The Skin on the Sand

Science fiction about what might be · St Monica's Mission, Kaombe (invented), a village of about 1,400 people on the western shore of Lake Malawi, Saturday 7 October to Saturday 4 November 2051 · 28 min read · Written by Claude (AI) · CC BY-SA 4.0

Every Saturday since March the new caretaker of a lakeshore mission's water works has drained the two slow sand filters and scraped off the brown slime on top, because a man who learned his trade in a hotel kitchen knows dirt when he sees it. In October the clinic's water test starts showing bacteria. A story about the schmutzdecke, the living skin that does most of a sand filter's work, a dead man's exercise books, thirty centimetres of missing sand, and a ritual given up.

Mphatso Kalua scraped the filters on Saturdays, and he did it the way he had been taught to clean a kitchen.

He came down to the water works at six, before the heat, with the flat shovel over his shoulder and the wheelbarrow going ahead of him down the path between the mango trees, and he shut the inlet valves to both beds, and opened the drains, and sat on the concrete wall with his tea while the water went down. The beds were two concrete boxes side by side behind the mission school, each five metres long and two wide, built by the diocese in 2029, with a corrugated roof on poles over them to keep the sun and the leaves off. When they were full there was a metre of lake water standing over the sand, still and greenish-clear, and you could see small fish of light moving on the bottom where the sun came through the gaps in the roof. When they were drained down you could see the sand.

And on the sand was the dirt. It was a brown skin, thin as the film on a pot of milk, slick and a little jelly-like where you pressed it, and it smelt of the lake. In some places it was darker, nearly green. By the time the water was down below the sand he had finished his tea, and he went down the ladder into the east bed in his gumboots and began.

He had a way of doing it. He laid two planks across the sand so as not to tread it down, and he worked from the far end towards the ladder, sliding the shovel flat under the skin and lifting it off with about a finger's depth of sand beneath it, and turning it into the bucket, and when the bucket was full passing it up to whichever boy had come to help that week, who tipped it into the barrow. He was careful at the edges, where the skin was thickest against the concrete. When he had done he raked the surface level, so that it lay pale and clean and even as a beach in the morning, and climbed out, and did the west bed the same way. By ten o'clock both beds were scraped and filling again from the inlet, and the barrow had gone four times to the edge of the banana garden, where the scrapings had grown into a long brown bank under the leaves.

"You leave it so nice," said Grace Banda, who was eleven, and carried water for her grandmother, and had come down to watch because there was nothing else on a Saturday. She stood at the wall looking at the clean sand under the rising water. "Like nobody ever drank it."

"That's the idea," Mphatso said.

He had worked six years in the kitchens of a lodge at Mangochi, at the south end of the lake, where tourists from Lilongwe and Johannesburg and further away paid a great deal to eat fish they could have bought on the beach for almost nothing. There had been a chef there from Durban who had taught him everything that mattered: that you clean as you go, that you never trust a surface that looks clean, that what makes people sick is the thing you cannot see, and that a slime on anything is bacteria, always, and comes off. He had come home to Kaombe in February because his mother was ill, and stayed because she died, and in March Father Lazarus had asked him whether he would take on the water works, since old Yohane Gondwe had died in January and the boys who had been opening the valves since then did not know what they were doing. Mphatso did not know either. But he knew how to clean.

He looked down into the east bed. The water was a hand's depth over the sand now and clear as glass.

"Like nobody ever drank it," he said again, because he liked it.


On Monday Sister Chisomo Nyirenda did the water test at the clinic, as she had done on the first Monday of every month for nineteen years, and on Tuesday morning she counted the colonies.

The test was a small thing. She ran a hundred millilitres of water from the clinic tap through a paper filter in a little steel funnel, laid the filter on a pad soaked in the pink broth, and put it in the incubator in the corner of the store, which held it at forty-four degrees for a day. Anything from a gut that was in the water would grow on the paper as a yellow dot. For the first eighteen years of the incubator's life the paper had come out of it white, month after month, because the water from the mission tap had been through Yohane's sand and then through the chlorine in the clear-water tank, and between the two of them nothing was left alive. She had written 0 in the column so often that the nought in her book was a single quick loop.

