A ghost which is repeatedly seen by multiple witnesses in a specific place, and sometimes on specific occasions, obviously provides much better scope for investigation. It is likely, then, that this phenomenon will be the first to be fully understood. A typical 'recordings' ghost appears to have no relationship with its surroundings (it passes through solid matter, sits where there is nothing to sit on etc). Its behaviour is repetitive and limited, and it appears to have no interest in those who witness it.

These forms are generally attached to specific locations, and most 'historical' ghosts fall into this category. For instance, an apparition of President Lincoln has been seen at The White House on numerous occasions, by the most level-headed of witnesses, including Winston Churchill and John Kennedy.

The degree of perceived 'realness' varies greatly from one apparition to another. Some seem to take an almost translucent form, while others may appear solid, but still have that spectral je ne sais quois. As one witness when interviewed for TV succinctly put it: 'I got the impression it was no more of an intrusion than people on TV. Real, but only to a certain degree.' On the other hand, there are those which appear to be so thoroughly mortal that the witness initially believes that they are looking at a real person. The former leader of the Liberal Party, Jeremy Thorpe, encountered an apparition of this kind while staying with friends in Trethevy, Cornwall. Returning to the dinner table, Thorpe looked back down the corridor and saw a monk in a brown habit entering the bathroom, and later asked his hosts who the monk was. It transpired that he was known as 'the Prior' and had been in residence for several centuries.

Sometimes the supernatural nature of an apparently 'real' person can be revealed in a more surprising way. John Spencer, a researcher in unexplained phenomena, recalls an incident in the summer of 1995 which took place during the investigation of an allegedly haunted building and which involved a new and fairly sceptic member of his team. 'He looked up and saw this girl in front of him and just thought she was a member of the team that he hadn't been introduced to. He started to say hello and she just disappeared. That shook him up quite badly.' Along similar lines, Spencer chuckled at the memory of a group of witnesses who saw a ghost purported to be that of the ill-fated Lady Jane Grey. 'She was dancing through the corridors, and really happy. Everyone thought she was a genuine young girl, thought it was lovely. . . until she went straight through this iron door.'

Spencer, like most other psychical researchers, leans towards the belief that 'recordings' ghosts are a kind of natural hologram, recorded and played back by mechanisms we don't yet understand. 'I don't think,' he muses, �that if someone sees Sir Walter Raleigh, it proves that there�s life after death... I don�t think there�s anything of his �soul�, just a picture, like a video.�

The key to understanding the �recordings mechanism�, if such a thing exists, is in discovering what triggers it. Says Spencer: �Logic demands that there be a combination of things, some of which are not. For instance, if it was just a matter of a particular combination of geography - a room constructed in a certain place only, say - then everyone who went into that room or building would be recorded and you wouldn�t be able to move in there for ghosts flying all over the place. So it must be the combination of geography and atmospherics - a particular type of location and a particular humidity of temperature - which might embed or play a recording. But again, you might suppose that it would happen more frequently. So you need another rare factor and that�s where you bring in the human mind. Perhaps it takes a particular type of mind to embed the recording, and later, to press the replay button. That, I think, has a certain logic to it, because it would explain why not everything gets recorded and not everything gets seen.�

His wife Anne Spencer, who also works in this field of research, asserts that a recording might not be triggered by the person whose image is recorded, but by someone seeing them. She cites a recent case of a witness who looked out of a hotel window onto a view of a pond and saw someone lying face down in the water. �He rushed down but when he arrived there was no-one there. When he was back at the window again, he saw the same image. It suggests that the image could have been recorded by someone seeing the man in the water - and to experience the recording, you need to be in exactly the spot where the original sighter was.