it is called Gustave Eiffel, and he is not an architect: he is an engineer.

An engineer who builds bridges

Eiffel is born in Dijon in 1832, graduates from the École Centrale in Paris and in 1867 founds his own company. He has a specific talent: building large, light iron structures, building them in the workshop from standardised pieces, and then assembling them on site. With this method he builds the Maria Pia bridge in Porto (1877), a 160-metre iron arch over the Douro, and the Garabit viaduct (1884), in the Massif Central, 122 metres above the river Truyère. He also designs the internal structure of the Statue of Liberty (1881–84), the iron frame that supports Bartholdi's copper skin: a project born from the ideas of Viollet-le-Duc, who died in 1879, and then taken over by Eiffel.

The idea of the tower does not come from him. In 1884 two engineers of his company, Maurice Koechlin and Émile Nouguier, design a very tall pylon made of four uprights curving toward the centre. The architect Stephen Sauvestre softens it with arches at the base and glazing on the platforms. Eiffel buys the patent rights, takes part in the 1886 competition and wins it. It is not a generous gesture, but it is the signature of a method: he did not invent the form, he invented the way of building it.

The Eiffel Tower — the numbers
ConstructionJanuary 28, 1887 – March 31, 1889
Original height300 m (today about 330 with the antenna)
MaterialPuddled iron · about 7,300 t
Pieces and rivets18,038 pieces · 2.5 million rivets
RecordWorld's tallest structure until 1930

How to build a tower in two years

The speed of the construction is the most surprising part. The pieces are designed with millimetric precision and prefabricated at Levallois-Perret, in Eiffel's workshop, then taken to the Champ-de-Mars and assembled with steam cranes and wooden scaffolding that follow the tower's growth. Every hole is already in the right place: on site nothing is cut or drilled, one only rivets. In little more than two years the structure reaches 300 metres, with a very low number of fatal accidents for the time. On March 31, 1889 Eiffel climbs to the top for the ceremony, on foot, up 1,710 steps, and hoists the French flag.

The tower was to be dismantled after twenty years, in 1909, according to the permit granted. It is saved because Eiffel, with foresight, turns it into a scientific instrument: he installs a weather station, an aerodynamics laboratory and, from 1898, wireless telegraphy experiments that soon become a permanent antenna. By 1909 its value for military and civil communications is such that the concession is renewed. Later Eiffel also builds a wind tunnel at Auteuil (1912), which still bears his name today.

The iron that became a symbol

The writers who attacked it get used to it. Maupassant, it is said, lunched at the tower's restaurant because it was the only place in Paris from which one could not see it. In the twentieth century Roland Barthes makes it a symbol that everyone fills with whatever meaning they like. The framework, with its parabolic form calculated to resist the wind, shows that an infrastructure can also be a monument. And the monument, in turn, has been a model for generations of engineers: without the Garabit viaduct and without Eiffel, the architecture of iron and later of steel would not have had the same road ahead.

«The Eiffel Tower is not beautiful because it was designed to be, but because it was calculated to stand up in the wind. Its curve is the curve of the bending moment, and whoever climbs to look at it up close understands that the form is not an image: it is an equation that has been given a skin of iron.»