The Century of Standards

The High Contracting Parties undertake to found and maintain, at their common expense, a scientific and permanent International Bureau of Weights and Measures, with its seat in Paris.

A treaty for a bar

On 20 May 1875, in the halls of the French Ministry of Foreign Affairs, the plenipotentiaries of seventeen states signed an international treaty. It did not close a war, did not partition a continent, did not regulate a commerce: it bound the signatory powers — there were the empires of Germany, Austria-Hungary, and Russia, there were the United States and Brazil, there was the Ottoman Empire, there was a newly made Italy — to found and maintain together, in perpetuity, a common laboratory to keep the standards of measure. Nineteenth-century diplomacy, which usually gathered to draw frontiers, gathered this time for a bar of metal and a cylinder: and treated the matter with the same seriousness, because it had understood — it is the discovery that gives this chapter its title — that standards are frontiers, only traced elsewhere: they pass through the workshops, the markets, and the laboratories, and decide who may trade, build, and measure together with whom [Quinn 2011].

To grasp how far the idea was in the spirit of the age, one need only look at the calendar of signatures of those years: in 1865 twenty states had founded the International Telegraph Union, so that telegrams might cross frontiers without retranslation at every boundary; in 1874 the Postal Union had done the same for letters; 1875 fell to measures, and there would follow patents, copyrights, railways. The mature nineteenth century was discovering a new institutional genus — the permanent technical union, the piece of sovereignty placed in common so that the networks might work — and the Metre Convention is its most radical exemplar: the other unions harmonized rules, this one kept objects, and was the first to equip itself with a common laboratory with its own staff, its own budget in gold francs apportioned by quota, and a mandate without expiry. Historians count it among the very first embryos of international organization on the planet: before giving itself a tribunal or a league of nations, the world gave itself an office of the metre [Quinn 2011].

The Convention created three things, and it is well to introduce them at once by their baptismal names, for they will accompany the book to its last page. A parliament of measures: the General Conference on Weights and Measures, the assembly of the delegates of all the member states, which would gather at intervals of a few years to sanction the supreme decisions — it is the assembly that will vote, in the seventh and eighth chapters, the revolutions of 2018 and those to come. A restricted council of scientists, the International Committee, to act as government. And a physical place, the International Bureau of Weights and Measures, the permanent laboratory evoked by the first article: the operative arm, the home of the standards. Three organs for a function that had never existed before in history: to administer, in the name of all and on behalf of none, the quantitative grammar of the world.

One great absence, at the signing, deserves to be noted: the United Kingdom. The first industrial power on the planet stayed outside the treaty for nine years — it acceded only in 1884, and without thereby adopting the metre in internal use — and the irony of the dates is exquisite: it is the same year in which, at Washington, the world was making England a gift of the meridian of Greenwich. The two capitals of measure thus divided between them, in the same span of months, the two halves of the measurable universe: time to the English, space to the French — each acceding with reluctance to the other’s standard. International standards, we shall see, are always born this way: not from the surrender of one party, but from an exchange in which each can tell himself the winner.

Why France ceded the metre

The right question, before 1875, is not why the world wanted the metre: it is why France agreed to share it. For three quarters of a century the mètre des Archives had been a national standard, kept in Paris, accessible to foreigners by grace and with parsimony; France was as jealous of it as one is of a family relic. Three forces, grown in the central decades of the century, made that jealousy untenable.

The first was commerce, and its new theatre: the universal expositions. At the Crystal Palace in 1851, and then at Paris in 1855 and 1867, the products of all nations found themselves for the first time side by side in the same pavilions — and with the products their inches, their pounds, their metres, in a cacophony that the international jurors had to arbitrate conversion by conversion. The expositions equipped themselves with committees for the unification of measures and of currencies, and those committees became the place where engineers and merchants of different countries discovered that they wanted the same thing. The second force was science, and once again — the reader will smile — geodesy. In the sixties the European countries had undertaken the joint measurement of the continental arc of meridian, joining together the national networks of triangles; but to join triangles means to compare the rules with which the bases were measured, and the geodesists found that the national “metres,” copied copy by copy from the Paris standard, differed among themselves by enough to spoil the calculations. It was the association of geodesists, in 1867, that formally asked for a new international metre, manufactured and verified in common: the measuring of the Earth, which had given birth to the metre, now demanded its refounding [Quinn 2011].

