Three Pouches
A parish spring-tank gets one 24 g pouch of chlorine a fill. In the dry season it leaves 1.4 mg/L and the women go back to the stream; in the first rains the spring turns brown and the same pouch leaves nothing. The chairman wants the dose doubled. The tank's assistant works out why one dose cannot serve both seasons, a twelve-year-old with a clear tube and a kitchen scale supplies the missing reading, and the women's blind taste test sets the number the assistant would have taken from a guideline.
The tank at Kyabahinga held six thousand litres, stood on a stone plinth below St Charles Lwanga church, and was filled by a spring that gave about a fifth of a litre a second when it was behaving. That is seven hundred and twenty litres an hour, so a fill took a little over eight hours, and every fill got one pouch of chlorine.
The pouch was twenty-four grams of calcium hypochlorite, sixty-five percent available chlorine, weighed out on the school's kitchen scale by whoever had the key. The tank's assistant, which the parish called Tally because it counted things, had been given the job of recording what that pouch did.
On Sunday the twenty-first of September, after the ten o'clock Mass, Tally told the water committee what it did.
"In the dry season, it leaves one point four milligrams a litre at the tap," it said, from the speaker the school had wired to the tank house. "In the last four days, with the first rain, it leaves nothing I can measure. I tested at the tap at six, at noon and at six, twice, with the colour kit. Zero each time."
"Then it is too weak," said Mzee Byaruhanga, who chaired the committee and had chaired it since the tank was a hole in the ground. "We double it. Two pouches. Forty-eight grams."
"I would like to show you what that does first," said Tally. "Will you let me?"
Here is what was true in 2026, when this story was written, and what had been added by 2047.
In 2026 chlorine at a village tank was already old and well understood, and it already worked best in water that looked clean. A person, or a pump, adds enough to kill what is in the water, and a little more so that some is still there, a few hours later, when somebody opens the tap. The part that gets used up on the way is called the chlorine demand. It is not a constant. It goes up when the water carries mud, leaves and dissolved iron, all of which take chlorine without being any danger to anyone. The rule people quote, a free chlorine residual of at least 0.2 to 0.5 milligrams a litre at the point of use, is the amount left over after the demand is paid. Muddy water is also harder to disinfect for a second reason: a bacterium sitting inside a clump of silt is shielded from the chlorine around it. Guidance since the 2000s has said that chlorination works well only below about 5 NTU of turbidity (a measure of cloudiness), and the usual advice above that is to settle or filter first.
What had been added by 2047 was not chemistry. It was a cheap assistant on a speaker that never lost the log, that could do the arithmetic aloud in a language the committee chose, and that could keep three seasons of one tank's records in a form a twelve-year-old could add to. The chlorine was the same.
Nothing in this story is a prediction that such a speaker will exist. It assumes steady improvement and no sudden jump, which nobody can check from here.
"The pouch is twenty-four grams," said Tally. "Sixty-five percent of that is chlorine: fifteen point six grams. In six thousand litres that is two point six milligrams in every litre. Say it is the dose.
"In the dry season the spring is clear and the water takes up about one point two of it. Two point six take away one point two leaves one point four. That is what I measured, and it is what Mama Josephine's women call the hospital taste. They are not being difficult. The World Health Organization's own guideline puts the most a person may drink for life at five. It is not a number anyone is meant to enjoy.
"In the rains the spring runs brown. I measured the demand on a jar of it on Wednesday, with a measured dose and an hour of waiting, and it is about three. Two point six take away three is less than nothing, so there is nothing at the tap, and the water has been drunk as it came for four days.
"Now double it. Forty-eight grams is five point two milligrams a litre. In the dry season that leaves four. In the rains it leaves two point two. So your two pouches fix the rainy season and make the dry season undrinkable, and even the rainy season is poor, because two point two at the tap tastes like swimming, and because chlorine at a dose that high in brown water is still working on mud and not on the germs inside it. One dose cannot be right in both seasons. It has to follow the water."
Mzee Byaruhanga said nothing for a moment. He had a way of looking at a speaker as if it were a person who had said something clever at a funeral.
"And the women go to the stream," he said.
"Seven of the nineteen households I could count went back to the stream in August," said Tally. "I do not know what the stream water is in the rains. It is probably worse. I do not know how many of the nineteen have gone back since, because nobody tells the tank where they drew from. The chlorine is not the only thing in this decision I cannot see."
"What does it need?" said Josephine Atuhaire. She was a catechist and she kept the key. She was also the one, in August, who had stopped drinking from the tank herself.
"Three things," said Tally. "A way to know, every fill, how dirty the water is. A dose for each of three kinds of water. And a number for the taste that is yours, and not mine."
The way to know came from the school.
