I traced the six-inch main with my finger. It ran from the well toward the pasture system, passed a series of stock-tank branches, and curved near boundary marker fourteen at the eastern edge of the ranch. The line stopped on Dad's map exactly where it should have stopped. Nothing crossed into the Dawson property. No notation showed a customer connection or any agreement to provide water to another landowner. Yet sixty-two houses had gone dry when I stopped a pump on my side of the boundary.
I returned to the pump house with a flashlight. The gauge settled at forty-one pounds per square inch, a disappointing reading for a system that used to operate close to fifty-eight under ordinary conditions. Pressure alone doesn't prove a theft; pumps age, demand changes, restrictions develop, and gauges can lie. I checked the instrument against a second gauge and compared readings at several ranch branches. The lower pressure was real, and it was more pronounced toward the section of line nearest Ridgecrest.
The next morning I called Daryl Voss, an independent hydrological engineer from San Marcos. Daryl and I had worked a municipal leak investigation years earlier, after a town discovered that its expensive new reservoir was supplying considerably more water than its customers were being billed for. He had a gift for interpreting instruments without allowing them to become substitutes for common sense. He was fifty-five, sun-weathered, laconic, and interested in difficult problems the way other people are interested in fishing.
"I've got a private well and a neighbor who loses water whenever I turn it off," I said. "I think someone tied into an old ranch main without authorization."
"You think," he repeated. "Or you know?"
"I know what the pressure is. I know what happened at the subdivision. I haven't found the physical connection."
"Good," he said. "We start with what you know. I'll bring the instruments Thursday. And Cole? Don't dig up the entire county before I get there. I'd prefer to know what we're looking for before we go hunting."
Daryl arrived in a dented white truck loaded with pressure loggers, flow-measuring equipment, acoustic sensors, and enough notebooks to make my father proud. We spent the first hour calibrating instruments because a bad reading can send a good investigation in exactly the wrong direction. Then we walked the line from the wellhead toward the boundary, checking branch points and recording how pressure behaved under controlled demand. The work was slow and technical, but it had a certain elegance. Pipes don't know how to lie; people are the ones who mislabel what pipes are doing.
At the wellhead, the outgoing system held close to its expected pressure. The first and second agricultural branches showed ordinary losses over distance. Farther along, the behavior changed. Near boundary marker fourteen, there was a sharper decline than the configuration of Dad's map could explain, especially when we isolated the ranch's documented branches for brief controlled tests. Daryl stopped beside the marker, crouched in the brittle grass, and looked toward the new houses beyond the fence.
"Something's pulling water here," he said. "And unless you've got a herd of cattle living underground, it isn't the stock tanks."