The Leaf Holds Its Water, Then Lets It Go

Water does not leave a withering leaf at a steady rate. In one batch monitored hour by hour, it very nearly stopped for eight hours in the middle, then left in a rush at the end. Between the sixteenth hour and the twentieth, the leaf shed about eight percent of the water it still held. Between the twentieth and the twenty-fourth, it shed forty-two percent.

The Trust has described this step, for a hundred and forty years, as waiting. That has always been the honest word for it. What was not known, until the wither was measured closely enough, is that the waiting has a shape.

The register of one batch

A 2026 study in Horticulturae¹ followed a single withering batch at Fujian Pinpinxiang Tea Co. in Fuding. The room was held at 25 degrees Celsius (about 77 degrees Fahrenheit), plus or minus one, and at 65 percent relative humidity, plus or minus five. The wither ran thirty-six hours. Samples were drawn at thirteen points across it, 650 in all, and the moisture content was entered each time as a fraction of the leaf's fresh weight.

The entries read like this.

At the start the entry is 0.781. By the twelfth hour it reads 0.547. The leaf had given up roughly thirty percent of the water it began with, and it had done so briskly.

Then the register slows. From twelve hours to sixteen it moves 0.547 to 0.475, a further thirteen percent. From sixteen to twenty, 0.475 to 0.438, a further eight percent, four hours for the smallest movement in the whole record.

Then it turns. From twenty hours to twenty-four, 0.438 to 0.252: a reduction of about forty-two percent of what was left, in four hours. From twenty-four to thirty-six, 0.252 to 0.106, another fifty-eight percent, the fastest relative loss on the sheet.

The shape of it is fast at first, then a long flat stretch through the middle, then most of the remaining work compressed into the back third.

What the entry does not cover

This is one batch, in one room held steady at 25 degrees, which is a regulated wither and not the sky's. That matters, and the Trust would rather say so than let a good measurement be borrowed further than it goes.

The three traditional ways of holding the leaf run on quite different clocks. Sun withering takes two to three days. Indoor natural withering runs fifty-two to sixty hours. The heated room finishes in about a day. Our withering guide sets out all three, along with the chemistry of the wait and the twenty percent moisture reading that ends it. A thirty-six-hour batch at a fixed temperature and humidity is closest to the last of those, and the shape of its curve is not a law the sun is obliged to obey. A cloudy afternoon or a dry wind redraws any of these numbers.

The finding worth keeping is the shape, not the figures: a long plateau, then a rush. It has now been measured once, carefully, in a room where nothing else was allowed to vary.

One more correction belongs in the same entry. The leaf looked still through those middle hours, and it was never still. Cell walls are breaking down, the oxidase enzymes are out, the polyphenols are falling and the amino acids rising, and that account has been on the record for years; the withering guide carries the figures. What is new is narrower and, to the Trust, more interesting. By the one measure the maker is actually waiting on, the measure that decides when the leaf comes off the rack, there is a stretch in the middle where almost literally nothing happens.

The maker's job through the plateau is to arrange the conditions, watch, and decline to intervene. That has always been described as work. It now comes with a measured interval in which the leaf, weighed, is doing very little to earn the attention, followed by one in which it does everything.

The study that produced these entries was not built to describe the curve. It was built to read moisture through the leaf, with near-infrared light, without destroying the sample to weigh it. The register can be kept now without taking the leaf off the rack.