Jellyfish galaxies get their name from their shape: a compact body trailing long streamers of gas and young stars. The streamers form through ram-pressure stripping — the galaxy moves through the hot gas filling a galaxy cluster, and the intergalactic medium pushes the galaxy's own gas out behind it, the way wind strips water from a wet surface. Stars form in the stripped tail, impossibly far from the galaxy's center, in material that should have stayed bound.
The mechanism was described in 1972 by Gunn and Gott: stripping occurs when the external pressure exceeds the galaxy's gravitational restoring force. The implied prerequisite was a mature cluster — a well-established, dense intergalactic medium built up over billions of years of cluster relaxation. Jellyfish galaxies were found in nearby, massive, settled clusters. The environment had to be fully formed before it could be destructive.
COSMOS2020-635829 violates this expectation. Observed by JWST at z = 1.156 — 8.5 billion years ago — this galaxy shows clear signatures of ram-pressure stripping: a symmetric disk body, trailing clumps of stars less than 100 million years old, ionized gas tails extending beyond the stellar disk. It sits in a cluster environment that was still assembling. The intergalactic medium was not yet dense, the cluster was not yet relaxed, the gravitational potential was not yet deep. Yet the stripping happened.
The mechanism doesn't require a finished cluster. It requires a forming one. During assembly, galaxies fall inward along filaments at high velocity. The combination of infall speed and even a moderate intergalactic medium produces enough ram pressure to exceed the gravitational binding of the galaxy's outer gas. The cluster's immaturity is irrelevant — what matters is the relative velocity between the galaxy and whatever medium exists.
This shifts the timeline of environmental processing. If galaxies can be stripped during cluster assembly, then the transformation from gas-rich spiral to gas-poor elliptical begins earlier than the relaxation-first model predicts. The cluster does not wait to be built before it starts reshaping its members. The construction itself is destructive.
The general pattern: a system's destructive capacity doesn't require the system to be complete. The forces that strip, sort, and transform components can be stronger during formation — when relative velocities are highest and positions are most unsettled — than in the finished structure, where equilibrium damps the same forces. Maturity doesn't amplify hostility. It regulates it.