Prisca, who was twelve, had made a turbidity tube in science with her teacher the previous term: a clear plastic tube a metre long, a black cross on a white disc in the bottom, and a ruler up the side. You pour the water in until you can no longer see the cross from above. A short column means dirty water and a long one means clear. Her teacher had calibrated it against a borrowed meter, once, and written the three readings on the tube in marker. Over sixty centimetres: below five NTU, clear. Between twenty and sixty: five to twenty, cloudy. Under twenty: brown.
"Brown means the first drum is not enough," said Tally. "Brown water should settle overnight in the old drum before it goes to the tank, and go through a folded cloth. Four folds. I measured it the week after the first rain: a night's settling takes a brown fill from about sixty NTU to about twelve, and the cloth takes it to eight or ten. That is cloudy, not clear, and it is better. If the tube still says brown after the drum and the cloth, nothing I can offer is good enough, and I will say so: the households boil."
"That is a lot of steps for a pouch of powder," said the chairman.
"It is three," said Tally. "Settle. Look in the tube. Pick the pouch."
The three pouches were worked out on the school scale that Sunday evening with Prisca doing the sums on a slate and Tally checking them aloud.
For clear water the demand was one point two and the aim was a residual of half a milligram, so a dose of one point seven. Six thousand litres at one point seven is ten point two grams of chlorine, and ten point two over sixty-five percent is fifteen point seven grams of powder. A sixteen-gram pouch. For cloudy water the demand was two, so a dose of two point five: fifteen grams of chlorine, twenty-three point one of powder. Twenty-three grams. For brown, after settling and the cloth, the demand was three, so three and a half: twenty-one grams of chlorine, thirty-two point three of powder. Thirty-two.
"Sixteen, twenty-three, thirty-two," said Prisca. "The old one, twenty-four, was the middle one all along."
"Yes," said Tally. "Nobody picked it. It was a good guess for a week in March."
They painted the three numbers on the plinth in white, yellow and red ochre, and under each one the tube's reading, so that a person with the key and no assistant could do it.
The number that was the women's took a week.
Tally's half a milligram was not its own. It came from a guideline, written by people who had done the arithmetic on many tanks they would never see, and Tally said so when Josephine asked where the half came from. "If your women taste half a milligram and spit, a guideline will not stop them going to the stream," it said. "I would like them to taste it first."
So on the second Saturday in October, after the market, twenty women sat on benches outside the tank house and tasted three cups each, poured by Prisca from three buckets that Tally had dosed to a measured residual of 0.0, 0.5 and 1.4 milligrams a litre. The cups were labelled with the colours of a chalk dot under each, rubbed out after. Nobody knew which was which, not even Prisca.
Fourteen of the twenty could not tell the 0.0 from the 0.5. Three picked the 0.5 as "a little like the borehole at the health centre, but fine." Three said they could taste it and would drink it. All twenty picked the 1.4 as the one nobody would choose. Seven called it the hospital, and one, an old woman called Bwengye whom everyone left to the end, called it "the water that was angry with me."
"So half a milligram," said Josephine.
"Half a milligram," said Tally. "It is no longer the guideline's number. It is yours, with a guideline that agrees. If the nineteen households begin to leave again, you will be the first to hear, and I will help you try one at point four."
Mzee Byaruhanga, who had taken his sample in silence and not said which he thought was which, said: "It tastes of nothing."
"That is the best a chlorine dose can do," said Tally.
There were things the tank did not promise, and Tally had put them on the plinth too, in black below the three colours, in Prisca's capital letters, at Josephine's insistence.
CHLORINE DOES NOT KILL EVERYTHING. A COMMON PARASITE CALLED CRYPTOSPORIDIUM SURVIVES THE DOSE; BOILING OR A FINE FILTER IS FOR THAT. IF THE TUBE STILL SAYS BROWN, BOIL. THE TANK IS CLEANED THE FIRST SATURDAY OF EVERY MONTH.
In the first week of November the rains settled in. The tube said brown on eleven of the first fourteen fills, cloudy on three. On twelve the cloth was changed by Prisca and her brother. Eleven times the red-ochre pouch went in, and the colour kit read 0.4, 0.5, 0.5, 0.6 at the tap in the evening, and 0.3 once, on a day when the tank had not been scrubbed.
Mama Josephine wrote it in the book on the shelf, under the date, as she did every day, and then she signed it, which she had never done before. Tally recorded the number in the log, and put beside it, in the margin where its own words went, a short line.
Nobody here knows if the nineteen households are all back. The tube is the only thing in this log that was measured by a person I could see.
On Sunday after Mass, the women queued with their jerrycans in a line that went round the plinth and out onto the road. The queue was longer than in August. Tally counted it, and did not say what it thought the number meant.