Four Thousandths Proud

The plate bearing the number 411 came back from the divider last Tuesday with its graduations cut at a hundredth of an inch, and I set it on the bench beside the theodolite to check the run of the teeth against the master. Under the lamp the divisions looked true. Under the micrometer they did not. The twelfth graduation stood four thousandths proud of its neighbours, and the twenty-third stood three thousandths shy, and between them the scale read a curve that no eye could see and every measurement would inherit. The eye learns to distrust itself in this way: not by failing, but by succeeding too well at the wrong question. It asks whether the line is straight, and the line is straight enough. It does not ask whether the man who cut it was tired at the twelfth stroke, and the answer to that is yes, he was, and the plate has carried the tiredness ever since.

The student will ask how we know the error is in the plate and not in the micrometer. We know because we measured the micrometer against the gauge blocks first, and the blocks against the standard, and the standard against the metre at Greenwich, and at each remove the error was smaller than the one we were chasing. The instrument is not the fault. The instrument is the only reason we can see the fault at all. This is the difficulty with trusting the eye: it must be calibrated before it can be distrusted, and the calibration is itself an act of the eye.

I will re-cut the plate on Thursday, and I will cut it in the morning, and I will stop at the eleventh graduation whether or not the work is going well. Tell Pemberton the twelfth division on the old plate is the one to keep - it is the only honest mark on the thing.