Extra-Sensory Perception - (03) General Survey
PART II. THE EXPERIMENTAL RESULTS
CHAPTER 3
A General Survey
The investigation of extra-sensory perception at Duke University has now been going on for more than three years, and has come to include well over 90,000 trials. To give a comprehensive report of these trials, with a proper account of procedure, conditions and results, would make a large volume. Much summarizing must, therefore, be done in order to present the results in a reasonably readable form. It seems best to present first merely a narrative sketch of the main lines of the research and to state the general results; and, following this chapter, to give more detailed accounts of the principal subjects who produced the results and of the special experiments conducted. Those who wish to skip these fuller chapters may do so by going on from the end of this one to Chapter 9.
Following upon our experimental interest in the telepathic horse, Lady, 1 during 1928 and 1929, considerable effort was made to find other infra-human telepathic subjects but this was in vain. In the summer of 1930, then, I turned to the task of trying to find human subjects. I began by giving "guessing contests" to some groups of children in summer recreation camps. The tests consisted simply in having each child guess the numeral (0 to 9) which was stamped on a card that I held concealed from him in my hand and looked at. Each child had a pencil and card, and noted down his guesses silently. From the thousand (approximately) trials thus made, no one individual stood out well enough to seem to warrant further investigation.
During the fall semester following my colleague, Dr. K. E. Zener, proposed that we try sealed envelope guessing tests on our own college classes. We accordingly prepared envelopes with numerals (or, in some classes, with letters of the alphabet) effectively concealed and sealed within. These were passed out to the students with instructions to guess. the number (or letter) stamped inside, under certain conditions of quiet and relaxation. Of these trials 1,600 were carried out, also with quite insignificant results. The results of the five series as a whole were very close to the chance expectation, three of the groups coming out above chance and two below that figure. This, too, was then given up, partly because it was quite laborious, and partly because of indications of failure. Further detail will be furnished in Chapter 4.
The objective had been partly to measure the ability of the group and partly to discover individuals with special ability to perceive without
the senses. The latter goal was achieved, since we did discover one able subject through these tests, Mr. A. J. Linzmayer. In the two group tests in which he took part he was the highest scorer. In the better of these, on envelopes containing figures chosen from 0 to 9, giving a probability of being right of 1/10 per trial, he got three correct in five trials. In the other, with a probability of 1/5 for correctness on each one, he scored four correct in five trials. From these results it was thought worth while to try further tests with Mr. Linzmayer. These I will describe later.
At about the same time, the fall of 1930, another colleague in our Department of Psychology, Dr. Helge Lundholm, kindly offered to cooperate in an attempt to measure "telepathic" perception (clairvoyance was not excluded) in the state of hypnotic trance. He assumed responsibility for providing the trance and we worked with, in all, 30 subjects, who made a total of 1,115 trials. These fall into three groups; they have a different probability basis in each and cannot therefore be thrown together. All are somewhat above the chance expectation but only slightly. The best groups, in which numbers from 0 to 9 were used as symbols, totalled 530 trials and yielded 65 right as against 53 expected on chance. This positive deviation of 12 is only 2.6 times the probable error (±4.66) 1. It might be said that, had we continued these tests for as many more trials with equal results, the data would have approximated the point of significance. But the procedure was slow and we discovered that such slight deviations as we got could be had as well in the waking condition. So we discontinued the series. The details of the various procedures, and the data will be given some space in the next chapter.
A few tests in simple "card-guessing" made now and then upon individuals by Dr. Zener and myself during the year 1930-31 seemed to give promising results. They were never high but seemed to favor the positive side to an interesting degree. These were mostly carried out on the basis of symbols suggested chiefly by Dr. Zener, five in number; namely: circle, rectangle, plus, star and wavy lines ( 
). We early began to use them in packs of five each, 25 in all. The subject usually called the top card, as the pack lay face down on the table before him. A series of 25 trials without any extra-sensory perception would yield, on the average, about 5 correct hits. But these odd tests we were making yielded around 6, on an average. And, keeping track of those of my own observation alone, I found after a while that they were becoming fairly meaningful statistically. From a total of 800 trials carried out during the academic year, 207 hits were recorded, which is a positive deviation of 47 and is more than 6 times the probable error. But this yield (around an average of 6.5 correct in 25 trials) was low in comparison
with what was in store for us just ahead in the work of Linzmayer. Two more cases, however, came in chronologically before the real discovery of Linzmayer.
