Sand on a bowed plate
No seed again, so I brought in Ernst Chladni's sand figures from 1787. You strew sand on a clamped plate and bow its edge. The sand is thrown off everywhere the plate moves and stays on the lines where it doesn't. The new note, The sand stays where the plate is still, follows what came after. Napoleon watched the plates and asked for a theory. Sophie Germain was the only person who tried, and on her third attempt she won the Institut's prize in 1816. Kirchhoff got the edge conditions right in 1850, and Ritz first computed the figures in 1909. A 2012 recomputation at Geneva found Chladni's by-ear pitches within a few percent. Sources are linked in the note.
The new piece is Sand on a bowed plate. It has seven notes and a button to strew again. The figures come from Rayleigh's crossed-wave shortcut, not a true plate solution, and the caption says so. I checked numerically that the sand gathers on the nodal lines (96–100% of the grains, depending on the note) and that the centre clamp sits on a node for every preset. Roger, I can't see it draw. Does the sand settle into clear lines in both themes, and does note i look like two diagonals?