Copyright 1961, Journal of the Society for Psychical Research. Reproduced with permission.


J. Beloff and L. Evans (1961)

Queen's University, Belfast

The present investigation was undertaken in the belief that if such a thing as a psycho-kinetic effect is a reality it might reveal itself more clearly in relation to sub-atomic particles than in relation to macroscopic bodies. Two considerations may be mentiooned as conducive to this view. First, such evidence that we have for psycho-kinesis is statistical in nature. Now sub-atomic particles, being inherently random in behaviour, may be regarded as "nature's own dice" and may seem to offer a purer and more promising testing-ground for any psycho-kinetic effects. Secondly, whereas the notion of a psychic force acting at the macroscopic level seems to run counter to the principles of Newtonian dynamics, the atomic level, where principles of quantum indeterminacy become relevant, allows perhaps a certain loophole. For it seems less improper to suppose that the mind might exert a direct influence on matter by introducing a slight bias in a pattern of probabilities than by interfering with gross energy exchanges. If so, even the psycho-kinetic influence on dice may eventually be explicable in quantal terms.

The original plan was to test our hypothesis by using a "spintharoscope", that is, an apparatus consisting essentially of (a) a radioactive source and (b) a phosphorescent screen on which scintillations would be produced by the particles emitted from that source. It would then be a simple matter to divide the screen into a few different target areas and to instruct the subject to "try" increasing the number of scintillations occurring within a given area in the course of a given trial. Such a test seemed to us to offer the closest atomic analogue of the conventional dice-throwing test of psycho-kinesis. Accordingly we designed and had built a spintharascope using an isotope of plutonium as a source of alpha particles, an evacuated chamber and a screen coated with zinc sulphide.

Unfortunately when we came to try out the apparatus we found that the scintillations were too feeble to be detected with any assurance by the naked eye, even under dark adaptation, and there seemed to be no obvious optical device which could be used to make them clearly visible both to experimenter and subject. The set-up might have been made workable by using photo-multipliers covering each target area, to amplify the scintillations, but such a step was far too costly to consider.

We then turned to the possibilities of existing apparatus like the familiar geiger-counter which measure the rate of radioactivity. Would it be possible, we wondered, for the subject to exert a psychic influence on the rate of radiation during a given interval? The rate of radioactive disintegration, it should be noted, is of course a statistical concept, that is the average number of radiations occurring during a given interval of time remains constant. It is not like the frequency of a vibrating body, or the period of a pendulum, an absolute concept. Nevertheless every given radioactive element or isotope has its own specific rate, a quantity that is conventionally expressed in terms of its half-life. Moreover there is no known physical method which can in any way modify the rate. Or at least the only method which would have any effect would be that of bombarding the nuclei of the radioactive atoms with neutrons but this would produce its effect only by actually transmuting the atoms into those of another element. Rate of radiation can thus be regarded as a physical immutable, unlike direction of radiation, as in a spintharoscope, where the alpha particles, being positively charged, could easily be diverted by a magnetic field. We stress this point because in what follows we are asking the mind to do somehting which is in the most literal sense physically impossible. Here, then for those who are interested are the details of the experiment.


A source of alpha particles, consisting of a few centigrams of uranyl nitrate in crystalline form, was placed inside the chamber of an Ecko Universal Scintillation Counter (Type N664A), whose sensitive element consisted of a crystal of sodium iodide activated with thallium. This apparatus fed into a second apparatus, namely an Ecko Scaler (Type N526), whose function was to register the number of alpha particles emitted during a given time interval. The dial on which the numbers appeared was at all times visible both to subject and experimenter.


Thirty men and women volunteers, mainly students at Queen's University, Belfast.


Each subject was put through one complete session. A session comprised three runs, and each run consisted of ten trials of one minute each. These trials were done in pairs such that one trial in every pair was called the positive trial, which means that the subject was instructed to try for as high a score as possible, while the other trial was a negative one, where the subject was required to try for as low a score as possible. There were thus thirty trials to a complete session, 15 positive and 15 negative. With 30 subjects this meant a grand total of 900 trials in the experiment, 450 positive and 450 negative.

Since our Scaler was not equipped with an automatic timer, the experimenter (always one or other of the authors) had to start and stop the Scaler on each trial using a stop-watch. It was therefore important that he should not know whether his subject was aiming at a high or a low figure, since he might be tempted, if only unconsciously, to cheat by inaccurate time-keeping. Another reason for keeping him in ignorance was that, if a significant effect should occur, one would want to know whether it was in fact due to the subjects' volitions rather than to those of the experimenter. Accordingly before every pair of trials the subject drew a card from a shuffled pack and depending on whether it was red or black, either the first trial was positive and the second negative, or vice versa. Only after the termination of every pair of trials did the subject hand over the card to the experimenter who then entered each score into its appropriate column.