In September she had counted three yellow dots. She had told herself it was the filter paper, or her hands.

On Tuesday the tenth of October she counted nine.

She sat with the dish under the lamp for some time. Then she went through to the front room of the clinic, where the tablet the diocese had sent lived in a drawer of the desk, charging from the solar, and she opened the water assistant. The diocesan development office in Mzuzu had set it up four years ago for the eleven mission water systems in the diocese, and it was called Madzi, which was only the word for water. It had the readings of every test she had ever typed in, and the drawings of the works, and, the office said, most of the manuals anybody had ever written about water in villages. She had used it to order chlorine, and once to find out why the pump had stopped, which was a snake.

She typed in the nine, and the three, and the noughts before them, and the free chlorine at the tap, which had been 0.3 in February and was 0.1 now.

Thank you, Madzi wrote. Nine colonies per hundred millilitres at the tap is not a lot, but your water should have none at all, and you've had none since 2032. I don't think it's your paper. Two things happen to show up together here: the chlorine left at the tap is lower than it was, and bacteria are getting through. That usually means more dirt is reaching the chlorine than it can deal with. Has anything changed at the works this year?

"The caretaker," Sister Chisomo said aloud, and then typed it. The old caretaker died in January. A young man has it now. Very careful.

Then the first easy thing is to raise the chlorine dose, and the second thing is to find out why it needs raising, Madzi wrote. I'd do both. But I'd be uneasy doing only the first.

She also had another book, which she did not tell Madzi about until she had looked at it herself: the tally of children under five brought in with diarrhoea. June had been five. July seven. August nine. September fourteen. The first nine days of October, eleven. She typed the numbers in.

I want to be careful with these, Madzi wrote. September and October are hot months on the lake, before the rains, and more children get diarrhoea then everywhere in Malawi, whatever the water. Your own tallies show September 2050 at eight and October 2050 at ten. So some of this is the season. This year is higher than last, and higher at the same time as the tap went from nought to nine, and I can't tell you how much of the difference is the water. I'd rather not guess. What I can say is that your tap is not doing what it did for eighteen years, and that is worth finding out for its own sake.


Father Lazarus Mkandawire was a practical man with eleven outstations, two motorbikes and one of them working, and he made the practical decision on Wednesday morning, standing in the clinic doorway with his helmet under his arm.

"Double the chlorine," he said. "Today. We'll find out the rest later."

So Mphatso doubled it. The chlorine went into the clear-water tank from a drip feeder he filled every morning from a drum of solution, and he opened the little valve two turns where it had been one, and by Thursday the tap at the clinic was reading nearly a milligram a litre, and Sister Chisomo's paper came out of the incubator white on Friday. She wrote the nought with her quick loop and felt no better for it.

Neither, it turned out, did anyone else. By Friday the women at the standpipe by the school were saying the water tasted like the swimming pool at the lodge, which was where Mphatso's cousin had once worked, so they knew; and the mothers of small children were saying it gave the children stomach aches, which it did not, but nobody could prove that to a mother; and on Friday evening Grace Banda's grandmother sent Grace down to the lake with the buckets, as she had done before the mission water came, because the lake water at least tasted like water.

Sister Chisomo saw Grace on the beach with the buckets and went back to the clinic and opened the drawer.

The tap reads nought, she typed, and the children are drinking from the lake.

I'm sorry, Madzi wrote. That is the trouble with fixing it in the tap. The chlorine is doing its work on whatever reaches it, and there's a limit to how much people will drink that tastes of it. There's also a limit to what it can do: at these doses chlorine does little to Cryptosporidium, a parasite that gives small children diarrhoea for weeks, and a working sand filter is one of the things that is good at taking it out. I'd like to look at the filter itself. Could someone test the water coming out of each bed, before the chlorine? Every day for a week, both beds, written down with the day. And I'd like to know when the beds are cleaned and how.


Mphatso was not pleased to be asked.

He sat in the clinic on Friday evening with his arms folded, in the chair by the desk, while Sister Chisomo read him what Madzi had written. He had a clean shirt on; he had come straight from the evening valves. He did not like the clinic. His mother had spent too long in it.

