This was the first time I had clearly seen a really 'plastic bend', since these are much rarer than the slow bends I had observed previously. I do not think there can be any question of fraud when a really plastic bend is produced under close scrutiny, unless there is serious chemical corrosion, such as that produced by mercuric salts. Even then, the metal behaves quite differently, becoming wet, discoloured and brittle, but hardly plastic. Chemical corrosion is accompanied by a change in weight; therefore I was pleased that I had recorded the weight of this spoon as follows:

Original weight 24.3526 g

Weight after bend by child 24.3533 g

Combined weight of pieces after fracture 24.3529 g

These variations are within the limits of weight hanges, both up and down, which have been observed in other bent specimens. The errors are due to dirt and to moisture condensation and evaporation.

To be quite sure of such softening, one must be able to inspect it at close quarters both with the eye and by handling.

A piece of metal that has been critically weakened by bending to and fro can be held rigid at the weak point between finger and thumb. A gradual series of bends can be made apparent by suitable manipulation, but all the time the weak point is held firm, and the movements are made against the restraining force of the flesh of the ball of finger and thumb. The conjurer Mr Randi has demonstrated this trick to me, and it is most effective. It caused me to pause and think: was the Uri Geller spoonsoftening, or others that I have observed, of the same character? No, the character was not the same, in two important details: first, there had been no opportunity for Geller to weaken the spoon by working it to and fro; and second, the part of the spoon which softened and bent was clearly exposed to view, and not held firm. In fact, I was able to hold the end of the spoon myself, and sensitively probe the softening by movements of the hands against the resistance of the metal. It was an uncanny experience.

I have more recently obtained a videotape record of a vertically mounted stainless steel teaspoon in a floppy condition. Metalbender Stephen North succeeded in softening a point on the spoon without even touching it. He then gently pushed and pulled the bowl about, so as to demonstrate the floppiness. Eventually the bowl fell off.

I have obtained similar videorecords of the bending of long thin aluminium strips, which are mounted vertically by standing them on pieces of Plasticine. Stephen holds his hand close to the metal but does not touch it. Suddenly a softening at one point causes a sharp bend, the upper end of the metal strip collapsing to one side under gravity. This type of rapid metalbending event, which can be seen on only a few frames of the videorecord, is not infrequent. It almost seems as though metalbending takes place either too slowly or too fast for the viewers to have an easy task observing it. No doubt this has contributed much to the slowness of its acceptance by sceptical people.

When a strange physical phenomenon is reliably observed and even instrumented, the possible causes must be assessed. First, there may be a natural physical cause, such as the relaxation of previously developed internal stress. In many cases of metalbending, such causes have been eliminated by previously annealing the metal. Second, there may be the conscious or unconscious 'action', by means as yet unknown, of one or more humans (or animals). Third, such action may be brought about by one or more discarnate entities-'spirits' of living or dead humans (or animals) or, alternatively, deities. Finally, the action may be brought about by one or more inanimate physical objects, such as haunted houses. In principle, any one of the above, or any combination of more than one, could have been the cause of the phenomenon.

It is difficult to see what proof might be offered of the responsibility of these entities. Even if communications were received, repeated demonstrations would have to be given, with the communicated message before each one.

The most that can be expected at present is that a link with a human can be established. Uri Geller is said to cause metalbending because it happens frequently in his presence, and sometimes when he thinks that he is intending it. But the phenomenon is said to be 'spontaneous' because sometimes it occurs when he thinks that he is not intending it.

A sudden event observed in an early session of Uri Geller in my laboratory was the bending without touch of a discshaped single crystal of molybdenum of about 1 cm diameter. This had been provided by Dr Anthony Lee of the Cavendish Laboratory, and I personally kept it secure in a plastic box in my pocket before its exposure to Geller's 'action'. Physicists David Bohm, Ted Bastin, Jack Sarfatt and also Brendan O'Regan were present as witnesses. Geller asked for small metal objects to be placed for him on a large metal plate, so we placed on the table a machineshop working plate. I took the crystal from its box and put it absolutely flat on the plate. Sarfatt extended his hand a few inches above the crystal and the other objects on the plate. Geller moved his hand above Sarfatt's, until a tingling sensation was reported by the latter. Geller tried to 'concentrate his action', and it was suddenly seen by the observers that the crystal had changed its shape, and was now slightly bent, through an angle of about 20°. I could not swear that the bending was not accompanied by a tiny metallic sound. But I was absolutely certain that neither Geller nor anyone else had touched the crystal since I placed it on the metal plate; nor did he drop anything on the metal plate.

