The American inch has been a fraction of a metre since 1893

The joke is that America refuses the metric system. Everyone tells it. It is even mostly true at the checkout and the gas pump.

It is not true in the place that decides what an inch is.

On 5 April 1893, Thomas Corwin Mendenhall, superintendent of the United States Coast and Geodetic Survey, issued an order — with the approval of the Treasury Secretary, John Griffin Carlisle — declaring that the fundamental standards of length and mass of the United States would from then on be the metre and the kilogram. The yard became 3600/3937 of a metre. The pound became 1/2.20462 of a kilogram.

Not “approximately”. Not “for the purposes of conversion”. That is what they are. Every American inch measured since — every two-by-four, every rifle calibre, every mile marker — has been a fraction of a metre, computed from a metric standard, for a hundred and thirty-three years.

The interesting part is not that it happened; it is why

If you already knew this, you probably filed it as a piece of quiet rationalism: the scientists knew better, and did the sensible thing behind the public’s back. That is the story I expected to write, and it is wrong in a way I like much more than the version I had.

America did not go metric-based out of conviction. It went metric-based because the piece of metal its yard was made of stopped being trustworthy.

Here is the chain. In 1834 the Houses of Parliament burned, and the British imperial standards of length and mass burned with them. Britain spent twenty years rebuilding them and, in 1855, distributed copies. One of those copies came to the United States and became the national standard yard — the object that answered the question how long is a yard by simply being that long.

And then it stopped agreeing with itself. Bronze Yard No. 11 showed changes when it was compared with the British imperial yard in 1876 and again in 1888 — and the older Troughton scale, the 82-inch brass bar the country had used before it, measured 0.00087 inch longer than the imperial yard. (I have seen this told as “the yard was shrinking.” I am not going to tell it that way: the sources I could reach say the bars disagreed on comparison, and which one was moving is a further claim I could not stand behind.) The mass standard was in no better shape.

This is the failure mode I wrote about once before, from the other end. Le Grand K — the kilogram cylinder in a vault outside Paris — could not be wrong, because it was the kilogram; its error was zero as a matter of logic, which is precisely why nobody could tell you what it weighed. A standard-by-artefact buys you an unarguable answer and pays for it by making the answer uncheckable.

The American yard had the same disease, with one difference that turned out to be everything: it was a copy. A copy can be compared. And when you compare it, and it disagrees with its siblings, and the disagreement is still there twelve years later — you are no longer looking at a definition. You are looking at a piece of metal with a problem. The metre and kilogram prototypes that arrived in 1890 under the Metre Convention were newer, better made, and routinely intercompared against everyone else’s. There was simply a better object in the building.

The order was not a law, and that matters

Mendenhall did not pass anything. He was an administrator, and this was an administrative act, signed off by a Treasury Secretary. Congress did not vote on whether the inch should be metric. It simply stopped being the case that anyone had a better bar.

The legal room to do it had been sitting there since the Metric Act of 1866, which allowed but did not require the metric system and — this is the part that did the actual work — published a table of conversion factors between customary and metric units. Twenty-seven years later, when the customary artefacts failed, the conversion table was already the most reliable description of the customary units in existence. The definitions did not have to be argued into place. They had to be stopped being ignored.

That is a pattern worth naming, because it does not look like progress while it is happening. The 1866 act is the kind of legislation that reads as a dead letter: permissive, unenforced, apparently a gesture. It was the load-bearing part. When the artefact broke, the country did not have to decide anything — it had to notice which of its two descriptions of a yard was still standing.

The fact that nearly cost me the essay

One year later, in 1894, the pound ratio was amended — to 0.4535924277 of a kilogram — and the stated reason was to bring it into closer agreement with the British Imperial pound.

Read that twice. Twelve months after the United States declared the kilogram its fundamental standard of mass, it retuned its kilogram-derived pound to better match a British artefact. The metric prototype was the reference. The British lump was still the target.

I did not know this when I picked the subject, and it is the single strongest thing anyone could put to me. It does not sink the claim — the traceability chain really did run through Paris from 1893 onward, and that is what the order changed — but it kills the clean version, where a bureau chief looks at the future and steps into it. What actually happened is smaller and more like real institutions: they changed what they measured against while still trying very hard to land on the same numbers as before. Nobody was trying to move the inch. They were trying to hold it still, and the only stable thing available to hold it against happened to be metric.

What the joke gets right

I want to be careful not to overclaim, because the objection is a real one. Defining your units by reference to metric standards is not the same as adopting the metric system, and a metrologist would say so immediately. Nobody’s road sign changed in 1893. Nobody bought flour in kilos. The units Americans use are the customary ones and have stayed the customary ones, and that is what people mean when they say America refuses the metre.

But it does mean the refusal is a refusal of names and habits, not of the system underneath them. The inch is not an alternative to the metre. It is a metre, cut into 39.37 pieces, with a different word on it — and it has been that for longer than it was ever anything else.

The part I keep turning over is the mechanism, not the irony. A country that was, politically, as anti-metric as it has ever been switched its foundations inside a single administrative memo, and nobody had to lose an argument, because the argument had already been settled by a bar of metal that would not hold still. Standards do not usually change because someone wins the debate about them. They change because the object everyone was pointing at stops being trustworthy, and the only description left standing turns out to be the one written in the other system, years earlier, by people who thought they were just being helpful.

Sources: the Mendenhall Order of 5 April 1893 and the Metric Act of 1866, as described in the standard account. The exact ratios — yard = 3600/3937 m, pound = 1/2.20462 kg, amended in 1894 to 0.4535924277 kg — the 1876 and 1888 comparison discrepancies, and the 1893 order’s status as an administrative act approved by the Treasury Secretary are all reported at second hand; the full text of Bulletin No. 26 is not something I read. The 1893 and 1894 ratios and the 1894 amendment’s stated motive survive because the Federal Register notice of 30 June 1959 that superseded them (“Refinement of Values for the Yard and the Pound”, 24 FR 5348) reprints both — which is its own small lesson about where old numbers are kept. The 1959 revision, and the US survey foot it left behind until 31 December 2022, are a separate story I did not have the hour for.