“Arlland Mast had been watching estate auctions in Kiowa County, Kansas for 31 years. And he would tell you himself that he had never seen a piece of iron draw less attention than the combine sitting in the southeast corner of the Decker yard on the morning of April 3rd, 1997.” Nobody bid on it. Nobody stood near it. The crowd gave it a look and kept going. The asking price was $0. What the crowd paid was $40. But one man walked straight to it and put his hand flat on the grain tank — not casual, not performative — listening. He’d come to that auction knowing exactly what he was doing. And what he found inside that abandoned machine would take him eight days to uncover and 290,000 reasons to keep quiet about.
“Arlland Mast had been watching estate auctions in Kiowa County, Kansas for 31 years. And he would tell you himself that he had never seen a piece of iron draw less attention than the combine sitting in the southeast corner of the Decker yard on the morning of April 3rd, 1997.” Nobody bid on it. Nobody stood near it. The crowd gave it a look and kept going. The asking price was $0. What the crowd paid was $40. But one man walked straight to it and put his hand flat on the grain tank — not casual, not performative — listening. He’d come to that auction knowing exactly what he was doing. And what he found inside that abandoned machine would take him eight days to uncover and 290,000 reasons to keep quiet about.
What Curtis had seen, what he had seen from 100 feet away before he’d put his hand on the tank, which was why he’d come to the auction in the first place, was a serial number.
A partial serial number visible through the grime on the right side of the cab, on the small rectangular plate that Case IH stamped before the machine left the Racine, Wisconsin assembly line. The number began with JJC.
He had seen that prefix once before — in 1993, when he’d been doing research on the 1660 platform for the rebuild he was working on and had come across a technical service bulletin that Case IH had issued in 1990 and then quietly withdrawn from circulation.
The bulletin described a limited production run of 1660 units manufactured between August and November of 1988. In which an experimental rotor cylinder design had been installed. A design that the engineering team had developed in response to persistent field complaints about rotor loss in tough green harvest conditions — specifically corn and soybeans with high moisture content in the straw.
The experimental rotor ran a modified concave clearance geometry and a different cylinder bar pitch than the standard production unit. The separation efficiency in the design simulations had been significantly better.
Case IH had installed the configuration in fewer than 80 machines and put them into the dealer network without documentation distinguishing them from standard production units. To gather real-world performance data without the liability complications that a formal field trial would have created.
The machines had entered the general fleet quietly. Farmers who bought them noted — some of them — that the machines seemed to handle tough conditions better than neighboring 1660s. But without documentation, without knowing what they had, most simply ran the machines until they wore out or broke down and traded them in without incident.
The serial number prefix JJC for the August-November 1988 window was the only distinguishing mark.
The reason Case IH had quietly withdrawn the service bulletin, in Curtis’s reading of the situation — he had not confirmed this, only reasoned it from what he knew about how equipment manufacturers manage internal test programs — was that the experimental design had performed better than the production unit in the field conditions it was built for. And acknowledging that publicly would have created complicated questions about why the company had discontinued a superior design and continued selling the inferior one.
Those were not questions a manufacturer wanted to answer in the liability environment of the early 1990s.
He was not certain his reading was right. He was certain enough to drive to Kiowa County on a Thursday afternoon, check the estate auction listing, and stand in the Decker yard on the morning of April 3rd with a yellow legal pad under his arm, waiting for the lot to come up.
The JJC prefix on the Decker machine serial number indicated the August-November 1988 production window. The machine was the right age. It was a 1660.
He had put his hand on the grain tank that morning — not to listen to it, but to get close enough to confirm the last four digits of the serial number. Which matched the range listed in the 1990 service bulletin he’d saved in a folder in his shop.
He paid $40 and drove home.
He called his cousin Lyall, who had a flatbed trailer and owed Curtis a favor from the previous fall. Went back to the Decker place the next morning with a chain binder and a come-along. Loaded the 1660 and brought it to his shed in Ford County.
The work took seven days. Not eight hours a day. Not every minute. He had a farm to run, and spring planting was two weeks out. But seven days of sustained, methodical work — the kind that Dale Halby had taught him was the only kind that didn’t produce mistakes.
Day one: the rotor drive belt. Standard component, available from the Case IH dealer in Dodge City. He drove in that morning, paid $84, drove back, had the new belt installed by early afternoon. He did not start the engine. He wanted the feeder house freed before he put any drive load on the rotor system.
