Which is odd, since the proof is trivial. Take a worn chain, lay a sample length flat on a surface and pull the free ends apart. It will be longer than a new chain. Now push the free ends together, keeping the chain straight. It will be the same length as a new chain, or fractionally shorter. The difference is the free play in the bearings, and the fact that it can be pushed back to its original length shows that the plates themselves have not got any longer.
Which is odd, since the proof is trivial. Take a worn chain, lay a sample length flat on a surface and pull the free ends apart. It will be longer than a new chain. Now push the free ends together, keeping the chain straight. It will be the same length as a new chain, or fractionally shorter. The difference is the free play in the bearings, and the fact that it can be pushed back to its original length shows that the plates themselves have not got any longer.