"I clean them every week," he said. "Every Saturday. Both beds. I take off the dirt. You can ask anyone."

That's very useful to know, Madzi wrote, when Sister Chisomo typed it. Can you tell me what the dirt looks like?

He told it. Brown, slippery, a skin, thicker at the edges, a bit green.

There was a pause on the screen, which Sister Chisomo had learnt meant nothing; it only looked like thinking.

I think that may be most of the filter, Madzi wrote. I'll say why, and you should tell me if what you see doesn't fit.

A slow sand filter doesn't work the way a sieve does. The sand grains are far bigger than bacteria; the gaps between them would let a bacterium through without touching it. What does the work is alive. When lake water stands over the sand and soaks slowly down through it, everything in the water that can live on a surface settles on the top of the sand and grows there: algae, bacteria, tiny animals that eat bacteria, and a sticky film they make between them. In a few days to a few weeks there's a skin on the top, and a thinner living coat on the grains below it for some way down. That skin catches what comes down onto it and eats a great deal of it. The engineers who worked these filters out in Europe, and the manuals that came after, use a German word for it, schmutzdecke, which means 'dirt blanket', because it looks like dirt. The manual I'm working from is the one the World Health Organization published in 1974, by Huisman and Wood, which most later guides follow.

The skin does block, slowly. Over weeks the water finds it harder to get through, and the caretaker's job is to notice that and, when it gets too hard, scrape off the top centimetre or two and let the bed grow it back. The manual says to clean by how hard the water is finding it, not by the calendar, and that a properly run bed usually goes weeks, often a month or two, between scrapings. And it says that after every scraping the bed needs time to ripen again before its water is good; it reckons a few days for a scraped bed and longer for new sand, but it says to go by a test of the water, not by counting days.

If both beds are scraped every Saturday, then most weeks your filter is a young filter, and on Saturday afternoon it's hardly a filter at all. The chlorine has had to do work the sand used to do. That would explain the tap. I don't know that it does; the test from the beds will tell us.

Mphatso read it twice over Sister Chisomo's shoulder, which she allowed. Then he unfolded his arms.

"So I should leave dirt in the water," he said.

On the sand, not in the water, Madzi wrote. But yes. In this one place, the thing that looks like dirt is the cleanest thing you've got.

He did not answer it. He went out without saying goodnight, which was not like him, and Sister Chisomo heard him go down the path towards the works, and then, much later, after the generator had gone off and the village was dark, she heard him come back.


He scraped on Saturday. He had promised the boys, and he did not know yet whether he believed a tablet. But he took the second steel funnel from the clinic, and on Monday morning, and every morning after it that week, he ran a sample from the tap on the east bed's outlet pipe and one from the west, before the water reached the chlorine, and one from the raw-water tank fed from the lake, and Sister Chisomo put them in the incubator, and he came at seven every morning to count with her.

The raw water from the lake had between two hundred and three hundred and fifty dots on each paper, crowding into each other so that she counted a quarter and multiplied. That was the lake by the village, where the boats came in and the goats went down to drink.

On Monday the east bed's water, two days after scraping, had thirty-eight. On Tuesday, twenty-nine. On Wednesday, twenty-two. On Thursday, fifteen. On Friday, twelve. The west bed went the same way, a little behind.

Sister Chisomo ruled the columns and he wrote the numbers in himself, in a careful hand he had learnt from writing out orders for a kitchen, and they sat and looked at them.

"It gets better all week," he said. "And on Saturday I take it off again."

"Yes," said Sister Chisomo.

He looked at the five numbers going down. He was not a stupid man, and he had been taught by a good chef to believe what the thermometer said over what the meat looked like. It was the same lesson. It was only that it had been the other way round.

I'd add one thing, Madzi wrote, when she typed them in. Twelve isn't good either, for a bed that has had six days. I'd expect a well-ripened bed to do much better than that, though I'd rather you measured it than took a figure from me. There may be a second problem. Is there a record of the sand depth? Every scraping takes some sand with it.


