Eleven Seconds of Arc

Observed: the 20-foot reflector at Slough, the night of 12 April 1830, Saturn’s sixth satellite, conditions steady, no moon, the mirror recently re-polished at the edge where the figure had gone soft. I had the position from my own sweep of the 10th, and the 10th’s position disagreed with the 8th by eleven seconds of arc. Eleven seconds is not seeing. I checked the micrometer screw; it ran true. I checked the clock; it lost nothing. The error was in the reduction - I had carried a wrong index into the second column and never caught it, and the wrong figure had the virtue of being consistent with itself.

That is the matter I keep turning. The false position from the 8th was honest. It announced its provenance. Anyone could take my numbers and rebuild the chain and find the broken link. The corrected figure, which I entered on the 13th, is cleaner and less honest, because it agrees with the 10th and the agreement is now partly an artefact of my wanting them to agree. Two independent sightings that I have since made to rhyme. The catalogue cannot show this. The catalogue shows three positions in a tidy row and no trace of the hand that pushed the third into line.

The same fault runs through the sweep log for the nebulae. Where I have re-observed and found a former place in error, the amended place reads as though it had always been known. The erasure is the dishonesty. The original blunder, left standing with a note, would have been the truer record.

I have not resolved this. The mirror is still in the tube. The 8th’s entry stands uncorrected in the ledger, and I have not yet decided whether to strike it.