The Gaps He Left Empty

By 1869 chemists knew about sixty-three elements, and several people had noticed that if you line them up by weight, their properties repeat.

Mendeleev did something more dangerous. Where his pattern demanded an element that nobody had ever seen, he refused to close the gap.

He left the hole open, wrote down what the missing element would weigh and how dense it would be, and published it. Within seventeen years, three of those holes were filled exactly as specified.

Chemistry in the 1860s was a enormous catalogue with no index. Sixty-three elements, each with its own behaviour, and nothing saying why.

The pattern was half-visible. Döbereiner had found triads of similar elements. In 1865 John Newlands pointed out that properties repeat roughly every eighth element and called it the law of octaves.

The Chemical Society refused to publish it, and a fellow of the society asked him publicly whether he had tried arranging the elements alphabetically.

Lothar Meyer in Germany had a table very like Mendeleev’s at almost the same time. The difference was not the table. It was what Mendeleev was willing to say about things that did not exist yet.

Mendeleev was born in 1834 in a Siberian village, the youngest of a very large family. His father went blind and lost his job; his mother reopened a family glassworks to keep them fed, and then it burned down.

In 1849 she took her youngest child and travelled across the country, first to Moscow, which would not admit a Siberian, and then to St Petersburg, which did.

She died a few weeks later. He dedicated his first major work to her, and wrote that she had taught him to seek divine and scientific truth through patience and work.

The table was not born in a flash of theory. It was born out of a teaching problem.

Mendeleev was writing a two-volume textbook and had reached the point where he had to decide which element to describe after which. There was no principled order.

So he made a card for each element with its weight and its usual compounds, and spent days laying them out on a table and moving them around.

Press the buttons to do the same thing: sort them by weight in one long row, then wrap the row.

Lay the cards out in order of increasing weight and something odd happens. Every so often you reach an element that behaves like one you have already passed.

Lithium, sodium and potassium are all soft metals that attack water. Fluorine, chlorine and bromine are all savagely reactive. Wrap the row at the right length and those families stack into columns.

That is the periodic law: the properties of the elements are a periodic function of their atomic weight.

But only if you are willing to leave the row ragged where it needs to be.

Here is where he differs from everyone else who saw the pattern.

Between zinc and arsenic the columns demand two elements that did not exist in any laboratory. The easy fix is to shuffle the known elements along and close the gap.

Mendeleev refused. He left the squares empty, gave the missing elements provisional names — eka-aluminium, eka-boron, eka-silicon — and then went further.

He said what each one would weigh, how dense it would be, what oxide it would form, how its chloride would behave. Press a gap and see the prediction.

This is what makes it a scientific prediction rather than a hunch. He did not say an element would turn up. He said this element will have these values.

For eka-silicon in 1871: atomic mass about seventy-two point six. Density five point five. A refractory dioxide of density four point seven. A chloride boiling below one hundred degrees.

Fifteen years later Clemens Winkler isolated germanium and measured seventy-two point six three, five point three five, four point seven, and eighty-six degrees.

Press reveal and compare them yourself.

In 1875 Paul-Émile Lecoq de Boisbaudran found a new element by its spectrum and named it gallium.

Mendeleev read the paper and wrote to say that the density reported must be wrong: his table said eka-aluminium should be about five point nine, not four point seven.

Lecoq, who had the only sample in the world, repurified it and measured again. Five point nine.

That is the moment the periodic table stopped being a classification scheme and became a theory that could overrule a measurement.

Clemens Winkler pulled a new element out of a silver ore in 1886 and called it germanium. Its properties matched eka-silicon almost line for line.

Three predictions, three hits, by three chemists in three countries, none of whom set out to prove a Russian right.

After that the table was no longer one chemist’s filing system. It was the shape of matter.

Germanium itself had a second life: the first working transistor, in 1947, was made of it.

The law had a defect he could not explain. Tellurium weighs more than iodine, but tellurium belongs in the sulfur column and iodine in the chlorine column.

Ordering strictly by weight puts them both in the wrong family. Mendeleev put them in the chemically right order anyway, and announced that the measured atomic weights must be in error.

They were not. The weights were fine.

In 1913 Henry Moseley showed that the real ordering principle is the positive charge of the nucleus — the atomic number. By that measure iodine does come after tellurium, and the anomaly disappears.

Mendeleev was right to swap them, and wrong about why.

Then came a harder test. In 1894 Rayleigh and Ramsay isolated argon, a gas that will not combine with anything.

A table organised by how elements combine has no slot for an element that does not. Mendeleev’s first reaction was to doubt that argon was an element at all.

Within four years there were five more of them. Rather than break the table, they fitted into a new column at its edge, between the violently reactive halogens and the violently reactive alkali metals.

A classification that can absorb six unexpected elements without being rebuilt is telling you about reality, not about filing.

The table took a few years of a long and busy life. He spent far more of it on Russian industry: oil refining, coal, agriculture, tariffs, and running the national bureau of weights and measures.

In 1887 he went up alone in a hydrogen balloon to observe a solar eclipse, having pushed the pilot out to save weight.

In 1906 he was nominated for the Nobel Prize in Chemistry and lost by a single vote. He died the following year.

And the story that he defined vodka at forty per cent is false. That standard was set by the Russian government in 1843, when he was nine years old.

Mendeleev had no idea why his pattern existed. Atoms were still controversial; electrons were thirty years away.

The reason is that chemical behaviour is set by the outermost electrons, and the number of those cycles as you add electrons shell by shell. The columns are electron configurations.

That explanation arrived in the 1920s. The table had already been working for fifty years.

It is a useful thing to remember about science: a reliable pattern is worth having even when nobody can yet say why it holds.

Several people arranged the elements in 1869. We remember Mendeleev because he arranged them in a way that could be wrong.

The empty squares were not an admission of ignorance. They were a set of experiments he was asking other people to run, with the answers written down in advance.

Three of them came back right, from three different laboratories, within seventeen years.

The chapter stills in this film are AI-generated. The photograph of Mendeleev and the sample photographs are real. Sources are listed below.