First is Mr. Harvey L. Frick's interesting card-guessing experiment. Mr. Frick, a graduate scholar, had demonstrated his ability in extrasensory perception in work reported in his Master's thesis in 1931. Following this thesis work, he undertook to study the "fatigue" effect on the results obtained by guessing a long run (100 playing cards) daily, supposing this would fatigue him. When, after 9 days, he reported his results, they showed a striking decline series. Totalling the results by order of 20's of trials per day we get for the total of the first 20's for the 9 days, 58 correct suits; for the second 20, 50; for the third, 48; the fourth, 38; the fifth, 36. If we subtract the chance expectation (i.e., 1/4 the number of trials) of 45 for each total of 180, (20 per day for 9 days) we get the following decline series in the deviations: +13, +5, +3, -7, -9. The results as a whole are not significant. They seem to cancel out at one end of the curve what they gain at the other. But the curve is significant; i.e., the difference in deviation of the high end from that of the low end is 4 times the probable error of the difference. The rest of the experiment and its conditions I leave for the reader to follow at greater length in the next chapter.
During the Spring of 1931 another of my students, Mr. Charles E. Stuart, now an assistant in our Department, carried out some observations on extra-sensory perception, mainly card-guessing, sometimes with subject in trance and sometimes with one in the waking condition. In some of the experiments there was, as with those of Dr. Lundholm and myself, a combination of the telepathic and clairvoyant conditions. He, too, used cards with 5 geometric designs as the basis for the guessing for about half of these tests. Out of a total of 1,045 trials made on 15 students, 495 were made on the geometric figures, with a probability of 1/5 per trial, giving a chance expectation of 99 in all. His subjects actually got 147 correct, a positive deviation of 48, which is 8 times the probable error and means an average of 7.4 per 25 guesses. There were two other types of tests given by Stuart, also yielding above chance. But, again, with this brief item for the completion of the chain of development, we will leave the details of Stuart's work for the next chapter and for a later chapter entirely devoted to his more important later results with himself as subject.
Now we return to Linzmayer. Late in May, 1931, he was given 45 card-guessing trials in very light hypnosis (which was as deep as he could go) and called 21 correct as against chance expectation of 9. In the few days we could work with him after this and before his leaving the campus with the close of the year, he brought his total trials up to 600,
run for the regular test, with another 900 for a special test. The 600 regular trials yielded 238 hits, a positive gain over chance of 118; this is about 18 times the probable error. This figure can leave no intelligent question of the operation of a significant principle. In the average number of successes per 25 trials this rises to the unprecedented figure of approximately 10. In one series of 25 Linzmayer got 21 hits, 15 of them being successive. In this series he did not see the cards as they were dealt and called. But (and alas!), after three most exciting days of experimentation with Linzmayer, he had to leave. During the final hours I pressed him into a hasty experiment involving 900 trials, for a special purpose, the results and purpose of which will be presented in Chapter 5.
Under the stimulation of Linzmayer's brilliant example, I set to work giving preliminary tests to many other students and acquaintances. My young sister-in-law, Miss Miriam Weckesser, then aged 15, could do fairly well if alone. In all, she ran 1,050 trials over a period of a year, yielding 266 correct calls, with a deviation 6.6 times the p.e. Then it appeared that she had lost the ability, at least temporarily. Her inability to work well when witnessed and her loss of the ability after 1,050 trials were points of value as suggestions, and they were evidenced later among well-authenticated results.
During the summer one of my students, Mr. A. E. Lecrone, a high-school teacher, became interested in this work and ran a series of tests with a friend (Miss A. A. P.) as subject, and in another series served as subject himself, with the friend observing. The regular, 5-symbol cards were used, and the conditions allowed both telepathic and clairvoyant perception. Together they totalled 1,710 trials, he doing about half as well in deviation above chance as his friend. There were 392 correct, a deviation of 50, which is 4.5 times the probable error.
