Let us say we have a type I source of entangled photons (so that we get a positive correlation if the polarising analysers are aligned). After leaving the source both photons are passed through a vertical polarising filter and continue in opposite directions. Anne is closer to the source and has a polarising analyser orientated at 45 degrees to the vertical. Bob is in the opposite direction and further away from the source than Anne. Bob's analyser is horizontally orientated. Ordinarily without entanglement, we would expect Bob to make no positive detections at his horizontal analyser, because the photons going towards him have already passed through a vertical polariser. However, since his polariser is orientated at 45 degrees relative to Anne's polariser, the entangled expectation is that Bob will get a 50% correlation with the positive detections made by Anne. The problem is that Anne could signal super-luminally to Bob by rotating her analyser. How is this situation avoided? Does placing two vertical polarisers near the source destroy the entanglement?