The reading that could not be trusted

An original account by an invented weather observer.

At first I entered zero rainfall. The collecting cylinder was empty, and the water level had not reached the first line on its scale. As I replaced the cover, however, I noticed a curled leaf wedged tightly across the funnel.

The empty cylinder could no longer give me a certain answer. Perhaps no rain had fallen, or perhaps water had been prevented from entering the gauge. I struck out zero rainfall and wrote no valid reading instead.

The correction looked untidy on the page, but it made the record more accurate. I was no longer claiming to know what the weather had done; I was recording what the faulty instrument could not tell me.