The periodic table is a law, not a filing cabinet — and the proof is that it described elements that did not yet exist, and was right. This is that story, operated rather than read: Mendeleev's 1871 bets scored against the elements that later filled his gaps, and — further back — Newlands' 1865 near-miss rebuilt on the instrument. Each one shown honestly: why it looked real, and exactly where it broke.
The moment. For most of history, chemists knew barely a dozen elements; by the 1860s the count had blown past sixty, as new methods — electrolysis, then spectroscopy (you can watch this in section 03) — pried them loose faster than anyone could organize them. Then in 1860 the Karlsruhe Congress finally settled a consistent set of atomic weights, and that lit the fuse: within a decade a handful of people, working independently, reached for that one newly-reliable number to find the hidden order — a French geologist (de Chancourtois, 1862), a London sugar-refinery chemist (John Newlands, 1865), and two chemists who landed on it within a year of each other (Mendeleev, 1869; Lothar Meyer, 1870). Weight was the best ruler they had. This page is what each of them saw with it — and, at the heavy elements whose weights were still wrong, where it failed them. The fix came in 1913, when a 26-year-old physicist, Henry Moseley, swapped weight for atomic number.
In 1869–71 Mendeleev left holes in his table and, for some of them, wrote down the atomic weight, density, and oxide of whatever would fill them — years before anyone found it. Here is the most famous case: two gaps under aluminium and silicon.
Six years earlier, John Newlands lined the known elements up by atomic weight and heard a tune: properties seemed to repeat every eighth element, a "law of octaves." He was ridiculed. Rebuilt on the wave — the 63 elements known in 1865, height = atomic volume (Meyer's 1870 curve), color = family — you can see he was right that the pattern is real, and see exactly where it broke.
Now color the board not by chemistry but by how each element was first caught. Press play and you're not watching chemistry — you're watching human capability. Each new method opens a whole region at once: Davy's electrolysis cracks the left of the table in 1807–08, then spectroscopy, radioactivity, and finally the accelerator, making elements that were never on Earth.
Method buckets curated from the sourced discoverer/era data; each element lights on the date it was first obtained by its crediting method — electrolysis on isolation (so Davy's cluster bursts at 1807–08), spectroscopy on identification. "Chemical isolation" is the large neutral background the dramatic methods stand out against. Long-tail dates are curated-approximate; the waves are the point.
Seven years before Mendeleev, Alexandre-Émile de Chancourtois wound the elements up a cylinder by atomic weight — one full turn every 16 units (oxygen's weight) — so elements 16 apart stack in a vertical column. His vis tellurique was the first real periodic system. The journal published it without the diagram, and the idea sank for a decade. Here it is rebuilt — the cylinder unrolled, each element at its weight (down) and its wrap-position (across).
Positions use modern atomic weights except indium and uranium, shown at their 1865 values (de Chancourtois worked in period weights that differ for the heavies) — but the decoherence is robust to that: the column breaks because heavy elements aren't 16-spaced in weight, true under either ruler.