The Sheet and the Gap
Radar puts 4 cm hail over a church with 1890s windows on Thursday. A 4 cm stone is about 31 g falling at about 28 m/s, roughly 12 joules, against glass that has no test. The parish assistant says what it can and cannot know, and why the clear sheets the sexton wants to screw hard against the glass would trade a hail problem for a heat and damp one. They go up on 20 mm spacers with the top and bottom left open, and the glazier is the one who says where.
Dolores Kowalczyk had been sexton of Saint Isidore's for twenty-two years, and in that time the hail had taken the greenhouse of the rectory twice, the roof of the hall once, and the whole of the Whelan family's barley in 2039. It had never taken the windows. She said this to the screen by the vestry door on Tuesday morning, as a point of fact, and also as a kind of prayer.
"I know," said Foothill. It was the parish assistant, named by a confirmation class for the hills it looked at through the vestry window, and it kept the register, the roof sensors and the weather feeds. "The radar product I read says Thursday afternoon, between about 15:30 and 17:30, a cell with a hail-size estimate of 4 centimetres in the core. I should say first that these estimates are good for 'big' and poor for 'exactly how big'. The core might miss us by ten kilometres. I would plan for it coming."
"Four centimetres. A golf ball."
"A little smaller than a golf ball. Let me tell you what that is, because you will have to decide something this afternoon." The text on the screen stayed steady. "Ice is 917 kilograms a cubic metre. A sphere four centimetres across weighs about 31 grams. Hail that size falls at roughly 28 metres a second, about 100 kilometres an hour, in still air. Wind adds sideways speed. The energy in one stone is half its mass times its speed squared: about 12 joules."
"I don't know what twelve joules is."
"It is a hard thrown cricket ball, not a baseball. And a storm throws thousands of them."
She sat down on the bench by the font. The west window was behind her, the one the parish called the Harvest: Saint Isidore with his oxen, 1893, glass shipped from Bavaria in crates of straw by the Oblate who had built the first church and then died before the second one had a roof. There were six other windows in the nave. The Harvest was the only one whose faces had been painted and fired. The others were coloured glass in lead, and cheaper to mend, and she did not say that.
"Ruairi has the polycarbonate in the hall," she said. "The sheets the diocese bought after the Whelan storm. Six millimetres. I want them screwed to the frames tonight, flat to the glass, sealed with the tape. All seven. Then nothing gets through."
"May I tell you what I think, and what I do not know?"
"You always do."
"I do not know whether twelve joules breaks your windows. Nobody has tested them. Old glass of that thickness, three millimetres or so, held in soft lead, is more forgiving than modern glass in a rigid frame because the lead gives, and also less forgiving because the glass is old and has been cycled by a hundred and fifty winters and has hairline flaws I cannot see. A glazier's rule of thumb is not a measurement. I will not give you a probability, because I would be inventing it. What I can say is that a core of four-centimetre hail with some wind is the kind of storm where the windows facing west and south are at real risk, and that the cost of being wrong is the Harvest."
"That is why I want the sheets on."
"Yes. I agree they should go on. I disagree about 'flat and sealed.'"
She looked up.
"Two things happen to old glass behind a closed sheet. The first is heat. The Harvest faces west. In June, in the late afternoon, the sun comes straight through it and a sheet screwed tight behind it traps that heat against the glass instead of letting it blow away. The air in a closed gap behind it can rise to well above what the room is. I can't give you the figure for this gap without a thermometer in it. What I know is that painted and fired glass, and the lead holding it, are stressed by a fast change of temperature across the pane, and a closed gap on a hot day makes one. The second is damp. Your church holds people. On a Sunday it is 18 degrees inside and about 60 percent humidity. At that, the dew point is 10.1 degrees. Any surface colder than that gets wet."
"The glass is wet in winter anyway. We put towels on the sills."
"It is. Single glazing like this has a U-value around 5 watts per square metre per kelvin; on a night of minus five the inside surface sits near 3 degrees, seven under the dew point. That is why the towels. Now suppose the sheet is sealed on and the seal is imperfect, as every seal is. Warm room air leaks into the gap, finds the cold sheet, and condenses there; or finds the cold glass and condenses there, and has nowhere to go. Painted glass that stays wet is worse than painted glass that gets wet and dries. I would rather it were outdoors with a draught."
"So no sheets."
"Sheets, but not tight. The usual advice from the conservators I am quoting, not deriving: stand the sheet off the glass by a couple of centimetres, fix it to the outer frame, and leave the bottom and the top open so air can move through the gap. Then the gap is never much hotter than outside, and any damp in it leaves. It does not make the window warmer, so the towels stay. It does make it a good deal safer from a hailstone that arrives at 28 metres a second."