I replaced the crystal in its box, which I returned to my pocket; a physical examination of the crystal would be necessary. I eventually found that a physical property (the magnetic susceptibility) of the crystal was anomalous, but I have not previously published this fact. It will be discussed in chapter 11.

Unfortunately a precise account of this event never appeared at the time, partly because an unauthorised and imprecise account was released to the press by one of those present. Sceptics had a fieldday on the basis of this account, and even now it is doubtful if many readers who remember it will be much influenced by my own version. But I myself was impressed by what I had seen, and was reasonably certain that it could not have been a piece of conjuring. This is partly because conjurors do not know about how to change the physical properties of very pure molybdenum; nor could it have been known just what investigation I was going to make.

But it must be admitted that there were no conjurors present at our session, and that numbers of that profession have written very sceptically about metalbending. Notwithstanding this, there is a less vocal minority of conjurors who are convinced of the reality of the phenomenon. Particularly is this the case in Frenchspeaking Europe, where metalbender JeanPierre Girard has given many 'demonstrations'.

One of the deepest impressions I received came from the observation of a bend by JeanPierre Girard. I had been asked by chemical physicist Dr Wolkowski of the University of Paris to participate in a discussion on the French radio, and the next day he telephoned to say that as a result of the broadcast a young man had come forward and demonstrated that he could bend metal. Girard was employed in selling pharmaceuticals, but was interested in conjuring and had performed as an amateur. He found himself able to concentrate deeply on pieces of metal, so that eventually they would bend. Very soon there was a television demonstration, of which I received a filmed version. Girard was seen to handle metal rods which bent; it was clear that manual force was not involved. Those of us who watched the film immediately saw that it could have represented a previously bent rod being slowly rotated so that the bend appeared. Such a procedure would have to have involved the film technicians, and when I mentioned this fact in a public lecture, the television company threatened legal action. Although the film was perhaps suspect, I determined to see Girard for myself. The opportunity came as late as July 1977, when I monitored a similar filming in Paris for Alan Neuman and the National Broadcasting Company of America (NBC). We made sure that the protocol was tight; large identified aluminium alloy bars were provided by the Pechiney Aluminium Company, and a suitable one was chosen on camera; after a filmed bending, it was returned on camera. The subsequent tests showed that the manual force that would have been required to make this bend was equivalent to 17 Newton metres (Nm) moment. The average limit of human strength under these conditions is 25 Nm, but a case has been reported of a man achieving a 38 Nm bend by force. The witnesses were unanimous that manual force was not used, and the published film shows this clearly (NBC holds the complete footage of the bending, and a report from me). Girard's pulserate was monitored during the event, and at the critical period it was as high as 160 beats per minute. The entire proceedings took place with Girard sitting at a flimsy glass table, which could very easily have been broken if strong force had been used on the metal bar.

It is never easy to achieve a metalbending event under the conditions required in the television studio. The delays and adjustments which are necessary to the production contrive to develop an atmosphere of excitement, impatience, confrontation and even stagefright. If video evidence is desired, it is much better to allow a child metalbender to accustom himself to the equipment for several days in his home, and practise metalbending in view of the camera, with one of the family operating the equipment. I have then been able to visit and monitor (for example, from Stephen North and Julie Knowles) the recording of stroking bends, and of bends without touch; also of strain gauge records of the type described in chapter 4.