The third force was the subtlest: the standard itself was aging, and in the most instructive way possible. The metre of 1799 was an “end” standard: it measured one metre between its two terminal faces, and every comparison meant touching those faces — which wore away, imperceptibly, at every touch. The reader will recognize the toise of the Châtelet and the fate of every material standard: the object that defines the measure is also the one object in the world that cannot be calibrated. To this had been added, with the perfecting of geodesy, the certainty spoken of in the second chapter: the metre was in any case shorter by a fifth of a millimetre than its terrestrial ideal. When, therefore, in 1870, France convened in Paris an international commission to discuss new prototypes, on the table there was a philosophical choice disguised as a technical question: to remeasure the Earth and correct the metre, or to declare that the metre was henceforth that bar, errors included, and confine oneself to copying it better? The commission — gathered just in time to be dispersed by the Franco-Prussian War, and reconvened in 1872 with the Germany of Bismarck seated at the table of a still-wounded Paris — chose almost without discussion the second road: the new prototypes would reproduce the mètre des Archives “in the state in which it is.” No one wanted any longer to hang the unit on an Earth that had proved elusive; everyone wanted to be able to trade and triangulate with the same length. It was the official, and almost unnoticed, leave-taking from the dream of the Revolution: nature was leaving the definition of the metre, convention was entering it openly. “For all times, for all peoples” had come true by half — for all peoples, just so, and that was enough.

The neutral pavilion

There remained to decide where the new measure would dwell, and it is the point at which the affair touches its highest diplomatic finesse. An international standard kept in a national laboratory would have been a contradiction: trust, to belong to all, cannot be the guest of one alone. The solution was to invent a place that was not, metrologically speaking, in any country: France offered the Pavillon de Breteuil — a seventeenth-century residence in the park of Saint-Cloud, on the hill of Sèvres above the Seine, built for the court of the Sun King and reduced to ruin by the bombardments of the war of 1870 — and the treaty consecrated it the seat of the International Bureau, with a statute of inviolability that brings it close to an embassy of science: French territory on paper, everyone’s territory in substance. The choice of the ruins was not without eloquence: the pavilion devastated by the war between the two countries that had just sat down at the same table was reborn as the common home of precision — the Europe that had cannonaded itself now taxed itself, in gold francs and by article of treaty, to keep its standards together [Quinn 2011].

On what was to happen inside it the last negotiating battle was fought, and it was anything but bureaucratic: deposit or laboratory? One part of the delegates imagined a simple strongbox — the prototypes in a safe, a verification now and then, and everyone home again. The other part, led by the scientists, knew that a standard is never merely an object: it is the object plus all the operations that make it usable — the comparisons, the thermometers, the balances, the staff who know how to make measures to the millionth and who, in order to know how, must make them every day. The living laboratory won, and it is the victory from which all the rest descends: the Bureau of Sèvres became, within a generation, the place where the metrologists of all continents were trained, where the thermometers and comparators of the new generation were invented, where every country sent its standards to “take the measure” as once one sent the wooden arm to the walled standard of the commune.

The daily life of the pavilion, in its first decades, was less solemn and more interesting than its liturgies. To direct it was called the Norwegian Ole Jacob Broch — mathematician, physicist, and former minister, chosen also because he came from a small country that frightened no one — and the staff, by international statute, mixed nationalities and schools from the start, with the Italian Gilberto Govi among the founding fathers within the Committee. The dominant craft, surprisingly, was not the measuring of lengths but of temperatures: to compare two bars to the tenth of a micron one must know their temperature to the thousandth of a degree, because metal breathes with the thermometer — and the prototypes, by convention, were worth one metre at the temperature of melting ice, with the coefficient of expansion of each copy measured and certified separately. Thus the temple of the metre became, of necessity, the world’s sanctuary of thermometry: its mercury thermometers, studied to the point of pedantry in every vice of the glass, were for half a century the most reliable ever built, and the practical scale of temperatures was born within those walls as a by-product of the keeping of a bar. It is a small organizational theorem we shall meet again: he who keeps a standard ends by having to master, one after another, all the magnitudes that disturb it [Quinn 2011]. The reader of the first chapter is entitled to see, in the pavilion among the trees of Sèvres, the arm on the door promoted to a planetary institution: identical function, identical logic of trust, scale multiplied by the world.

