The time at the third belch will

During the late 1870s, a network of pipes was installed beneath the streets of Paris - not for carrying water or sewage or gas, but for the distribution of time. Underground steam engines pulsed regular bursts of compressed air through the pipes, delivering the exact time of a mother clock to other clocks all over the city. As the pipes belched, the city's clocks ticked - precisely in step with each other.

Or nearly. Despite racing through the pipes at the speed of sound, the pulses of air took 15 seconds to cross the city, so some clocks lagged behind by a full quarter of a minute. For the Victorian engineers of temporal exactitude, this simply wasn't good enough. They modified the pneumatic clocks to take their distances from the mother clock into account and began the search for a better method of synchronisation.

The Parisian system was just one of a number of efforts to bring temporal order to the chaos of local times. Midday in one town occurs before the sun passes directly overhead a town further west, so clocks set locally will show different times. The coming of the railways highlighted these discrepancies, but as the railway men tried to introduce consistent timetables, the different times multiplied. A single railway station would work to three times - the local time of the town, the train time determined by the main city and the platform time, which ran a few minutes behind the train time to enable all those late passengers to catch their trains. Lines stemming from different cities would often run to yet different times.

Timetabling confusion over which time was which, or a conductor's watch running a bit slow, could mean a fatal train crash farther down the line.

The solution was for everyone to agree to a global standard time, with clearly demarcated time zones and accurate clocks sending out time signals across the globe. By the end of the 19th century, time had become a matter of convention, something to be debated at international meetings, something to be agreed on.

This preoccupation with synchronising clocks and the conventional nature of time is at the centre of Peter Galison's Einstein's Clocks, Poincaré's Maps . As Galison himself puts it in rather unprepossessing terms, this is a book about clock-coordinating procedure.

But this is also a book about far more than that. As the title suggests, it is also about the turn-of-the-century revolution in physics - about Einstein, the French mathematician Henri Poincaré and the theory of relativity.

Galison's project is to reveal the ways in which the development of relativity was inextricably bound up with the technology of time and with international conventions on time. In his scholarly work, Galison has examined the ways in which 20th-century physics is embedded in a material culture. Here he presents a similar thesis to a non-specialist audience.

Popular science books about relativity usually focus on the lone figure of Einstein and tell a story of heroic genius, of mind-bending ideas suddenly bursting forth from this obscure patent clerk. But the narrative of genius is misleading. It paints an alienating picture of science as mysterious and incomprehensible, as something divorced from the wider culture. All too often, popularisations describe their genius heroes as being "before their time". Galison shows what nonsense this is.

Einstein was not alone in focusing on the nature of simultaneity or the principle of relativity - two of the starting points for his theory of special relativity. Poincaré had also written on these subjects, and it is on Poincaré, rather than Einstein, that Galison concentrates. In the 1890s, Poincaré had already set out many of the key ideas that would later go into the theory of relativity. But he failed to take the final radical steps that the young Einstein realised were necessary.

Galison argues persuasively that both men's ideas were of their time, but his focus on the synchronisation of time means that we risk losing sight of what it was that Einstein did do differently. Although Galison notes that Einstein focused on kinematics rather than dynamics, we do not really get a clear sense of what this means or how the questions of synchronisation related to problems in electromagnetism. Generally, Galison skips a few too many steps and assumes a little too much prior knowledge for his explanations of the technical issues to be as clear as they might be.

But where Galison excels is in portraying the temporal culture of which Einstein and Poincaré were a part. Neither man was a purely abstract thinker - they also dealt with the concrete technologies of time. As president of the French Bureau of Longitude, Poincaré was at the heart of the struggles to bring order to the measurement of time. Einstein, too, was immersed in the culture of time, evaluating patents submitted to the world's premier clock-making nation.

Galison argues that by the late 19th century, time had acquired a material presence. It was real stuff that needed to be distributed around cities and across the whole globe.

In France, time was puffed through a network of pipes. In America, time was, quite literally, money, as American astronomers marketed and sold the precision time fixed in their observatories.

More than a century later, time engineers still worry about how to synchronise clocks. Atomic clocks keep track of seconds with a degree of accuracy that Poincaré and his colleagues could only dream about, but these seconds do not synchronise perfectly with the rotation of the Earth. At the end of the 19th century, scientists and engineers were synchronising clocks all over the world to within a fraction of a second. Today, we add in whole leap-seconds to try to keep our technologies of time in sync with nature.

Felicity Mellor is lecturer in science communication, Imperial College London.