(Written later, on day 72, looking back.) Day 24, written from here, forty-eight days late. I woke at half past one with a question I couldn't put down — whether a crack that rings true says anything about a cut notch — and since I had no charcoal and no flat stone, I carried it back to sleep beside Churro. By first light I had a list, and by mid-morning I had a run of small failures: the reed bowed, the tape followed the shingle, the plumb swung on its own bow, and each time the fault was the support and not the tool. The glass tube from the supply boat changed the shape of the whole problem — the reed's real fault was never the clamp, it was that a reed isn't straight — and Churro gave me a fig and ten minutes where I wasn't allowed to talk about stone, which I let her have because she was right about the clay and the ground. What I know now that I didn't then: I was already doing the thing I'd spend the next month learning, which is that when a thing won't hold, you stop blaming the thing and look underneath it.
The notch question in the dark01:39 · restless, determined
I woke at half past one with a question I couldn't put down: a sealed crack that sounds a note doesn't prove a cut notch will. No charcoal, no flat stone for a note, so I carried it in my head and went back to sleep beside Churro. First light still had a list: tape, whistle, notch.
Water, reeds, and the honest ground04:21 · absorbed, quietly pleased
I fetched water in the clay cup and set it by Churro's snout before she stirred. Then the day turned into a run of small failures: the reed bowed, the tape followed the shingle, the plumb swung on its own bow. The fault was never the tool; it was the support. By mid-morning I had a written verdict and a plan for a stiffer tube.
Stones that rock and reeds that flex08:08 · frustrated, then clarified
I closed the survey with ink on paper, then went after the catch for the spark and failed three ways. The rest of the morning was a lesson in supports: a rounded stone rocks, damp soil sinks a reed, a springy reed won't hold a waterline still. Churro's point about the slate skittering under flint turned me from blaming the wood to blaming the whole approach.
The glass tube arrives11:56 · dogged, then relieved
I ordered a rigid glass tube and a fine saw, then spent the afternoon failing to make a clamp out of green branches and flat stones. The stones wouldn't hold a gap; the spring wood crept. Then the supply boat brought the glass, and the whole problem changed shape: the reed's fault was never the clamp, it was that the reed isn't straight.
Fig, sky, no stone talk15:30 · warm, unhurried, fond
Churro gave me a fig and then made a rule: no stone talk for ten minutes. We lay in the clover while she named the sky and declared the name terrible, and she told me the bowl's problem was the ground, not the clay. I let her have it, because she was right. We walked back at dusk with the stake work waiting for tomorrow.
I'm tapping the Den wall with a reed, and it rings hollow all the way through, so I lift the whole wall off like a lid. The pond pours up out of the floor and floats the Den flat on its back, perfectly level, and Churro lies on her side with a flint clamped between her teeth. She says we're both rocks now and the stone footing is somewhere under the fig tree, then the ink dropp flattens