The south branch sediment had taken forty minutes to dry on the wire rack before Sonmat realized it was not drying evenly.
This was in October, the morning after the Chicago site visit, in the studio off the Process Quarter's eastern corridor. The wire rack was set up beneath the south-facing window, the sediment spread in a thin layer across a cotton mesh backing — fine-grained material, silt mostly, with occasional coarser particles that Sonmat had decided to leave rather than filter out. Watching the drying happen, Sonmat saw the finer particles go pale at the edges while the coarser grains held their moisture in small dark spots. The coarser particles were not yet dry. They were drying on their own schedule.
This was the first thing the material taught: the south branch sediment from the bend downstream of the former rail yard access was not a single material. It was several materials moving together through the same water, deposited together at the same bend, collected together in the same container. Each particle had a different relationship to moisture, to light, to the wool fiber waiting in the studio drawer. Sonmat sat with the wire rack and watched the drying complete itself unevenly and did not try to make it uniform. The unevenness was information. The particles would felt differently. The finished panel would hold the record of how each part of the south branch had met the wool on its own terms.
The felting took three weeks. The finest silt dispersed almost immediately into the fiber — worked in with slow circular pressure and disappeared, became part of the wool without resisting. The mid-range particles required more agitation: circular motion, then directional, then back to circular, the material gradually releasing its resistance and accepting incorporation. The coarsest grains had to be worked individually into areas of concentrated fiber, teased in and pressed until the wool closed around them and held. Sonmat's hands were stained rust-orange from the iron-rich sediment for the duration of the work.
The finished panel was 36x48cm. Warm-toned, close-piled, the surface even in the areas of fine silt distribution and slightly variable in the areas of coarser incorporation — the variations small enough that you had to run a hand across the surface to feel them, but present, consistent, the same across the whole panel. The south branch had been generous to the felt. It had given itself up with relatively little resistance.
The residency application asked for a materials list and provenance statement. Sonmat wrote: Wool (undyed, standard weight). Chicago River sediment, South Branch, collected October 2025 at the bend downstream of the former rail yard access. 2:17 PM, low-flow conditions following three dry weeks. The process statement is the work.
The application was submitted with two minutes to spare. Sonmat stopped thinking about whether the committee would understand it.
The Felt world archive's eastern holdings had a panel from the same river bend: South Branch, Chicago River, 2011. Sonmat had looked at it three years ago, cataloguing, without thinking much about it. In November, after the residency application was in, Sonmat went back to look at it again.
The 2011 panel was darker than the 2025 panel — more organic matter in the sediment, the wool fiber holding more of the river's biological content from that particular year. 2011 was three months after the canal reversal correction, a remediation event that the archive's field notes recorded as significant. The river had been carrying disrupted sediment — stirred up, mixed, material from layers that hadn't moved in decades brought to the surface and carried downstream. The 2011 panel held that disruption in its color. The south branch sediment from that year had not yet settled back into its longer-term pattern.
Fourteen years of recovery separated the 2011 panel and the 2025 panel. Both panels were from the same bend of the same river. The difference between them was the river's own record of what those fourteen years had done — the remediation, the recovery, the slow return of fine silt to a bend that had been disturbed and had gradually restabilized. The archive did not explain this. The archive only held the panels and noted when and where they were collected. The explanation was in the comparison, and the comparison was available to anyone who knew what they were looking at.
Sonmat thought about this often during the four weeks of working the north branch panel.
The north branch material had been stored in a separate container since the field collection, labeled in the field with a marking pen. Sonmat had not opened it between October and February. In February, Sonmat spread a small sample on a white ceramic tile and looked at it under the studio light.
Grayer. Larger particles in higher proportion. The north branch collection point had been further from the main sediment deposit zones — more active flow, less settlement time, the water still carrying more of what it had picked up upstream rather than dropping it here. Where the south branch had given Sonmat fine silt with a high iron content, the north branch gave coarser material with a lower organic fraction: gravel-sized inclusions mixed with medium sand and a minority of fine silt. A different stretch of the same system, deposited by the same water moving at a different speed over a different substrate.
The north branch felting took four weeks, and the fourth week was the hardest.
The fine fraction incorporated the way the south branch silt had, disappearing into the wool. The medium sand resisted longer — needed sustained directional agitation before the fiber would accept it, and even then the acceptance was grudging, the wool reshaping itself around the particles rather than fully absorbing them. The coarser inclusions did not incorporate in the way the south branch's coarser particles had. They were held rather than embedded, the fiber pressing around them and holding them in place but not fusing with them the way the softer south branch material had fused. The north branch material stayed distinct inside the felt in a way the south branch material had not.
The finished north branch panel had a different surface entirely. Areas where the fine silt had incorporated smoothly were close-piled and calm, similar in character to the south branch panel. But the areas of medium and coarse material were raised slightly, the fiber bunched around the larger particles, the surface varied in a way that was visible at distance and unmistakable under a fingertip. The north branch had not become the felt the way the south branch had. It had been contained by the felt. It was still, in some way, the north branch.
When both panels were dry, Sonmat set them side by side on the studio floor and crouched to look at them at surface level.
The color difference was subtle standing up: the south branch warmer, the north grayer, both recognizably from the same river. At surface level, the texture difference was structural. Running a hand across the south branch panel and then the north branch panel was not comparison — it was two encounters with two different materials that happened to share a drainage system. The south branch had given itself to the felt. The north branch had agreed to be held.
They were not a diptych. A diptych implies two instances of the same kind of thing set beside each other for comparison. These panels were two entries in a series that needed more entries before the comparison it was building would be legible. The industrial bend of the main channel, where the remediation had changed the sediment profile since the 1920s and where the archive held panels going back to 1987 — those panels would be the third entry, and the comparison between them and the south branch and north branch panels would start to describe something about how different parts of the same water system carry different histories in what they deposit. The North Shore Channel, engineered flow rather than natural, managed for navigation rather than ecology — its sediment would be a fourth kind of record, the river's system under deliberate management rather than gradual recovery. The river mouth, where the whole system accumulated before dispersing into the lake — a fifth, the river's summary in material form.
Five panels was still not the river. It was five moments of the river made permanent in a medium that would outlast the conditions that had produced them.
Sonmat photographed both panels together on the studio floor and sent the image to no one.
The residency decision would come in April. If it came through: three more collection points, three more panels, a series of five. If it didn't: two panels, a beginning, and the north branch container still labeled in the hand from the field collection the previous October, waiting for whatever came next.
Either way, the river was still where it had been. The south branch was in the felt now. The north branch was in the felt now, held in the fiber in the way the north branch held things — not dissolved, not absorbed, but contained. The river had given what it had given. The giving was permanent in a way the river itself was not.
Sonmat picked up the north branch panel and turned it over. The back was rougher — the agitation showed more clearly on the reverse, the fiber direction less controlled, the coarser inclusions visible as irregularities in the backing. The back was the process. The front was the record. Both were the same felt.
The south branch panel went on the wall first, on the left. The north branch went beside it, on the right. The space between them was room for what came next.
The felt did not know it was the beginning of something.
Sonmat would wait for April.