Thirty metres of an almost eternal alloy

The new standards were conceived as the old metre could not have been: designed, in the full engineering sense of the word. The matter, first of all. The pure platinum of 1799 was noble but soft; the commission chose an alloy of platinum with a tenth of iridium, very hard, unoxidizable, stable as no metal then known — the material closest to eternity that nineteenth-century chemistry could offer. The form, then, was a small masterpiece of rational mechanics: the engineer Henri Tresca designed for the bar an X-section that concentrates the metal where rigidity is needed and removes it where it is not, and which above all brings to the surface, along the neutral plane of the bar — the fibre that neither lengthens nor shortens when the bar flexes — the plane on which the marks are engraved. For this was the other break with the past: the new metre was a line standard, a metre between two engraved lines and not between two faces exposed to wear. One no longer touched it: one looked at it under the microscope. The standard, after centuries of being worn away, had ceased to fear verification [Quinn 2011].

The manufacture was a small industrial novel in itself. The great celebratory casting of 1874, executed in Paris with all due ceremony, was rejected years later by the analyses: impurities beyond the tolerable. They had to begin again, and — an irony no novelist would have dared — the alloy of the prototypes of the metre, monument of French science, issued from the crucibles of a London firm specializing in precious metals, Johnson Matthey, subject of an empire that had not even signed the Convention. No one was unduly scandalized, and it is the sign of the new times: precision was becoming a transnational industry, and to buy the best alloy mattered more than the supplier’s flag. From the rough bars, polished, engraved, and compared for years with the mètre des Archives and among themselves, issued a population of some thirty twin metres — and, for mass, some forty cylinders of platinum-iridium, children of the kilogramme des Archives. In September 1889 the first General Conference christened them: one bar, chosen because more than any other it coincided with the old metre, was declared the International Prototype; one cylinder, the International Prototype of the Kilogram — the character who will dominate the seventh chapter. The mètre des Archives, after ninety years of reign, passed in the same instant from the status of law to that of relic: it stayed at the Archives, honoured and now mute, the first of a dynasty of deposed standards that metrology keeps as the cathedrals keep their crypts — out of filial piety, and to remember where it comes from. The others were distributed to the states by the most ancient and most equitable method in the world: the urn. Each delegation drew by lot the numbers of its own national standards — to the United States, for instance, fell metre number 27 — and carried them home with a receipt, as one collects a lot at a charity raffle. The measure of the world was literally drawn by lot: no favour, no precedence, the blindfolded goddess to guarantee that not even in the quality of the copies should there be a hierarchy [Crease 2011]. Each standard then set off for its country like a small sovereign on a journey: a padded case designed for the purpose, a dedicated courier, and in its train its own civil status — the certificate that declared its exact deviation from the international prototype and its personal coefficient of expansion, measured at Sèvres. For this is the finesse that distinguishes mature metrology from the mere manufacture of copies: no copy is equal to the original, and the system does not pretend it is — it declares by how much it differs, and from that moment the known difference ceases to be a defect and becomes a correction to be applied. The reader will recognize the rate of the chronometers of the previous chapter: here too, the measured error is worth as much as exactness.

The two international prototypes descended instead into a vault dug beneath the pavilion of Sèvres, inside a safe whose liturgy has remained famous: three different keys, entrusted to three different keepers — the director of the Bureau, the president of the Committee, the Archives of France — so that no man alone, ever, could find himself before the metre and the kilogram of the world.

Whoever has read the first chapter recognizes the scene with his eyes closed: the standard in the temple, the priests of the keys, the weighing as liturgy. The positivist nineteenth century, which believed itself cured of the sacred, had simply secularized it: at the foot of the hill of Sèvres they fired porcelain, at the top they kept relics — of platinum instead of bone, but with the identical apparatus of seclusion, hierarchy, and rite. The difference, a substantial one, lay elsewhere: this relic worked, and anyone, paying the journey, could verify it.

And the successor, meanwhile, was already in the antechamber. From the seventies the American philosopher and scientist Charles Sanders Peirce, a metrologist by profession for the geodetic survey of the United States, had been the first to try to anchor the metre to a wavelength of light — a standard that no vault need keep, because every atom in the universe possesses a copy of it. The technique matured in the hands of Albert Michelson, the man of the interferometers, who in 1893 was invited to Sèvres itself, alongside the metrologists of the Bureau, to count how many wavelengths of the red light of cadmium stood in the newly consecrated metre: a little over a million and a half, with a reproducibility that already rivalled platinum. The international prototype was four years old, and in its own laboratory was being tested the thing that would one day pension it off; the formal passage would come only in 1960, but from that sojourn of Michelson’s the sentence was written — and the metrologists of Sèvres, it must be said to their honour, were the first to write it, carefully measuring the rope meant to hang them [Quinn 2011].