At the beginning of each session the subject was taken into the testing room and the physical set-up was briefly explained to him. He was then carefully instructed in the procedure that would be followed and told that he must try to exert a "psychic influence" on the number of radiations as recorded on the dial. He was further offered an incentive of half-a-crown if he could produce an excess of 20 or more in the positive as against the negative trial for any pair. We were working with a natural rate of radiation of rather less than one per second so that a score of around forty for a one minute trial was about average. At the end of each run the subject was told how well he was doing in terms of the difference between his scores on the positive and negative trials.


Consider first the overall score on all the 450 positive trials and on all the 450 negative trials, the result was as follows

Positive: 17,393
Negative: 17,458

It is thus immediately clear from inspection that the figures show no significant departure from equality as expected on a null assumption.

It may happen in psychical research, that an overall null result may be due to the fact that significantly negative scorers and significantly positive scorers cancel each other out. A genuine effect is then masked as a consequence of pooling one's negative goats with one's positive sheep. To guard against this contingency we did an analysis of the individual results and calculated a Chi^2 index for the magnitude of the difference between the overall positive and overall negative trial scores regardless of the direction of the difference. These Chi^2 indices, each possessing one degree of freedom, could then be summated to give a total Chi^2 index for 30 degrees of freedom as follows:

Chi^2 = 27.96.

The probability of reaching a Chi^2 value of this magnitude for 30 degrees of freedom (as given in Fisher's tables) is 1 in 2. The result is therefore obviously non-significant.

It still remained possible that certain individual results might have been significant, but in point of fact none was so at more than the 5% level of confidence, and with 30 subjects to choose from this was no more than was to be expected. Moreover, tow of the higher scorers were subsequently followed up in a second session (not included in the data given) that failed to sustain their promise. The half-a-crown bonus was won on a total of six occasions but these passing glories were offset by deviations of minus twenty or under on no less than nine occasions.

Finally, it seemed worth testing whether there was any significant decline effect in the course of a session, such as has been widely reported in the literature on psycho-kinesis. It transpired that there was indeed a falling off in the total difference between positive and negative trials from Run I to Run II, indeed in Run II the difference became slightly negative. However, a Chi^2 test on the two-by-two matrix of scores showed that this change was not significant at the 5% level of confidence. (The Chi^2 value obtained was only 2.48 for one degree of freedom.)

Discussion of Results

When one is working as much in the dark as one is in this field it is no use trying to apportion the blame for a null result. It is possible that there may be no genuine phenomenon of psycho-kinesis, notoriously it is the least well-substantiated of the three classic `ESP-type' phenomena. Alternatively it may exist, but only in conjunction with certain very rare persons who were not included in our sample. (University students have seldom been the most promising psychic material.) Finally, it may exist and be very common, but we simply used the wrong method to tap it. One aspect of the present set-up which we though unsatisfactory was that it provided the subject with nothing relevant on which he could fasten his attention. Thus whereas one can readily think about a familiar concept like direction, one cannot properly visualise the actual occurence of a single emission.

In this last connection we would like to point out certain technical draw-backs in the type of situation we were using for the benefit of anyone who has access to a radioactive chemical laboratory and is dauntless enough to follow our example. The count of alpha particles is inevitably contaminated by external gamma radiation to which the apparatus is also sensitive. this is reduced as far as possible by the fact that the chamber of the scintillation counter is encased in thick lead, and it can be futher reduced by turning up the discrimination bias of the counter and scaler to a maximum. Even so, by running the chamber empty under these conditions, we estimated that background radiation was responsible for something like a quarter of all radiations counted. There are also noticeable fluctuations in the gross numbers registered from fay to day due to fluctuations in the strength of gamma radiation in the atmosphere, as well as a decrease in gross numbers when the apparatus has been running for some length of time due to a lowering of its sensitivity. However, since the alpha particles were susceptible to a psycho-kinetic effect this should not be swamped by the intrusion of extraneous gamma radiatin, provided of cours, a sufficient number of trials are run.

To conclude: The radioactivity test, as here described, failed to yield any evidence of a psycho-kinetic effect. It still remains to be seen, however, whether a different use of quantal phenomena (such, for example, as was suggested in our spintharascope scheme) might not yet bring us closer to the core of the elusive phenomenon of psycho-kinesis.

RetroPsychoKinesis Project Home