Birkbeck College (University of London)

Based on a paper delivered at the Second International Conference of the SPR at Cambridge, March 1977

The detection of dynamic strain signals in paranormal metal-bending “action”(1) has enabled experimentation to be carried out on the distribution of the action around the metal-bender. A resistive strain gauge is mounted on or in a small metal specimen, which hangs from its screened electrical connections; electrical bridge, amplifier and chart recorder enable dynamic strain data to be collected. It is found that the “action” occurs in bursts of extension or contraction, rather randomly distributed in time, often at a rate of about 50 per hour.
An early result of experiments with several independent embedded resistive strain gauges was the finding(2) that dynamic strain pulses were frequently registered simultaneously on two entirely separate metal specimens. The data were interpreted in terms of a crude physical model of a “surface of action”, at points on which strain pulses occur. The configuration and movement of this surface with respect to the metal- bending subject can then be studied.
A further series of studies (3) was conducted with several independent strain gauges on a single piece of metal. The linear array of gauges along a thin strip of metal respond simultaneously to the ”action”, but the signals are strongest in the centre of a “region of action”, which extends about 10 inches, and can move slightly from event to event. A series of experiments was also conducted with strain gauges on opposite sides of the metal strip. For a “pure bend” the signals should be equal and in opposite directions, but these conditions were only obtained in the limit of infinitely thin metal strip; for strips of thickness in the region of 1 cm, one strain gauge receives a much smaller signal than the other, as though the penetration was incomplete.


Since there is no very strong reason for expecting the signals to be of magnitude or sense uniformly distributed across the thickness of the metal, it appears important to investigate the profile of strains in thick metal specimens. We therefore mounted six sensors throughout the thickness of a metal specimen, and connected them to six separate battery-operated bridges and amplifiers, using two synchronised three-pen chart recorders. Two different methods of fabricating the sensor were used: in one the strain gauges were mounted inside slots milled in the metal; in the other the strain gauges were stuck to metal strips, which were interleaved with spacers stuck together with epoxy resin; both methods are detailed in Figure 1.
1. Cross-sections of metal specimens a) Laminar, and b) Solid eutectic alloy, containing six strain gauges for hpa1investigation of dynamic strain profile.
Three experimental sessions were held with 13-year-old Stephen North, and simultaneous strain signals were recorded at six strain gauges. The thick metal strip was suspended horizontally radially in front of the subject.
The subject was seated and was able to reach forward his hand towards the metal strip; normally he would hold the left hand within a few inches of the end closest to him, but part of the time he did not hold out either hand, and many signals were observed during these periods. One author would superintend the working of the equipment, leaving the other free to witness the action without touch produced by the subject. Members of the subject’s family were often present. The normal time for an experimental session was 60 to 90 minutes.
When a metal strip is bent, either elastically or plastically, the convex face experiences extension (positive strain), whilst the concave face experiences contraction (negative strain). At the curved neutral surface, in the centre of the strip, there is no change of strain. If a positive strain were represented by an arrow of length proportional to its value pointing to the right and a negative strain by an arrow pointing to the left, then the normal bending situation could be represented by a series of arrows, as shown at the left-hand side of Figure 2.
2. Typical profile of dynamic strain signals across the thickness of a metal specimen. Also shown at the left is the profile that would result from a pure bend. Signals correspond to quasi-forces of about 20 gm weight.
If an array of synchronous paranormal bending signals represents a pure bend with the top face convex, then they could be shown graphically as a set of data points extending diagonally from bottom left to top right. A pure extension would be represented by a set of data points situated vertically one above the other.
The arrays of signals recorded in experimental sessions are more complicated than the above. A selection of typical arrays, actually a series from session S, are shown in Figure 2.hpa2
A simple classification of arrays may be made on the basis of the number of changes of direction in the line joining the data points. Arrays with no changes are bends; those with one change are symbolised by the letter V, those with two changes by N, those with three by W and those with four by M.
In Table 1 the numbers of signals arrays of each type are listed. The action inside the metal is more complicated than might have been supposed. The phenomenon might more appropriately be described as paranormal “metal-churning” than “metal-bending”. A strain is locaised to a depth of less than a millimetre, and may often be accompanied by a strain in the opposite sense at a neighbouring strain gauge.