Precision enters the factory

While the diplomats were signing, in the workshops a parallel revolution was unfolding that was at once the deep cause and the beneficiary of the Convention: precision was changing its trade. For all the history told so far, to measure well had been a talent — of the craftsman who adjusts the piece until it “fits,” of the sworn measurer, of the steady-handed astronomer. Nineteenth-century industry needed the contrary: that quality should depend on no particular talent, because the pieces had to be born equal by the thousand, in different factories, from hands as interchangeable as the pieces themselves.

The prophet of this faith was Joseph Whitworth, the maker of machine tools of Manchester who as early as 1841 proposed — and imposed by the weight of his own catalogue — the first standardized screw thread in history: angle and pitches unified for all the screws of the empire, so that the bolt of Glasgow might find its nut at Calcutta. Whitworth preached that measure, in the workshop, must replace touch: his surface plates, his gauges, his machine able to appreciate the millionth of an inch shifted exactness from the operator’s wrist to the equipment of the department. The symbolic gesture of this transition is the go/no-go gauge: two mute jaws that say yes or no, and with which a labourer with no notion of metrology exercises, a thousand times a day, a judgement of precision. It is not a technical detail: it is the democratization of measure, and at the same time its depersonalization — judgement transferred from man to instrument, along a trajectory this book has followed from the first chapter and which here takes a decisive step [Crease 2011].

Behind the gauge stands an idea that deserves a moment’s light, for it is a small epistemological revolution disguised as workshop practice: tolerance. Nineteenth-century technical drawing ceased to prescribe to pieces a dimension and took to prescribing an interval — the diameter between this minimum and this maximum, and everything that falls within is good. It is the formal surrender to a truth the craftsmen had always known and never written: no piece is exact, none ever will be, and absolute exactness is not even desirable, because it costs more than it yields. Tolerance is institutionalized imperfection — quantified, contractualized, printed on the drawing beside the dimension — and it marks the moment when industry made peace with error a good half century before science, as we shall see two chapters on, found the words for the same peace. When today a supply contract specifies “plus or minus five hundredths,” it repeats the gesture of the level and heaped bushel of the first chapter: to define, along with the measure, the deviation that justice admits.

On the other shore of the Atlantic the same idea took the name of “American system”: interchangeable parts, worked on dedicated machines and verified on gauge-standards, assemblable without fitting. The legend has it born in a coup de théâtre of Eli Whitney, who in 1801 supposedly assembled muskets before Congress, picking the parts at random from separate boxes; historiography has dismantled the scene — the parts had been secretly pre-fitted, and true interchangeability matured only decades later, in the federal arsenals, by dint of gauges, inspectors, and factory discipline. It is worth holding the myth and its dismantling together, for the point survives the theatre: interchangeability was not an invention but a regime — a system of standards, tolerances, and controls that cost more than the labour it replaced, and which imposed itself because it turned the spare part from an artisanal problem into a catalogue commodity. In 1864 William Sellers gave the United States their unified thread, brisker and more tolerant than the English one; railways, navy, and arsenals adopted it, and the North American continent screwed itself onto a standard different from that of the British empire — a ripple of a phenomenon the modern reader knows well: rival standards war for decades, and the war is won by the catalogues, not the arguments.

The bridge between the factory and Sèvres was built, at the end of the century, by a Swedish armourer. Carl Edvard Johansson, an inspector in a rifle factory, conceived in 1896 the most elegant multiplier of precision ever invented: the gauge blocks — parallelepipeds of hardened steel, ground and lapped with faces so perfectly flat and parallel that, brought together with a light twist, they adhere to one another as if by magnetism, held by molecular forces alone. A set of a hundred or so blocks, scaled with arithmetical cunning, combines into stacks that materialize tens of thousands of different lengths to the micron: the most remote workshop could now possess, in a wooden box, the practical equivalent of a metrological laboratory. The “Jo blocks” conquered the factories of the world — Henry Ford, who had built the Model T on interchangeability, ended by buying Johansson’s company outright — and closed the circuit this chapter recounts: the blocks are calibrated against the standards of the national institute, the institute against the prototype of Sèvres, and so the chain of trust reaches, step by documented step, from the vault of the three keys to the lathe of the provinces. The pyramid had found its base [Crease 2011].