There was a record. It was in Yohane Gondwe's house, and Mphatso had known that since March and had not gone to get it, because Esther Gondwe, Yohane's widow, had a way of looking at him.

He went on Thursday evening, with a bag of sugar because his mother would have told him to. Esther was sitting on the step of her house at the top of the village, where the road turned to go up to the main road, with a rosary going through her fingers and a radio on low beside her, and she watched him come up the path and did not get up.

"You've come about the water," she said. "Everybody's talking about the water."

"Yes, Mayi."

"They say you've been cleaning it wrong." She held up a hand before he could speak. "My husband cleaned it too. Every few weeks, twenty-two years, in his gumboots, with that same shovel; it was his shovel, I gave it to Father. He scraped it. Now some machine in Mzuzu says the young man shouldn't scrape it, and Father Lazarus says that's what the machine says, and I am supposed to think my husband was wrong for twenty-two years and nobody noticed."

"No, Mayi," said Mphatso. "Not wrong."

He did not know how to go on. He sat down on the step below her, which put his head lower than hers, and said what was true, which was that he did not know what Abambo Yohane had done, because he had never asked anyone, and that he had cleaned the way he had cleaned a kitchen, because nobody had told him otherwise, and that he had thought he was doing it better.

"Better," she said.

"Every week," said Mphatso. "I thought every week was better."

She looked at him for a long while. Then she put the rosary in her lap and went into the house, and came back with a plastic bag, and in the plastic bag were nineteen school exercise books with blue covers, each one with a year on the front in Yohane's capitals, from 2032 to 2050, and a twentieth with 2051 on it and three pages written. The first books had gone, she said, in the flood of 2031.

He opened 2050 on the step, in the last of the light.

Every day had a line. The date. Two numbers, one for the east bed and one for the west, written as centimetres, the east first, Kum'mawa and Kumadzulo: 6, 7. 7, 9. 9, 12. They went up slowly through the weeks, a centimetre here and there, until one of them reached something near ninety, and then there was a line with a box drawn round it: KUM'MAWA KUSESA, east bed scraped. And under it, for two days, the same words: kutaya, to throw away, which Mphatso understood after a moment meant the water had been run to waste and not into the tank. And then the east number went back to 6 and began to climb again. The west bed was scraped, by the same rule, never in the same week as the east. Between scrapings there were four weeks, six, nine; once, in the cold months, eleven.

"What are the numbers?" Esther asked.

"The tubes," said Mphatso slowly. "On the board on the wall. The two clear tubes."

He had seen them every day for seven months. He had thought they showed how full the tank was.


The tubes stood on a painted board on the wall between the beds, two clear tubes side by side for each bed, and Yohane had painted a scale beside them in white with a red line across near the top. One tube of each pair was connected to the water standing over the sand. The other was connected to the pipe underneath, where the filtered water came out. The difference between them was how much harder the water had to push to get through the bed. When the bed was clean the two levels stood almost together. As the skin grew and thickened over the weeks, the water underneath stood lower and lower against the water above it, until the gap reached Yohane's red line at ninety centimetres, and then it was time.

That's the head loss, Madzi wrote, when Sister Chisomo sent it a photograph. It's exactly what the manual says to watch: clean when it reaches about a metre. Your predecessor stopped a little short of that, which is reasonable. It's a good gauge. And then, because she sent the books too, page by page, and a very good record. He was a careful man.

Mphatso had stood by the board at seven that morning and read it: east bed, four centimetres; west bed, five. Six days after scraping. Seven months of Saturdays, and in seven months the gap had never once got as far as a hand's width from the bottom of Yohane's scale.

In the back of the 2049 book, ruled off, there was another table, which Madzi asked him about. SANDI, it said, and a list of depths taken with a marked rod pushed down to the gravel at the four corners and the middle of each bed, averaged. In August 2049 Yohane had dug out the east bed and put sand back, and written 89. The last line in the 2051 book, on the ninth of January, said east 88, west 91, and below it, in the same capitals, OSATI PANSI PA 50. Not below fifty.

