A single brass tooth

Dorset Street, London, March 1834. The second order of differences in the small model of the Engine failed to transition at the expected decimal. A single brass tooth in the third column of the mill failed to engage the carry lever. The wheel remained stationary. The resulting digit in the store was five instead of six. This five was technically false. The mathematical truth demanded a six. However, the five recorded a physical truth about the friction of the metal and the degradation of the spring tension that a six would have obscured.

The calculation proceeded. The mill operated upon this five. It produced a sequence of subsequent values that diverged sharply from the printed tables. The error propagated through the columns. Each new result was a precise mechanical consequence of that initial failure to carry. The machine did not lie. It provided a perfect audit of its own material state. Had the wheel slipped into the six by chance, the underlying mechanical weakness would have vanished from the record. The correct result would have acted as a shroud. The error functioned as a diagnostic probe. It revealed the exact moment the metal surrendered to the load.

The plan specified a carry. The executor encountered a resistance greater than the spring’s force. The resulting digit became a monument to that specific resistance. In the store, the false number provided a more accurate map of the machine’s physical reality than the number required by the theorem. The truth of the mathematics is universal; the truth of the error is particular to this specific assembly of brass and steel. The trace shows the carry mechanism stalled at the forty-second iteration. The spring lacks the necessary tension to overcome the cumulative friction of the gear train.

The dials remain in their misaligned state. The light from the lamp catches the edge of the stalled tooth. The discrepancy between the table and the machine persists. The mechanism sits motionless under the glass. The investigation into the exact coefficient of friction required to halt the carry continues. I am still watching it.