One very unusual record was obtained by my colleague David Robertson of the action of the metalbender Julie Knowles on a very long thin strip of aluminium (50 cm X 8 mm X 0.8 mm). Such strips can be waved about flexibly and are bent very easily, so I have used them for practice and encouragement of the metalbending children whom I have studied. Julie held the strip upright by one end and remained still; it soon became apparent that the top end was forcing itself towards her, and then springing back again. She was able to press the strip quite hard against this invisible force, and also to move the strip about and turn it to the horizontal; it was obvious from the videorecord that the 'invisible force' was not in fact produced by a thread or 'moti', as it is called by Indian conjurors. Eventually Julie forced the strip sufficiently hard against the 'invisible force' to bend it right over in a smooth arc of nearly 180°.

Here is a final example which illustrates well the spontaneity, even capriciousness, of metalbending happenings; also the difficulties that we encounter in bringing ourselves to come to terms with them.

During 1978 a Japanese boy, Masuaki Kiyota, was invited by a television company to visit London and attempt to 'demonstrate' some optical phenomena of the type to be described in chapter 22. I decided to examine his metalbending, and brought with me two freshly purchased stainless steel teaspoons and two of my own household dessert spoons; all of these I quickly identified by nicking their bowls in various places, using wirecutters. I also traced their outlines on paper and located their magnetic poles (chapter 11). I kept them securely in my pockets during Masuaki's visit.

On the first day I pulled out a dessert spoon from my pocket, quickly checked the identification by touch and offered it to him during a break in the filming at Birkbeck College; he straightway placed it in his lefthand trouser pocket and withdrew his hand; I could see a bulge, presumably (but not certainly) made by the spoon. No one else who might have received the spoon from him came near us. He then asked to be shown to the toilet, so I took him there myself, unaccompanied. Before we had reached the toilet, while we were walking together along the basement corridor of the College, Masuaki pulled a spoon from his pocket and gave it to me; there was a single 180° twist in its neck, which appears in Plate l.la (2). By touching the nick in the bowl, I found that it was my spoon, and I kept it in my hands until I returned it to my pocket. Of course I did not see the twist take place, but my observation was that the necessary tools (such as hand vices or wrenches) could not have been used on his person without my noticing; therefore I deduced that Masuaki's twisting of the spoon, whether carried out with two hands, with one, or with none, must have involved local softening of the metal.

Similar twisting of the spoons shown in Plate l.la (1 and 3) took place during Masuaki's visit, but I was not personally responsible for observing them. The last teaspoon I continued to keep in my pocket, and after checking the identification I offered it to Masuaki two days later in a taxi while I was sitting beside him. He played with the spoon onehanded, all the time in my field of vision; within two minutes it became twisted (Plate l.la (4)). It happened too fast for me to get a really clear visual image of the happening. I again checked my nicking and returned it to my pocket.

The identification of a spoon by touching the nicked bowl is easy and quick, perhaps embarrassingly so. Later during the day I pulled a teaspoon from my pocket without checking it and showed it as evidence of the taxi event to another investigator. I had earlier explained my identification technique, and he was able to see that it was in fact another teaspoon altogether (Plate l.la (1))! I pulled out the 'taxi' teaspoon and this time there was no trouble with the identification. My face was undoubtedly red, but this is no reason why the observation should be invalidated.

I describe these events in order to show the difficulties of observing spontaneous phenomena. Had Masuaki been sat down in front of the television cameras and invited to bend metal, there would not have been a very high chance of success. But Nippon Television in Tokyo have during 1979 achieved some good videorecordings of twists.

My description of metalbending events will not in subsequent chapters be arranged in correct historical order, but rather by classification of the different types of physical phenomena and the experiments necessary to obtain some understanding of them; these 'notouch' experiments have mostly taken place in the homes of metalbending children, to which I have carried portable equipment, attempting to approximate to 'laboratory conditions' without removing the child from his natural environment.

The chapters will read as though the sessions have been one continuous success story, but in fact this has not been the case. There have been many long hours of observations without any 'paranormal effects' to show for them. My policy has been to set up the most suitable social and psychological conditions for metalbending to occur. I have no detailed recipe for success, nor do I consider that public demonstrations are likely to be successful, except in rare cases.

Plate 1.1a (Opposite above) Stainless steel spoons twisted by Masuaki Kiyota Plate 1.1b (Opposite left) Enlargement of twist in spoon 1

Plate 1.1c (Opposite right) Enlargement of twist in spoon 4