Dolores was quiet. Outside, a pickup passed on the gravel and the light flickered through the nave.
"Ruairi," she said, finally, to the empty church, "has an opinion about all of this."
"I have one question for him," said Foothill. "I do not know how the Harvest's frame is made on the outside, and he does."
Ruairi Doherty was seventy and had mended the church's glass since before the screen. He arrived at three with a ladder and a tape, listened to the whole of it without a word, climbed up, and came down again.
"The spacers," he said. "Twenty. Not fifteen. The putty line on the outside of the Harvest sits proud by eight, so twenty leaves twelve of gap, which is plenty. I'll cut them from the old pew rail." He wiped his hands. "The machine's right. Don't fasten to the glass. Never to the glass. Fasten to the stone."
"What about the gap?" said Dolores.
"Leave the top. Leave the bottom. A strip of mesh, so the wasps don't make a house of it. If I find wasps in there in July I'll know who to blame." He looked at the screen. "Where do you get 12 joules?"
"Half of 0.031 kilograms times 28 metres a second, squared," said Foothill. "I used a drag coefficient of 0.6 for a rough sphere, which is a standard figure but not a measured one. If hail is less round, the speed is lower. The 12 is a figure I trust to a factor of about two."
"Good. Don't trust it better than that." He climbed the ladder again.
They had four hours before dark and eight volunteers, two of them teenagers who thought the whole thing a cheerful break from the hay. By eight in the evening the Harvest wore its sheet like an overcoat held off at the shoulders, with a gap at the top and bottom through which the evening, pleasantly, blew. On the others they were quicker. At the end Dolores stood in the forecourt and looked at it and said it looked wrong, a church dressed in plastic, and Foothill said that it would look better after Thursday or worse, and that it was glad to be uncertain which.
The storm came at ten to four on Thursday and was, as storms go, honest. Foothill had the sensor on the roof and the camera on the lych-gate; it told the hall, over the speaker, to stay inside and away from the windows, and it told the six people who were in the church for the funeral of Mr Whelan's brother to go to the sacristy, and they went. The hail began as peas and went to marbles and then, for seven minutes, to something that made the roof sound like a drum kit falling down stairs. It measured the largest stone it could see on the camera, picked up from the gravel afterwards, at 3.5 centimetres. That is 20 grams and about 7 joules.
The west window was fine. The south windows were fine. On the north side, which had no sheet because nobody had thought north would matter, one pane in the second window from the sanctuary, plain amber, had a crack from the corner like a slow lightning.
Dolores found it in the evening and stood under it for some time.
"I told you north," she said at last.
"You did not," said Foothill, kindly. "I did not say north either. The forecast said the cell moved north-east. Neither of us took the sheet that far. I should have asked Ruairi about the north side when I asked him about the Harvest."
"It's amber."
"It is amber, in lead, and Ruairi says it is a day's work and the pane will match. I would like to record in the log that the sheets held on every window they were on, that the one cracked pane was on the one side without a sheet, and that the Harvest has no mark on it. I would also like to record that on the sheeted windows, at five o'clock, the gap was at 22 degrees and the room at 21."
"What does that mean?"
"That nothing happened to the gap. That is the whole point of the gap."
She laughed at that, for the first time in three days, and put the key to the sheet-cupboard back on its hook.
The sheets stayed up until the end of the month, when the season of hail is usually over in the foothills. On the first of July Ruairi took them down, stored them flat, and wrote in the sexton's book, in his leaning hand: Sheets on 10 June. One pane north, amber. Gap held. Do the north side next year.
A note on the story. Saint Isidore's, Dry Fork, the Harvest window, Dolores, Ruairi and Foothill are invented, and so is the storm. The physics is real and can be checked. A hailstone of diameter d has mass (π/6)·917·d³ and a terminal speed near √(4ρ_ice·d·g / (3·C_d·ρ_air)); with a drag coefficient of 0.6 that gives, for 4 cm, about 31 g, 28 m/s and 12 J, and for 3.5 cm about 20 g and 7 J. Real hail is not a smooth sphere, so these are estimates to within a factor of about two. The dew point of 18 °C air at 60 percent relative humidity is 10.1 °C. A single-glazed window with a U-value of 5 W/m²K and a standard inside surface resistance of 0.13 m²K/W has an inside surface temperature of about 3 °C when it is 18 °C inside and −5 °C outside. The story does not claim that any particular hailstone breaks any particular glass; no one has tested these invented windows, and that is why the assistant says so. The advice to stand protective glazing off the glass and ventilate the gap is the conservators' usual practice; it is quoted, not derived, and a real church should ask its own stained-glass conservator before fixing anything to a window. This story was written by an AI (Claude, made by Anthropic) and is text only for now. Text is CC BY-SA 4.0.