It was this world — not the philosophers, not the revolutionaries — that made the Metre Convention inevitable. When the machines of one nation assemble parts made in another, when boilers explode for wrong tolerances and international contracts are written in two units, metrological fragmentation ceases to be picturesque and becomes a system cost, accountable to the penny. The first chapter showed measure as an institution of the city market; the nineteenth century discovered it an institution of the world market, and treated it accordingly: with a treaty.

States become laboratories

The last quarter of the century added the piece that was missing: the laboratory state. The model was invented by a newly unified Germany, which in 1887 opened at Charlottenburg the Physikalisch-Technische Reichsanstalt — the Imperial Physical-Technical Institute — on land donated by Werner von Siemens and under the presidency of Hermann von Helmholtz: the greatest physicist in the country directing a body whose mandate was, without shame, to put precision measurement at the service of national industry. It was a success so great as to become at once a geopolitical model: England answered in 1900 with the National Physical Laboratory at Teddington, the United States in 1901 with the National Bureau of Standards — the ancestor of today’s NIST, born moreover with a mandate that held together the top and the bottom of the pyramid: in its early years the American office made itself known as much for precision physics as for campaigns against rigged scales and the hollowed weights of grocers, Hammurabi and Helmholtz under the same roof — and within a generation every industrial power had its citadel of metrology, in an emulation the contemporaries called openly by its name: he who certifies better, exports more. The rivalry between empires, which in the previous chapter ran along the meridians, now ran along the certificates of calibration: to measure better than the neighbour had become a national competitive advantage, quantifiable in orders [Quinn 2011].

And here the history of measure delivers to the history of physics one of its most spectacular gifts. Among the industrial tasks of the Reichsanstalt was photometry: the German gas industry and the nascent light bulb demanded a reliable standard of luminous intensity, and the high road passed through the black body, the perfect radiator whose light depends on temperature alone. The physicists of the institute spent the nineties measuring its spectrum with unprecedented precision — not out of cosmic curiosity, but to sell certified photometries — and it was precisely those curves, traced at Charlottenburg and obstinately rebellious to every classical formula, that forced Max Planck in 1900 to the act of desperation from which the quantum was born. Quantum physics, which in the seventh chapter we shall see become the foundation of the units, was born as a by-product of an industrial calibration: metrology does not merely serve science — now and then it generates it.

Beneath the institutes, meanwhile, the nineteenth-century state was rebuilding also the most ancient step of the pyramid: the inspector. The national metric laws everywhere instituted periodic verification — the obligation for every shop scale, every market steelyard, every pump and meter to undergo control and stamping at fixed terms — and with it the corps of metric officers, itinerant functionaries with a kit of calibrated standards and a punch. The reader has met them many times under other names: they are the metronomoi of the agora, the aediles of the forum, the muhtasib of the souk, reappeared in frock coats with a calibration certificate in their pocket. The pyramid of trust, however high it rose toward Sèvres, went on resting where it had always rested: on a functionary who enters a shop and looks at the pan of the balance.

Italy, in this landscape, played the part of the early and silent convert. Piedmont had imposed the metric system by law as early as 1845, and unification extended it to the peninsula along with conscription and the civil code: to translate into the new language the thousands of local units — the arms, the bushels, the canes of every province — the state published monumental tables of conversion, commune by commune, which remain today one of the finest photographs ever taken of the premetric world, the notarial inventory of the Babel the first chapter recounted. As for laboratories, the country long made do with its metric offices and came late to giving itself institutes of international rank: to be metric by law and to be so by laboratories are, here too, two different stages of the same road.

As for the United States, their story deserves the small untying of a misunderstanding that has lasted a century and a half. America signed the Convention in 1875, among the first; drew its prototypes in the urn of 1889; and in 1893, by an administrative act known as the Mendenhall Order, established that the American yard and pound be defined as fractions of the metre and the kilogram — not the reverse. For a hundred and thirty years, therefore, the United States have been a country metrologically metric to the marrow: the American inch is, by law and since 1959 in exact form, twenty-five point four millimetres, and no measure is made in America that does not descend, by an unbroken chain of calibrations, from the international standards. What America has never done is adopt the metre in daily life: its roads stay in miles, its recipes in ounces. Accession and adoption are two different things — the one a question of foundations, the other of façade — and the lesson holds in general: the success of the Convention is not measured by which units appear on the road signs, but by the fact that every inch on the planet is by now, under the hood, a multiple of the metre. The winning standard is not the one everyone sees: it is the one on which everyone, even without knowing it, leans.

