The Bullet Cluster (1E 0657-56) is widely regarded as the “smoking gun” for dark matter. Two galaxy clusters collided at a relative velocity of ∼ 4700 km/s approximately 150 million years ago. After the collision, the hot intracluster gas (visible in X-rays) was stripped and decelerated, while the gravitational lensing mass (measured through weak and strong lensing) remained associated with the galaxies and passed through. The spatial offset between the X-ray gas and the lensing mass has been interpreted as evidence that the dominant mass component is collisionless—i.e., dark matter.

This chapter presents the τ-re-read. In Category τ, dark matter does not exist (the Sector Exhaustion Theorem, the relevant chapter). Yet the Bullet Cluster observation is real. The resolution lies in the distinction between boundary mass and gas mass: the gravitational lensing signal is generated by the boundary holonomy H_∂[ω] of the cluster galaxies, which is collisionless not because it consists of unknown particles, but because boundary holonomy is a topological quantity that does not interact through the EM sector. The offset between X-ray gas and lensing mass is a boundary-mass offset —a natural consequence of the distinction between Sector D-sector (gravitational) and Sector B-sector (electromagnetic) readouts.

The second half of the chapter addresses large-scale structure. The cosmic web—the filamentary network of galaxies spanning hundreds of megaparsecs—is derived as a Wilson skeleton : the network of large-scale holonomy loops on the base τ¹. Baryon acoustic oscillations (BAO) are the periodic imprints of these holonomy loops on the matter distribution.