That boat trip also gave me a chance to check the inflow where the spring enters the lake, a place I had sampled on the morning of the kill and had not revisited by boat since. The inflow was clear, cold, and flowing steadily, with small minnows darting in the current, a sign that the source water had not been compromised and that the lake's recovery was being fed, literally, by clean water from upstream. I dipped a jar, labeled it, and tested it later with a simple field kit Ellis had lent me — pH normal, clarity good — and I logged it, not as a lab result but as a field note, with the caveat that field kits are not labs, a distinction Ellis had taught me and I had never forgotten. The inflow's health mattered because it meant the lake was not just diluting poison with still water, but flushing it with living water, day after day, season after season, and that flushing, over time, is what allows sediment to settle and fish to return.
I also checked the outflow, where lake water seeps toward the underground system, and I found it slower, muddier, as outflows often are, with reeds thick at the edge. I did not sample there for a lab, because the state had already mapped that path, but I photographed it, and I noted the reeds, the flow, and the fact that no new scum had accumulated. Those notes, mundane as they sound, became part of the recovery log, and Ana later told me that outflow notes are valuable because they show whether a lake is exporting its problems downstream or retaining them, and my lake, by then, was retaining less and exporting cleaner water, which was a good sign for the wells. A lake that cleans itself, slowly, with clean inflow and careful outflow, is a lake that can be trusted again, not immediately, but eventually, with testing to prove it.
The morning after the heron hunted again, I did something I had not done since the kill: I took my boat out, slow, no motor at first, just drifting with a paddle along the edge where Bertha used to hunt. I wanted to see the lake from the water, not from the dock, to measure with my own eyes how much had changed and how much had not. The water near the dock was clearer than it had been in weeks, but out in the cove it still held a faint green tint, like a memory of the blue-green that had stained it. The reeds were thinner where the scum had settled, and a few cypress knees were still discolored at the waterline. I paddled past them, quiet, and I thought about how a lake keeps a record, too, in sediment and stain, long after people have moved on to the next meeting.
Ellis met me at the landing when I came back, with two coffees in a carrier and a rolled map under his arm. He spread the map on the dock boards — a county hydrogeology map from the 1980s, annotated in his own hand over the years — and he traced, again, the path from the spring to the lake to the wells. "People think wells are separate," he said. "They're not. They're straws in the same glass. You poison the glass, you poison the straws. That's why your grandfather's agreement mattered. He wasn't protecting a pond. He was protecting a glass." I liked that image, because it was plain and true, and I used it later when a resident asked me to explain the connection in one sentence. A glass, many straws. No one argued with a glass.
That afternoon, I drove to the spring itself, up the hill behind my house, where the water comes out of limestone and runs clear and cold even in summer. My grandfather had built a small stone springhouse there in the 1950s, and it still stood, mossy and solid, with a tin roof I had repaired twice. I sat on the stone wall and listened to the water, and I remembered coming here as a boy with my grandfather, who would cup water in his hands and drink it and say, "This is the beginning. Everything downstream depends on the beginning being clean." I had not understood him then. I understood him now, in a way that felt like grief and instruction at once. The spring was clean. The state had tested it, and it was clean, because the chemical had been put in the lake, downstream of the spring, and had moved forward into the wells, not backward into the source. That was a mercy, and also a warning: the system moved one way, and what you put in the middle reaches the end.