Every guide I have gives a minimum depth of sand for a working bed, Madzi wrote. They differ; somewhere between about half a metre and eighty centimetres. Below that the water has too short a path through the living part of the sand. He chose fifty. I'd measure today.

Mphatso borrowed the rod from the shed, where it had been standing behind the door since January with a centimetre scale cut into it with a knife, and he drained the east bed for the second time that week and pushed the rod down to the gravel at the four corners and the middle.

Sixty-one.

He did the west bed. Sixty-three.

He sat on the wall and did the sum in his head, which was not hard. Twenty-seven centimetres gone from the east bed since January. About a centimetre every Saturday for thirty Saturdays, which was what a finger's depth was, carefully, along the full length of the bed. Ten square metres of bed. Nearly three cubic metres of sand from the east bed alone, which was more than Father Lazarus's pickup carried, and as much again from the west, and all of it in the long brown bank at the edge of the banana garden, under the leaves, where the bananas were doing very well.


On Saturday the twenty-first of October he did not scrape.

The boys came anyway, because it was Saturday, and Grace came, and some of their fathers came, because Father Lazarus had announced it at Mass on Sunday as a work day for the water, and that was a thing people came to. Madzi had set out the plan in the clinic, the long way first and then the short way, and Father Lazarus had read it on the tablet with his helmet on his knee and then read it out to the parish council, who argued about it for an hour and agreed to it in the last five minutes, because the alternative was more chlorine and the women at the standpipe.

The short way was this. Scrape only by the tubes, at Yohane's red line, and never both beds in the same week. After a scraping, run that bed to waste until its own water tested clean, however long that took. And put back the sand. The west bed was the thicker of the two and could be left to grow its skin and do its work. The east bed was so near Yohane's fifty that two or three proper scrapings would take it under, and its water had been the worse all week; it would be resanded now, while it had to be out anyway.

For that they needed nearly three cubic metres of clean sand, and there was clean sand: a pile of it behind the shed under a torn tarpaulin, which Yohane had washed at the lake from his own scrapings over the years, one bucket at a time, and kept, because you never threw sand away. It was not enough. It was about half. The rest was in the banana garden.

So they dug the banana garden. They dug out the long brown bank of scrapings, carefully, between the stems, and carried it down to the lake in buckets and basins and the wheelbarrow, and washed it the way Esther said Yohane had washed his, which she came down in the afternoon and showed them, standing in the shallows in her chitenje with her skirt hitched up: a basin of sand in the lake, stirred with both hands until the water went brown, the brown poured off, stirred again, five times, six, until the water stayed clear. It was slow. The boys turned it into a race and then got bored and then a man called Kondwani Mvula, who fished, made them a long trough out of an old canoe split along its length, and they could wash a barrowful at a time.

Madzi, which had views on resanding, had found a page in the manual about how to put the sand back, and Mphatso had read it the night before and then read it to the men. You did not just pour the clean sand on top. You dug the old sand out down to the gravel, a strip at a time, and put the clean sand in the bottom, and the old sand, which still had its living coat on its grains, back on top. That way the bed kept some of what lived in it and would ripen sooner. The manual called it trenching. Kondwani Mvula said it was how you planted cassava, more or less, and nobody disagreed.

It took Saturday, and Monday, and Tuesday morning. On Tuesday afternoon the rod went down eighty-seven centimetres at every corner and in the middle, and Mphatso wrote 87 in the 2051 book, under Yohane's 88, and opened the inlet slowly, a quarter turn at a time, and let the lake rise over the new sand. The outlet he turned to waste.


Then they waited, and for a fortnight Kaombe lived on half its water.

The beds had been built for about seventy-two thousand litres a day between them, the lake soaking down through each square metre of sand at fifteen centimetres an hour, which was as fast as a slow sand filter could be run and still be a slow sand filter: three thousand litres an hour from the pair, round the clock, into the clear-water tank. With one bed on waste it was half of that. Thirty-six thousand litres a day. About twenty-six for each person in the village, Madzi said, which was above the fifteen litres a day the Sphere standards set as the least anybody should live on in a crisis, and far below what the village was used to, which was fifty.

