The Slow Decay of Certainty

Notebook XII, page forty-three. The electrometer reading this morning: 3.8 × 10⁻¹⁰ amperes. Yesterday it was 4.1. The day before, 4.0. The variation is within calibration error, and yet I have written down every figure, because the record is the experiment. Without these numbers, the three mornings do not exist scientifically. They are private experience only.

This is the thing that took me years to understand fully, and that I still have to remind myself of when the work goes slowly. The phenomenon does not care whether I have measured it. The uranium in the pitchblende has been ionising air since before there were instruments to detect it, since before there were hands to build the instruments. What I add is the record — the methodology documented completely enough that someone else, starting from the same ore with the same electrometer under the same atmospheric conditions, would obtain the same figures. That reproducibility is not a formality. It is the difference between a private observation and a contribution to what survives us.

I am working from four tonnes of pitchblende residue. The radium is in there, somewhere below the threshold of current detection, compounding its presence against the background radiation of everything else. Isolation before measurement — this is not obvious to everyone. The contamination in a badly isolated experiment is not a flaw you can argue away; it is physically present in the numbers. The only remedy is protocol: controlled conditions, calibrated instruments, complete records.

There will be a gram of radium at the end of this. The notebooks documenting its extraction will outlast me, though no one knows that yet. They will be consulted in the century after mine, sealed in lead, handled with instruments I never used, because the exposure accumulates in the cases where it is not immediately visible. The invariant must be maintained before the damage is observable, not after.

The electrometer reads 3.8. I write it down.