Table 1. Numbers of signal sextets of different profile complexity in Stephen North sesssions.
Profile TypeSpecimen typeabbNumber of Gradient Changes
Bend or stretch 51110
V 191021
N 111182
W 418113
M 1524

It is of course important to verify by other experiments that the signals do not arise from paranormal action on the electronic equipment or even on the pen of the chart recorder. Two subsidiary experimental programmes were mounted for this purpose. In the first a galvanometer mirror was mounted on a very thin spring steel strip, with strain gauge attached. One end of the spring was attached to a horizontal surface under a glass dome, and an optical beam from a helium-neon laser passed through the dome and was reflected through it again from the mirror and onto a scale. The optical path was about 5 m. Small movements of the light spot were seen to synchronise with strain gauge signals, and some ringing was observed, due to the long- period mechanical resonance of the system.
In the second experiment a dummy strain gauge was included with real strain gauges on a metal strip. Typical signals were observed on the real strain gauge chart record throughout the session, but no signals at all were hpa3recorded from the dummy gauge. A resistive thermal sensor (Micro- Measurements type STG 50D) of 70 Ohms resistance was connected with series compensation in one channel of the electronics. In physical appearance the thermal sensor is very similar to the resistive strain gauge (Micro-Measurements type EA09 125 BT 120); but its resistance is insensitive to strain although highly sensitive to temperature. Sudden temperature changes are unknown to us in paranormal metal-bending sessions, although temperature drifts usually occur, arising from environmental causes. Paranormal strain signals, on the other hand, are sudden, in that they are sharp-fronted pulses.
Although these experiments have vindicated the interpretation that the paranormal action is an internal strain in the metal or strain gauge or both, it must be stressed that dummy gauge techniques have not been used as standard practice in all the experiments reported previously. It has been our custom to introduce a dummy gauge at irregular intervals, averaging one session in ten.


hpa4We now describe the use of the strain gauge as a detector of directional effects in paranormal metal-bending. The experiments on the distribution of signals along the length of a metal strip(4) had been conducted using strips of width 8 mm, only just sufficient to mount the strain gauges. The assumption was made in analysing the data that the extensions and contractions were directed entirely along the long axis of the specimen (typically 40 cm x 7.5 mm x 0.75 mm). It was decided to test whether this assumption would be valid on wider metal strips. We therefore experimented with a strip just sufficiently wide for a strain gauge to be mounted across its width; the dimensions were 13.5 cm x 18 mm x 0.75 mm. Signals were in fact recorded from action on this strain gauge, but they were much smaller than those recorded on the strain gauge mounted along the major axis. When a wider strip was used, larger signals were recorded. It was therefore decided to experiment systematically on the directions of the dynamic strain vectors in sheet-metal specimens.
On long thin specimens, with high axial ratio, there could be a psychological factor which favours the production of strain vectors directed along the major axis. This factor, whose existence has not been proven, might be investigated on a subsequent occasion; but in the first experiments, the safest course would be to investigate a round or square disc specimen in which the only psychological factors would be the orientation of the disc with respect to the subject and to the horizontal.
There is also an important physical factor which affects the strain. When tension (positive stress) is applied along a diameter of a circular disc, an extension (positive strain) will be recorded along the diameter, but an equal contraction (negative strain) will be recorded along the orthogonal diameter. A more complicated stress produces strains of different magnitudes along orthogonal diameters. The strain tensor has many components, but it is possible with three strain gauges to determine two orthogonal components.
Figure 3 shows the arrangements of the sensors and defines the angles and magnitudes of the strains. A solution of the problem of determining the magnitudes and angles theta and theta + 90 of the extension and contraction vectors Ie and Ic from the vectors I1, I2 and I3 recorded at the “rosette” of strain gauges has been given by Mr. Jankowski. The equations appear with Figure 3. Directions of strain gauges 1, 2, 3 mounted in rosette. Equations connect the measured signals I1, I2, I3 with the extension Ie and orthogonal contraction Ic.
Two different configurations of the disc with respect to the subject have been used. In the first the disc is hung hpa5vertically and radially in front of the subject. In the second stiff electrical connections are strapped upon the bare forearm of the subject, who is presented with another target for his “action”. He normally extends his arm to hold his hand about six inches from the target, and the disc rests slightly above the hairs on the forearm.
Each triplet of dynamic strain pulses represents a positive or negative strain in a direction theta and a corresponding strain at theta + 90 These may be represented on a diagram by two radial lines of lengths corresponding to the strain magnitude. The signals during a session then appear as in Figure 4, which may be termed a “star diagram”. We can see by inspecting such a diagram just what was the directional character of the signals in the session.
Figure 4. “Star diagram” for Stephen North session N. Solid lines represent extensions, broken lines contractions. Lengths represent magnitudes; numbers indicate the session numbers of the two corresponding orthogonal signals. Calibrations show that the largest signals correspond to a quasi-force of about 100 gm weight.
In the sessions with Stephen North there were no very strong preferences of direction. In particular we note session U, in which a square metal disc was suspended vertically from one corner. As appears from the histogram of Figure 5b, there is no particular preferred direction, neither horizontal, vertical, nor parallel to either side of the square.
Similarly, there is no particularly preferred relationship between the magnitudes of the corresponding extension and orthogonal contraction signals; for a circular disc suffering a single radial stress vector we would expect the corresponding signals to be approximately equal and of opposite sign. The histogram of ratios, shown in Figure 5a, demonstrates the absence of preferred ratio. Indeed there is a certain number of pairs of corresponding signals which are of the same sign, that is, both extensions or both contractions. These are not included in Figure 5a. Such signals arise from a stress more complicated than a single radial vector, and consist of at least a pair of orthogonal vectors (i.e. simultaneous pulling or pushing by two pairs of hands). The proportions of such signals are to be found in Table 2.
5. a) Histogram of ratios of corresponding extension to contraction signals obtained during Stephen North sessions. b) Histogram of directions of signals (irrespective of their magnitude) obtained in session SNU.