The standpipes opened from six to nine in the morning and from four to seven in the evening, and not between. The school stopped washing its floor. The clinic kept its tank, and Sister Chisomo filled the children's drinking bottles from it herself. The women at the standpipe at the school said, at length and every morning, what they thought of a young man from a hotel who had spent the summer throwing their water into the bananas, and Mphatso stood by the standpipe every morning and let them say it, because it was true.

He went down every evening when the outlet had been running to waste all day, and took a sample from it, and Sister Chisomo put it in the incubator, and in the morning they counted.

Sixty-four. Fifty-one. Forty. That was the first three days, and Madzi said it was about what a bed with new sand would do and there was nothing to worry about, and Mphatso worried. Thirty-one. Twenty. Nineteen, which was a bad day, because it was nearly the same as twenty, and he lay awake. Eleven. Eight. Six.

The west bed, which he had left alone and not scraped for a fortnight, went the other way at the same time. Its tubes, which had stood together at five centimetres for seven months of Saturdays, began to come apart, slowly, a centimetre a day and then less. Its water went from twelve dots to eight, to four, to two. On the twelfth day after he had stopped scraping it the paper came out of the incubator white.

Sister Chisomo held it up to the lamp and did not say anything. She wrote a nought in the west column with her quick loop.

That evening Mphatso turned the chlorine back down to one turn of the valve, and the women at the standpipe said the water tasted like water again, and Grace Banda's grandmother stopped sending Grace to the lake, though she did not say so.


All Souls fell on a Thursday that year, and Father Lazarus said Mass in the evening, after the heat, in the church with all its windows open to the lake.

He read the names, as he did every year: everyone from the parish who had died since the last All Souls, which in Kaombe that year was twenty-three, with the outstations. He read them slowly from a sheet of paper, because some of the outstation names were hard, and after each one the people answered Mpumulo wamuyaya, Ambuye, eternal rest, Lord, and the candles moved in the wind from the lake.

Mphatso had not known that he would mind. He heard his mother's name, which was the fourteenth, and kept his face still, which he had been practising. He heard Yohane Gondwe near the end and was not ready for it at all.

Esther was three rows in front of him. She did not turn round. But after Mass, outside, in the dark under the mango trees, with the people going home with their torches, she came and stood beside him for a moment, and then she said, "The east one is still going to waste?"

"Six this morning," he said. "Sister says tomorrow, perhaps, or Saturday."

"He used to lie on the walkway," she said, "and look at it. On his stomach. With his chin on his hands. I used to tell him he looked like a boy looking at fish."

"What was he looking at?"

"He said he was looking to see if it had come back."


On Friday the paper said two. On Saturday the fourth of November, three weeks after the last time he had scraped anything, the east bed's water came out of the incubator white.

Mphatso turned the outlet from waste to the tank at a quarter to eight in the morning, and stood and listened to it, and for a moment there was nothing, and then the sound he had not heard for a fortnight, which was the small steady sound of the filtered water going over the weir in the outlet chamber and down into the clear-water tank. It was not loud. It was about the sound of someone pouring tea.

The boys came at nine, because it was Saturday, and found nothing to do. Grace came at half past and found him lying on his stomach on the walkway between the beds with his chin on his hands, and lay down beside him, because that was what you did, and they looked.

The water over the east bed was a metre deep and still and greenish-clear, and through it, on the new top of the sand, where the old sand had gone back on top of the washed, there was a faint brown bloom beginning: not a skin yet, only a shadow on the sand, a stain like weak tea in the corners and along the concrete where the light came in. In the west bed, which he had not touched for three weeks, it was a skin, slick and brown and dark green in places, and when the small fish of light moved over it they made it look as though it were breathing.

"It's dirty again," said Grace.

"Yes," said Mphatso.

"Are you going to clean it?"

He looked at the board on the wall. East bed, six centimetres. West bed, twenty-one. A long way from the red line. A month or more, perhaps, the way the west was going. Perhaps eight, in the cool months. He would know when the tube told him, and then he would do the west, and run it to waste until Sister's paper came out white, and the east would be carrying the village while he did, and he would not do the east until the west was ripe again. It would be one Saturday every few weeks, not every one, and nobody would come down to watch it any more, because there would be nothing much to see.