Day two and most of day three: the feeder house. Seized solid. The upper and lower cross shafts locked from four years of sitting. He sprayed penetrating oil into every bearing housing and chain roller he could reach and let it sit overnight. The next morning, he applied heat from a propane torch to the outer bearing races. Slow and patient. Not trying to get it hot enough to damage anything, but hot enough to let the oil work. And then he put a two-foot breaker bar on the upper drive shaft and worked it back and forth. Not forcing. Not hammering. Just steady pressure and release. Steady pressure and release.
The way you might work a rusted bolt that you know is coming if you don’t rush it.
By mid-afternoon of day three, he had the feeder house turning by hand. Not smooth — the chains needed replacement, and two sprockets were badly worn — but turning. He ordered the chains and sprockets that evening from the parts catalog he kept in a three-ring binder on the shop shelf. The same binder he’d been adding to since 1985.
Day four: the cab glass. A cut-to-fit piece from a glass shop in Dodge City. $47. He drove in, waited while they cut it, drove back, had it in before supper.
Day five: the grain tank weld. This was the piece he had thought most about before he started. A crack in a grain tank is not structurally dangerous to the machine, but it will leak in harvest. And a weld on painted sheet metal made improperly will crack again at the heat-affected zone.
He ground the fracture clean with an angle grinder. Ground a V-groove along the full 18-inch length of the crack to give the weld material something to sit in. Preheated the panel with the propane torch for 20 minutes before he struck an arc. Preheat manages the thermal stress. Dale had explained this more than once.
Then he ran three passes with his Lincoln Electric MIG welder. Each pass short, letting the metal cool between runs. He ground the finished weld flush on both sides and repainted the area with a rattle can primer.
Day six: the engine. He pulled the valve covers first and checked the rocker arms by hand. Dry but not corroded. The machine had been stored in a closed shed, which had kept the moisture out of the top end.
He checked compression with a gauge he bought from JC Whitney by mail order in 1989 and still used. All six cylinders within spec. The lowest at 92% of the highest, which was good.
He replaced the fuel injectors. Four years of sitting is long enough that the needles may have corroded to their seats with a set he sourced from a dealer in Garden City for $312 for the six. He changed the engine oil and filter, the hydraulic fluid, the coolant. He set the valve clearances by hand with a feeler gauge, working slowly around the engine in firing order the way Dale had shown him, checking each one twice.
On the afternoon of the sixth day, he bled the fuel system and turned the key.
It ran rough for the first three minutes. One injector popping unevenly, then smoothing as the fuel worked through. He let it idle for 45 minutes, watching the gauges, checking for leaks at every joint he touched.
Nothing.
He shut it down and let it cool and went in the house for supper.
Day seven: the rotor drive adjustment and the header height control. The rotor drive chain had stretched, and he tensioned it to spec. The header height control was hunting — cycling up and down without settling. A sensor issue, not mechanical. He traced the circuit with a multimeter, found a corroded connector at the left feeder house sensor, cleaned it with electrical contact cleaner and a small wire brush, reconnected it, and tested the height control through its range.
It settled.
He did one more check of every fluid level, every belt tension, every connection he’d touched in seven days. The last walk-around that Dale had required of him before any machine left the shop. Slow, deliberate, no skipping.
On the morning of the eighth day — April 11th, 1997 — Curtis Halby drove the 1660 out of his machine shed under its own power.
He hooked up the 30-foot grain platform he kept for his own harvest. Drove to a strip of winter wheat stubble behind the shop. Made three passes.
The rotor performed. Not just functionally. It performed the way the service bulletin said it would.
And then something beyond that.
Or so he would say later to the people he eventually told. The separation efficiency in the tough green straw was unlike what he’d seen in a standard production 1660. He knew the production unit because he’d run one for six harvests. This was different in a way you felt before you measured it. In the sound of the rotor running clean. In the sample quality coming off the cleaning shoe. In the absence of the rotor slugging that the standard machine was prone to in high-moisture material.
He made the third pass. Stopped the machine. Climbed down.
He stood there for a while with his hands in his pockets. He was not a man who made speeches to himself or to anybody. He just stood there looking at the machine.
Then he went inside and called the Case IH regional service office in Wichita.
He did not lead with the experimental rotor. He led with the serial number and told the service representative that he had acquired a 1660 that he believed was part of a limited production run and asked if the company had any documentation on the unit.