Table 2. Statistics of Directional Vectors
SessionShape of specimenRemarksNumber of extensions accompanied by contractionNumber of extensions accompanied by extensionNumber of contractions accompanied by contraction
SN NSquare side 5 cm 3340
SN OSquare side 5 cm 1858
SN USquare side 5 cmTwo permanent deformations3743
GC 2Square side 5 cm 250
SN ZCircular 3.7 cm diam.Arm mounting1140
SN CCCircular 3.7 cm diam.Arm mounting 1 sensor fractured1160

It would appear from both types of experiment that there is turbulence in the strain, localised to distances of the order of a millimetre or less. Perhaps on same occasions there is action on the strain gauges alone, hardly penetrating into the metal. As has already been noted, the phenomenon might be termed “metal-churning” rather than “metal-bending”; and it is possible that the churning takes place on a much smaller scale than we can resolve with arrays of strains gauges. It requires metallurgical investigation of the structure to settle the question; but early investigations (5) disclosed no significant differences between paranormally and mechanically bent metal specimens.


The first well-substantiated claim that paranormal action can produce structural change was made by Crussard and Bouvaist,(6) who investigated the adult French metal-bender Jean-Pierre Girard. Probably the most significant report they published was that physical change could be induced in aluminium alloys AU4SG-T351 and AU4G-T351, (Aluminium with small proportions of copper and magnesium, heat treated in a specified manner, for use in supersonic aircraft.) without actual bending occurring. However there was some permanent deformation of the metal, in that the strip became thinner by several microns. The “action” was detected by measurements at many hpa6points of the microhardness of the metal; an easily detectable enhancement reaching as much as 10 per cent along 30 mm was observed on each of the two sides. Examination by scanning electron microscope of the foils from the hardness-enhanced section revealed a high density of loop and other dislocations.
This claim is of such significance that it has been a matter of great importance that it be confirmed or disproven. We therefore invited metal- bender Jean-Pierre Girard to London and under witnessing exposed a similarly prepared specimen of AU4G-T351 to his action. A small decrease (3 microns) in the thickness of the specimen was monitored during exposure. Afterwards the microhardness of both sides of the specimen was independently measured by Dr. Desvaux at the Electrical Research Association and by Dr. Bouvaist at the Pechiney laboratories. The comparison between their data is seen to be satisfactory in Figure 6, and a local hardness enhancement is clearly seen. Similar experiments with metal- bending children are in progress.
6. Paranormal hardness variation at centre of either side of aluminium alloy strip. Open data points taken by Dr Desvaux (Leatherhead), closed data points by Dr. Bouvaist (Voreppe). Solid lines represent mean of pre- exposure data (typical points shown as crosses).
When a metal is bent, physically or paranormally, the work-hardening at the bend is associated with new dislocations. These are not usually loop dislocations, which can be produced in exposure to nuclear radiation. Metals in which dislocations are induced do not always bend, but internal strains are set up, and if these are of appropriate magnitude and configuration, plastic deformation can occur.
We could make a reasonable supposition that a “primary” event in the paranormal metal-bending process is the formation of loop and other dislocations. These need not necessarily result in actual plastic deformation. In this way both hardness enhancement and bending could be classed as the same type of event, differing only in geometrical configuration of the action. We have seen that the elastic dynamic strain signals differ continually in their geometrical configuration (1,2,3). It is likely that these represent the dynamic strains caused by the production of the dislocations; when there is a sufficient gradient of residual strain, the yield point is reached, and permanent deformation occurs.
Thus the paranormal metal-bending action, albeit limited macroscopically to a “surface of action” (2), which is still a useful model in current experiments, is essentially an internal action. We might claim that it was “information” which brought about dislocations in the lattice; this “information” may well be connected with the role of the observer in quantum mechanics(7) and might be termed “mentally induced tunnel effects”. Only on rare occasions would there appear to be a macroscopic quasi-force field acting on the metal.
The detailed structure of the surface of action is a kind of “curtain of raindrops”, enveloping certain lattice points and causing the onset of loop dislocations. However, these produce hardening and not softening of the metal, and we recall that there is good evidence, both in early observations(7) and in unpublished observations of Crussard and Bouvaist, of temporary and permanent softening, sometimes so extreme that it has been referred to as “plasticization” of the metal before fracture. If we hypothesise that the surface of action is sometimes capable of specific action at grain boundaries (taking on a kind of irregular honeycomb configuration), then a softening mechanism similar to the well-known quasi- viscous creep becomes possible. In this high temperature process extremely thin layers of atoms at the grain boundaries actually liquefy; but in the paranormal plasticization the temperature rise is unobservable, thus implying extreme specificity of the action, which would have to ensure that only a minute proportion of the atoms were in the high temperature phase. Electron micrographic evidence for the liquefaction has been obtained by Crussard and Bouvaist in the paranormal permanent softening process.
The direction of work in progress is to test the hypotheses that the dynamic strain signals correspond in number and strength to the density of dislocations produced in the lattice, and to the degree of liquefaction at the grain boundaries.


