diff --git a/docs/formats/core-data-regions.md b/docs/formats/core-data-regions.md
index 7f88d98..c79967e 100644
--- a/docs/formats/core-data-regions.md
+++ b/docs/formats/core-data-regions.md
@@ -2184,3 +2184,64 @@ it in the file -- consuming bytes 4..8 and shifting everything after. The struct
 literal named the fields in the right order; the statements did not. It is fixed
 with a comment saying so, since "read the `let`s in file order" is the kind of
 thing that gets undone.
+
+## Volumes: a candidate table, and why it is not yet called one
+
+Volumes were the only asset type never located at all, so this is a search
+report rather than a result.
+
+**Ruled out first: there is no engine-side name.** No level file contains the
+string `Volume` outside UI text ("Music volume", "Volume de la musique") and
+Italian item descriptions. So the format cannot be identified by a label and
+must be found structurally.
+
+`gameplay.bin` on disc is the already-decompressed level gameplay data. Its head
+is a count (`336` on level 0) and then 16-byte records from byte `416`. Those
+records carry `(type, pad, offset, id)`; the `id` field groups them, each id
+having exactly two records pointing into two parallel offset arrays. The offsets
+they point at resolve into a text pool -- the first ones land mid-string in
+dialogue ("it Megacorp Lab"), so this is the dialogue and class-dispatch side of
+the file, not geometry.
+
+The structural hit is a table of **144-byte float records**. Level 0 has one at
+`799204`:
+
+    20 records of 144 bytes, 19 populated and one all-zero terminator
+    34-35 of every record's 36 float slots are world-plausible
+    +0x0c holds a world position, varying per record:
+
+      [0] (206.5, 284.1, 116.0)     [6] (244.0, 407.2, 128.0)
+      [1] (219.3, 247.0, 107.5)     [7] (265.3, 443.7,  13548.6)
+      [2] (248.2, 198.4,  51.8)     [8] (185.1, 282.6, 104.9)
+      [3] (247.2, 203.6,  82.4)     [9] (185.8, 282.7, 102.6)
+      [4] (155.7, 372.7, 121.1)    [10] (260.8, 423.6, 130.9)
+      [5] (265.0, 445.3, 131.7)    [11] (247.0, 211.5, 100.6)
+
+A 144-byte record with world-space positions and float radii is exactly the shape
+a volume trigger region should have, and a matching table exists in every level
+examined (levels 0-9), each preceded by an all-zero record, with counts from 23
+to 612.
+
+**It is not called volumes here.** Nothing yet ties this table to volumes
+specifically. It could equally be camera bounds, spawn points, or path nodes --
+all of which would present 144-byte float records with world positions and no
+label. The distinguishing evidence has not been gathered, and the honest state
+is *candidate*.
+
+### Two detector bugs worth recording
+
+Both were mine, and both produced confident nonsense before being caught.
+
+1. The first validity test accepted zero as a world-plausible float. Every
+   zero-filled region then scores perfectly, and the cross-level check returned
+   runs of 225 to 1,514 records -- obviously padding. Requiring **non-zero**
+   floats (>=20 of 36) immediately separated real tables from memory.
+2. The stride claim rested on four eyeballed offsets 144 apart. Gap 144 was not
+   in the top eight gaps of the detector that found them; it survived because
+   three consecutive observations agreed, not because the measurement supported
+   it. The corrected detector then found longer runs, which is the opposite of
+   what the eyeball version claimed.
+
+The same shape as every other correction in this log: a clean, small, plausible
+number read as conclusive because it was not checked against a control that
+could have disagreed.
diff --git a/docs/roadmap.md b/docs/roadmap.md
index b71b71d..e95716e 100644
--- a/docs/roadmap.md
+++ b/docs/roadmap.md
@@ -407,7 +407,17 @@ Phase 1 hands over the extents; this phase works out what each one contains.
       `+0x20` is self-limiting: entries run while strictly inside the region.
       **The word at `+0x0c` is not a count** -- it disagrees with the table on
       most levels and reads 196,617 on level 2 against 4 sub-objects.
-- [ ] `volumes` — **not located.** Nothing examined so far presents itself as a
+- [~] `volumes` — **still not identified, but a candidate exists.** Every level's
+      `gameplay.bin` carries a table of **144-byte float records** with world
+      positions at `+0x0c`, preceded by an all-zero record (levels 0-9: 23 to 612
+      records). That is the shape of a volume trigger region and it is not tied
+      to anything else in the file -- but camera bounds, spawn points and path
+      nodes would look identical, and there is no engine-side `Volume` string to
+      discriminate with. **Recorded as candidate, not volumes.** Two detector
+      bugs on the way are written up: a validity test that counted zero as
+      world-plausible scored zero-filled memory perfectly, and the 144-byte stride
+      rested on four eyeballed offsets that the gap histogram did not support.
+      Not located. Nothing examined so far presents itself as a
       volume: `core_data`, the chunk region and `gameplay.bin` all yield
       structures that are accounted for by something else.
 - [~] `level_overlay` — header, constant block and level data parsed