Graham Watkins

A summary of my major research in Parapsychology

I was at this time (early 1970s) a collegue of J.B. Rhine, a staff member at the (now the Rhine Research Center) Foundation for Research on the Nature of Man--the descendent of the Duke Parapsychology lab--in Durham, NC. We were struggling with the problem of trying to work with "special" subjects, subjects who came to us claiming unusual abilities; some in the field object to the use of such subjects, but as I said then and as I still contend, if you're studying human ability to slam-dunk a basketball, you study Michael Jordan, not the Man in the Street.

On the other hand, it has also been my contention that you cannot properly study these presumed abilities by allowing such "special" subjects to do their "tricks;" I have myself been singularly unimpressed by some of the more famous "special" subjects, like Uri Geller. Finding and identifying the "tricks" such subjects perform may be impossible, and, in my experience, the ability to do sleight-of-hand and what appears to be real undocumented ability often goes hand- in-hand.

For this reason--and because our earliest "special" complained that tests like the Zener cards and the (Helmut) Schmidt electronic random number generator were boring to her--my associates and I devised what came to be known as the "mouse-ether" experiments. In these studies, the task the subject is presented with is that of waking a mouse unconscious due to the effects of ether more rapidly than it might otherwise awake--this was presented to the subject with emphasis on the fact that the mouse was unconscious due to the toxicity of the ether, and, though the subject knew the animal would not die in any case, the task was framed in the context of a "psychic healing" process.

Quite early, we encountered a problem with these studies; that is, the mice we were using varied rather wildly in their individual responses to ether. Using a constant dose, some individual animals would remain unconscious for less than a minute while others might sleep for half an hour. Although our early experiments were suggestive, this extreme variability in the animals themselves created such large variances that statistical tests were unimpressive. There are of course "outlier" procedures for handling such variance, which would have been acceptable in my original discipline--animal behavior--but were an obvious source of criticism and dismissal in Parapsychology.

For that reason, we undertook a long-term--and expensive--selective breeding program designed to develop mice that were consistent in their response to ether. Fortunately mice breed rapidly, and efforts were rewarded in a matter of several months; from a base Swiss-Webster stock we'd established a breed that, using our then-standard dosage of USP ether, seldom varied more than fifteen or so seconds in the time they remained unconscious.

In the actual experiments, the mice were placed side-by-side, separated by a vertical wooden partition, in front of a one-way glass through which the human subject could see and "work on" the target animal. Before a series began, the subject was given a card with the target animal indicated on it; the operators in no case knew which animal was the target. At first, the data was taken by the operators, using stopwatches, and basing their time on the animal's first coordinated step; later, photocell-and-lamp automated timing was introduced. The data thus acquired was analyzed using standard analysis of variance; independent variables were experimental/control, left/right, first half/second half (the human subject was given a break between "halves" of a run of sixteen pairs), and male/female (mouse--females recover more quickly on the average). The analyses were done by technicians at Duke University who were unaware of the nature of the experiment, and the P-values derived from these analyses often exceeded 10 (to the minus) 9; that is, the subjects (eventually four in number, two men, two women) could apparently cause their "target" mice to awake more quickly, often dramatically so. "Unselected" subjects showed no such ability. In later years (see the references) two different operators, using our specially-bred mice, were able to replicate our experiments without difficulty. Subsequently, as far as I know, no laboratory has chosen to make the investment of time and money to create a breed of mice consistent in their response to ether anesthesia. This does not, however, invalidate our initial studies.

Criticism was leveled at the experiments on the grounds--absurdly, it seemed to me--that ether was an "outdated" anesthetic (we could've used cyanide, after all. Ether was merely convenient). To answer this, though, the experiments were replicated using sodium pentobarbital (injected into the tail vein) and halothane(varathane). Pentobarbital had the disadvantage of causing the mice to remain unconscious for periods up to an hour or more, causing the subjects to complain about fatigue. But the mice had been bred for consistent response to ether, not these other chemicals. Nevertheless these studies were completed, and the results were the same--if not quite so dramatic.

