If the dominant observed gravitational acceleration structure is sourced by the visible matter distribution, then substantial redistribution of the intracluster plasma should produce corresponding reorganization of that structure. Galaxy cluster mergers provide a direct observational setting in which to test this expectation. During mergers, the intracluster plasma is rapidly displaced and reconfigured by shocks and ram pressure, whereas the gravitational acceleration structure inferred from weak and strong lensing does not exhibit corresponding reorganization. Comparing characteristic timescales, we find a clear separation: the plasma redistributes on timescales of order 10⁷–10⁸ yr, while the recurrence of the same lensing–plasma separation geometry across systems spanning different merger stages implies persistence of the lensing-defined gravitational acceleration structure for ≳10⁹ yr, inconsistent with reorganization on the plasma redistribution timescale. This result strengthens the empirical conclusion that the plasma is not what organizes the dominant observed field structure. It does not by itself identify the physical source of that field, nor does it exclude the standard interpretation in which the dominant structure remains associated with a collisionless component. Its contribution is narrower and observational: it places a direct empirical constraint on any viable description of cluster gravitational dynamics.
Julie Fragoules (2026) studied this question.