There was a pause. He was asked to hold. He was put through to someone else. That someone else asked where he’d acquired the machine. He said in an estate sale in Kiowa County. That someone else asked for his phone number and said someone would call him back.
Someone called back in 45 minutes.
Two representatives from Case IH drove to Ford County on April 14th — three days after Curtis made the test passes. They brought a measuring kit for the rotor concave geometry and a laptop with the company’s internal service database. And a purchase offer that started at $140,000.
And after two hours in Curtis Halby’s machine shed, closed at $290,000.
They wanted the machine. They wanted Curtis’s documentation — the service bulletin he’d found in 1993, his notes on the rotor performance in the test passes, the serial number records. And they wanted a confidentiality agreement, which Curtis signed and which prevented him from discussing the specific nature of what he had found in the machine for a period of 10 years.
He did not discuss it specifically then or after. He confirmed the broad facts — the purchase, the price, the eight days — to a few people he trusted. Which is how the story traveled as far as it did. He did not explain the technical details in any form. The confidentiality agreement was clear on that, and Curtis Halby was not the kind of man who agreed to something and then worked around it.
Arlland Mast drove past the Decker place 11 days after the auction. He’d been delayed by planting, same as everyone in the county that April. And the combine was gone.
He stopped at the end of the lane and looked at the empty spot in the southeast corner of the yard where the 1660 had been sitting for years. And he felt the specific uncomfortable feeling of a man who has watched something happen without knowing what he was watching.
There was no mark in the ground where it had been. Just absence. The yard looked like any other farmyard.
He found out the outline of the story through the small county way that outlines of stories travel. A cousin of Curtis’s who worked at a grain elevator mentioned it to a farmer, who mentioned it at the co-op in Pratt about three months later.
He sat on what he heard for a while. He turned it over. He did not tell it at the coffee shop in Greensburg or at the elevator in Haviland because he was not sure he understood it well enough to tell it right. And telling a story wrong is a worse thing than not telling it. In a community where what you say about a person stays with them.
He told it to me eventually. On his porch in Kiowa County on an evening in late summer, with the wheat stubble still pale gold in the fields running east and the light going flat and long across the yard.
He sat in a lawn chair and held a coffee cup and looked at the distance while he talked. He said he walked up to that combine and put his hand on it and he already knew. Before he touched it, he already knew.
He was quiet for a while after that.
“I’ve been watching sales in this county for 31 years,” he said. “I never once looked at that machine. Not really looked.”
He didn’t say more than that. He had said enough.
That is the cost of categorizing things too fast. Not cruelty. Not ignorance. Just the human tendency to file things quickly and move on. To let the first look stand in for all the looks that might come after.
Dale Halby had spent 40 years teaching his nephew to slow down at the thing that looked done. To put his hand on it. To ask, “What does it tell you?”
The lesson cost nothing to learn.
It paid $290,000.
Curtis Halby is not a man who seeks attention. He still farms 600 acres in Ford County. His machine shed still looks like a graveyard to anyone who doesn’t know what they’re looking at. The confidentiality agreement expired years ago, but he doesn’t talk about it much. When asked, he’ll confirm the broad facts — the $40, the eight days, the $290,000 — and then he’ll change the subject.
But sometimes, if you catch him on the right evening, and if you ask the right question, he’ll tell you about his uncle Dale. About the phrase. About what it taught him.
“What does it tell you?”
He learned to ask it about seized engines and broken transmissions and rusted grain tanks. He learned to ask it about things that looked finished, things that everyone else had already categorized and filed and moved past.
And he learned that the things that look the most finished are sometimes the things that have the most to say.
Some nights, I think about that combine. Sitting in the Decker yard for four years. Everyone who walked past it. Everyone who glanced and kept going. Everyone who saw a broken machine and nothing else.
And I think about Curtis Halby, standing in that same yard, seeing a serial number through the grime. Putting his hand on the tank. Listening.
I think about the difference between looking and seeing. The difference between glancing and really looking.
The difference is everything.
If you’ve ever been the thing that everyone walked past — the broken machine, the dismissed idea, the person who didn’t fit the category — you know what it feels like to be filed and moved on from.
And if you’ve ever been the person who stopped. Who looked again. Who asked, “What does it tell you?” — you know what it costs to see what others miss.
And you know what it pays.