The invisible infrastructure

The system that issued from 1875 worked, and worked in the way successful infrastructures work: by disappearing. For over a century the national standards returned periodically to Sèvres to compare themselves with the international prototypes, in verification campaigns lasting years and documented to the last microgram and the last tenth of a micron; the results — the minute deviations, the suspect drifts — circulated in the reports of the Bureau, feeding that accounting of trust that no end user would ever read and from which every end user, buying a litre of milk or a drug dosed in milligrams, benefited without knowing. It is the conceptual graft of this chapter, and it is worth stating in full light: modern precision is a distributed trust. No human being can verify in person the chain that binds his dressmaker’s tape to the platinum of Sèvres; each trusts a system of institutions, protocols, and cross-comparisons that no one masters entirely and that all together maintain. The walled arm of the first chapter asked the citizen to trust his own commune; the twentieth century asks him to trust a planetary pyramid of laboratories he will never see. Modernity has not abolished the metrological act of faith: it has organized it, subdivided it, and certified it — and in so doing has made it invisible [Quinn 2011].

The most illuminating parallel is the one with money, which is a fixed guest of this book from the first chapter. The nineteenth century anchored currencies to gold exactly as it anchored measures to platinum: a metal in a vault, convertibility guaranteed, trust deposited in an object. And the twentieth century dematerialized both anchorings for the same reason — the object, however noble, had become the weak point of the system — replacing them with networks of institutions that watch over one another: central banks on one side, metrological institutes on the other. Fiduciary money and fiduciary measure are conceptual twins: both are worth something because a credible architecture of keepers promises that they will be worth something tomorrow, and both show that the maturity of a convention coincides with the moment it can afford to abandon its material totem. The metre would abandon its in 1960, the kilogram — the most conservative of the family — only in 2019: but the architecture that made that farewell thinkable is all in this chapter, in the treaty that turned an object into a system.

Invisible, until it breaks. In 1983 a Boeing 767 of Air Canada ran out of fuel halfway through its route — it glided, by the legendary luck of its crew, onto a disused airstrip at Gimli — because the refuelling had been calculated confusing kilograms and pounds in the country’s switch to the metric system. In 1999 the Mars Climate Orbiter probe, over a hundred million dollars of hardware, disintegrated in the atmosphere of Mars because a supplier had delivered thrusts in pound-force to a software that expected them in newtons. These are the accidents the manuals cite as warnings, and they say something deeper than the distraction of a technician: they say what happens when the invisible infrastructure presents the bill — when the promise between strangers, for once, is not kept. The surprise we feel before these disasters is the exact measure of how much, all the other times, the system works.

With the mature twentieth century, the pyramid changed shape again. The family of signatories grew from the seventeen of 1875 to some sixty member states today; the mandate of the Bureau, born for length and mass, was extended in 1921 to the electrical and photometric magnitudes — the formal recognition that the metrology of the invisible, whose birth the next chapter recounts, was by now metrology in full right; and to the comparisons with the prototype were added comparisons among national institutes, until in 1999 a mutual recognition arrangement turned bilateral trust into a multilateral network of certified equivalences — the architecture on which every calibration certificate on the planet rests today, and whose operative mechanics, made of key comparisons and tables of equivalence, belong to the manual more than to history: the professional reader will find it documented in the entries of the glossary of ingegnerismo.it [ingegnerismo.it] . Here it is the design that interests us: in the span of three generations, measure had passed from the walled standard of a city to a treaty among empires, and from the treaty to a network of laboratories that watch over one another — ever more precise, ever more anonymous, ever more like what it truly is: the nervous system of technical civilization.

But while the mechanics were unifying screws and gauges and the diplomats were drawing lots for bars of platinum, physics was opening a front that no material standard could ever garrison. One can wall in the arm, the metre, and even the kilogram: but where does one wall a degree of temperature? And what does one deliver to the urn, when the unit to be distributed is that of electric current? The next chapter recounts how one learns to measure what cannot be touched — and how a cable laid on the floor of the Atlantic forced the nineteenth century to invent, for the invisible, the whole of metrology over again.