A more serious criticism lay in the distribution of "targets" during a series. At the outset-- when the experiments were less formal--we had started by assigning a "target" side (left or right, subject's perspective) for the first eight in a series of sixteen, with an automatic shift to the other side after the break. Later--after the strain of ether-consistent mice was well-established--the effect was, in fact, rather dramatic, and thus it wasn't unusual to see the subject bring fourteen or so of their mice up before the controls got up. Therefore, the "blind" was quickly broken. We'd learned, though, that a random sequence of targets through the experiment--LRRLRLRL, for example, for a "half"--caused the subjects to complain of fatigue and caused the effect to be minimal to non-existant.

Wondering why this was so, we ran a series of studies in which the subject was not present during the second half of the experimental run. To our surprise, the effect continued--on the same side as the subject had earlier been focused on. Further studies showed a logarithmic decay in this effect requiring more than a half-hour before extinguishing; over several series of such studies, correlation coefficients between first and second halves (second half=no subject present) were highly significant statistically.

In subsequent studies, we determined that the effect lost potency over distance, but in a linear fashion, not as an inverse square; we also found, to our surprise (this finding was quite accidental at first) that the presence of large masses of grounded metal in the vicinity of the targets seemed to suppress the effect, suggesting some sort of possibly electromagnetic mediation, although we could never demonstrate this in any satisfactory fashion. In addition, we took a series of measures of the physiological state of the subjects when they were "working." This showed a remarkable consistency over the four; there was little or no alpha in the EEG, the heart rate and respiratory rate were significantly elevated, and the fingertip plethysmograph (a measure of periferal vasoconstriction) was dramatically reduced in amplitude. In addition, the EKG showed some odd but consistent changes; the T-wave (associated with repolarization of the heart) was altered in all four subjects, but in opposite directions between the men and women. Since we were working with so few subjects, this finding can only be considered suggestive.

Having these parameters in hand, we then undertook to attempt to train unselected subjects, using biofeedback techniques, to simulate the physiological state of the "special" subjects. Doing this, we were able to achieve successful runs (statistical significance at or below the 0.01 level) with these previously unsuccessful ordinary subjects--through never to the dramatic extent of the "special" subjects.

Some time later--when there was a great deal of interest in the Russian woman Nina Kulagina, who was said to be able to move small objects mentally--a woman from New York named Felicia Parise was observed by Chuck Honorton and others to do the same. Honorton brought her to our lab, where we were able to work with her briefly; for us, she was able to move a compass needle several degrees off North. There are of course many ways to fake this, but: when Ms. Parise left the experimental setting, the needle remained, standing off North. Removing it from the environment caused it to gradually swing back as it was taken farther away; returning it caused it to move off North again. Finally--in Ms. Parise's total absence--the effect was allowed to "decay." As expected, it took about thirty minutes--exactly the decay factor we'd seen previously in our mouse-ether experiments.


Watkins, Graham K. and Anita M.: Possible PK Influence on the Resuscitation of Anesthetized Mice. Jour. Parapsychol. Vol XXXV No. 4: Dec. 1971. pp 257-272

Wells, Roger, and Judith Klein: A Replication of a "Psychic Healing" Paradigm. Jour. Parapsychol. Vol XXXVI No. 2: June 1972. pp 144-149

Watkins, Graham K, Anita M. Watkins and Roger Wells: Further Studies on the Resuscitation of Anesthetized Mice. Research in Parapsychol., 1972. pp 157-159

Watkins, Graham K. and Anita M.: Apparent Psychokinesis on "Static" Objects by a Gifted Subject. Research in Parapsychol., 1973. pp. 132-134

Wells, Roger, and Graham K. Watkins: Linger Effects in Several PK Experiments. Research in Parapsychol., 1974. pp. 143-147

The Rhine Research Center

Return to the Graham Watkins home page.