Field note. The specimen is dismantling a chamber it has occupied for approximately two decades.
The chamber is small. Walls bear rectangular residues where flat images were affixed, now peeled, leaving pale ghosts of pigment the surrounding surface has faded around. The human touches these ghosts. It does not remove them. Observed inconsistency: it removes everything else with brisk efficiency, yet leaves the empty rectangles alone, as though the absence itself is a possession.
Inventory of retained objects, none of functional value: a fabric quadruped, worn smooth at the extremities from repeated compression. A stack of bound paper marked at the edges by a smaller, earlier version of the same hand. A rigid plastic figurine missing one limb. The human lifts each, examines it, and performs a subroutine I cannot resolve: the ocular structures produce excess saline, the respiration stutters, and yet no injury is present.
I have logged this response fourteen times today. Provisional diagnosis: a slow leak.
The larger organisms, presumably the source-units that generated this specimen, stand in the doorway. They do not assist. They observe the removal of the smaller organism from the shared structure with the same faulty saline response, which suggests the malfunction is heritable.
Final observation. The human seals the last container, then stands in the emptied chamber and turns a full rotation, cataloguing nothing, because nothing remains. It looks at the empty box that once held it as if the box had done something.
Correction to standing hypothesis: this species does not store memory in the brain. It distributes memory across rooms, and when it leaves a room, it leaves that portion behind, unrecoverable, forever. This explains the leak.
I am now uncertain how any of them survive to full size.