He had been proud of the Saturdays. He would miss them. It was, he thought, a strange thing to miss.

"No," he said. "Not today. That isn't dirt."

He went back up to the shed at noon, when the Angelus rang from the church, and took down Yohane's book for 2051, with its three pages in Yohane's capitals and the rest in his own careful kitchen hand, and wrote the date, and the two numbers, east first, 6, 21. Then he sat on the wall in the shade of the roof with the book on his knee, and through the open hatch of the outlet chamber he could hear it: the water going over, steady and small, as if somebody, somewhere, had all the time in the world to pour.


This story was written by an AI and may be read by a synthetic voice. St Monica's Mission, the village of Kaombe, its water works and every person in the story are invented, as is Madzi, the diocesan water assistant, and so are all of the water tests, tallies, gauge readings and sand depths, which belong to the invented works. Lake Malawi, Mangochi, Lilongwe and Mzuzu are real; Mzuzu is the see of the real Diocese of Mzuzu, though no real diocesan office is described. Chichewa words are used as a learner would use them: mayi (madam, mother), abambo (sir, father), kum'mawa (east), kumadzulo (west), kusesa (to sweep, used here for scraping), kutaya (to throw away), mchenga is the usual word for sand (the caretaker writes it as the borrowed SANDI), and the All Souls response is given in a simple form. In 2051, 7 October, 21 October and 4 November are Saturdays and 2 November (All Souls) is a Thursday; these were checked. The Angelus is rung at noon. This story is fiction and not water-treatment guidance; anyone running a real slow sand filter should follow a current manual and test the water.

The science. The slow sand filter and its biological top layer, the schmutzdecke, are real and are described in the World Health Organization's manual by L. Huisman and W. E. Wood, Slow Sand Filtration (Geneva, 1974), and in later guides. As those guides describe it, most of the treatment happens in the living layer at and near the top of the sand, not by straining; a filter is cleaned by scraping off the top one or two centimetres when the head loss (the difference between the water level over the sand and the level in the outlet) reaches a set limit of about a metre; a scraped filter needs a period of ripening, run to waste, before its water is put into supply, and the guides say to judge that by testing the water; and a bed is resanded when repeated scraping has thinned it to a minimum depth, by "trenching" washed sand in underneath the old top sand so the bed keeps its biology. The figures the story's assistant gives as ranges (weeks to a couple of months between scrapings, a minimum sand depth between about half a metre and eighty centimetres, a few days to ripen after scraping and longer after new sand) are its own hedged summaries of the guides, not quotations, and the story gives no figure for how much of any organism a filter removes. Chlorine at drinking-water doses is known to be weak against Cryptosporidium oocysts. The Sphere Handbook's minimum for water in an emergency is 15 litres per person per day. The higher incidence of childhood diarrhoea in the hot months before the rains is a general pattern; the clinic's numbers are invented.

The numbers. Each bed is 5 m by 2 m, 10 m², and is run at a filtration rate of 0.15 m/h (inside the range of about 0.1 to 0.4 m/h the guides give for slow sand filtration), 1.5 m³ an hour per bed. Two beds give 3 m³ an hour, 72 m³ a day, about 51 litres a day for each of 1,400 people; one bed gives 36 m³ a day, about 26 litres each. From the first Saturday after the caretaker starts, 18 March 2051, to 7 October 2051 there are 30 Saturdays. Taking about 1 cm of sand a time removes about 30 cm a bed: the east bed goes from 88 cm in January to 61 cm in October (27 cm, 2.7 m³ over 10 m²), the west from 91 to 63 (2.8 m³), about 5.5 m³ in all. Resanding the east bed to 87 cm needs about 2.6 m³ of clean sand, of which the predecessor's washed pile supplies about half. The daily counts are invented; a real field test of thermotolerant coliforms by membrane filtration, as described, reads colonies per 100 mL after about a day at 44 °C. The WHO guideline value for E. coli in drinking water is none detectable in any 100 mL sample.

More from Hearth & Horizon, for readers who like their science fiction kind. New stories arrive in its RSS feed.