J. B. Hasted, “An Experimental Study of the Validity of Metal Bending Phenomena”. Journal of the Society for Psychical Research. Vol 48, No 770, 1976. pp 365-383.
2 J. B. Hasted, “Physical Aspects of Paranormal Metal Bending”. Journal of the Society for Psychical Research. Vol 49, No 773, 1977. pp 583-607.
3 J. B. Hasted, Zeit fur Parapsychologie und Grenzgebeite der Psychologie, Vol 20, 1979 pp 173-184. Also, in course of publication in New Horizons.
4 P. Barnes, J. W. Jeffery, O. Bateman, T. Gate, T. Southern, Birkbeck College, University of London, private communication, 1974.
5 C. M. Crussard and J. Bouvaist, Memoires Scientifiques Revue Metallurgique, 1978, February, p 117.
6 J. B. Hasted in “The Geller Papers”. Ed. Charles Panati, Houghton Mifflin Co. Boston. 1976, pp 183-196, 197-212.
7 E. P. Wigner, “Remarks on the Mind-Body Question” in The Scientist Speculates, Ed. I. J. Good, Heinemann, London, 1961. and “Symmetries and Reflections”, Indiana University Press, Bloomington, 1967.
We are grateful to the New Horizons Research Foundation, and to the Society for Psychical Research for financial support of these studies.


Follow Uri

Scan to Follow Uri on Twitter

Latest Articles

Read All Latest Articles
Amazing Lectures! uri lectures
Motivational Inspirational Speaker
Motivational, inspirational, empowering compelling 'infotainment' which leaves the audience amazed, mesmerized, motivated, enthusiastic, revitalised and with a much improved positive mental attitude, state of mind & self-belief.

“There is no spoon!”

The Matrix

“The world needs your amazing talents. I need them”

Michael Jackson

“Uri Geller gave an absolutely resonating talk on his life and career. He had every single magician in the room on the edge of their seats trying to digest as much information as they could. Uri emphasized that the path to frame is through uniqueness and charisma and that professional entertainers must be creative in their pursuits of success and never shy away from publicity.”

Tannens Magic Blog

“The man is a natural magician. He does everything with great care, meticulous misdirection and flawless instinct. The nails are real, the keys are really borrowed, the envelopes are actually sealed, there are no stooges, there are no secret radio devices and there are no props from the magic catalogues.”

James Randi (In an open letter to Abracadabra Magazine)

“Absolutely amazing”

Mick Jagger

“Truly incredible”

Sir Elton John

“Eternity is down the hall And you sit there bending spoons In your mind, in your mind”

Johnny Cash

“I Have watched Uri Geller… I have seen that so I am a believer. It was my house key and the only way I would be able to use it is get a hammer and beat it out back flat again.”

Clint Eastwood

“Better than watching Geller bending silver spoons, better than witnessing new born nebulae’s in bloom”


Urigeller_facebookDo you have a question? Contact Uri!