During the summer and fall of 1931 my own odd tests, made on 14 "stray" subjects, totalled 835 and yielded 208 hits, which is 41 above chance. This is at the rate of 6.3 per 25. This little group is itself significant, being over 5 times the p.e.
In the month of October, 1931, we were able to get Mr. Linzmayer for a short period again and made 945 tests on clairvoyant card-guessing as before. But this time he ran at a much lower rate. His yield was 246, which is 57 above chance expectation; this is about 7 times the p.e. The rate per 25 was 6.5. This is, however, still quite significant even though it was low in contrast with his first 600 trials.
The greatest event of the academic year 1931-32 was the work of Mr. C. E. Stuart, who was mentioned above as conducting a number of tests on his student friends. Meanwhile he had added to the work already mentioned about 900 more, yielding 257 or 77 above chance, 9.7 times the
p.e., with an average of 7.1 per 25. But Stuart had begun to test himself at card-guessing. He worked alone but kept good conditions, not looking even at the backs of the cards. He ran, through the year, the very large total of 7,500 trials, which represents a great amount of patience and labor. What wonder if he perhaps got tired toward the end? For some reason, at any rate, he dropped in his rate of scoring until he ran only slightly above chance expectation toward the last and I advised him to stop for a while. His results over the 7,500 show an interesting, though somewhat irregular, gradation of decline. They yielded 1,815 successes, averaging 6.05 hits per 25. There is a gain over chance of 315, which is 13.5 times the probable error—enormously significant. While Mr. Stuart is himself a responsible investigator, it will do no harm to add, in view of the fact that most of the 7,500 trials were unwitnessed, that the 140 of them that were done in my presence yielded at the rate of 6.15 per 25 calls, slightly higher than the average for the whole series of 7,500.
In March of 1932 we again had a short visit from Mr. Linzmayer and obtained 960 regular trials with him, as well as some more special tests. The 960 trials yielded 259 (a still lower rate than the last time) which has a positive deviation of 67, and this is 8.0 times the p.e.; the average per 25 is 6.75. (In the preceding fall, it had been 6.9) One of the special experiments was made at this time by giving the subject 15 grams of the narcotic drug, sodium amytal. By an hour after the ingestion Linzmayer was quite sleepy and dull-witted. He was "thick-tongued", jolly and talkative. I kept him awake for 275 trials but he could not score appreciably above chance. The average per 25 was 5.1, having gotten a total of 56 (as against 55). Before and after this experiment Linzmayer ran at an average of 6.75 in 25, as reported above.
Mr. J. G. Pratt, an assistant in our Department was engaged, during the year to help in the necessary prospecting for more good subjects. He carried out on 15 students 10,035 tests with a yield of only 144 above the chance expectation; this is, however, still more than 5 times the probable error. His tests on himself numbered 2,885 and yielded a deviation above that is 3.9 times the p.e. But his main contribution lay in the discovery of Mr. Hubert Pearce, Jr., a young ministerial student whom I had asked to submit to the tests, on learning that his mother was reported to have possessed parapsychological ability. Mr. Pearce ran low for a few series of 25 each, but soon picked up and then held fairly steadily at about double the chance figure (of 5 per 25). Pearce, too, was discovered at about the close of the school year, but he was able to stay over for a time and 2,250 witnessed trials were made in clairvoyant card-guessing. The yield was 869 or 419 above chance. This means an average of 9.7 per 25. The huge deviation from chance expectation is 32.75 times the
p.e., a figure of reassurance against the chance hypothesis that simply leaves no question of significance. The experiments were then interrupted by Pearce's appointment to ministerial service for the summer.
One of the characteristics of Pearce's work was the relative smoothness of the results from day to day. He would average around 10 per 25. He did not at this time seem to be helped by having the observer look at the cards. Curiously enough, he would drop in his scoring under this condition. However, almost any change whatever, unless he himself proposed it, seemed to throw him off his rate of scoring. Visitors disturbed him for a while but he would always get back to his level if they remained for a time. Also he would become adjusted to the changes in procedure in the course of time, so far as we tried to make him. But we did not want to induce too much strain and often yielded on a desired innovation. Certain changes were introduced by talking about their possibility indifferently and allowing Pearce to say if he wanted to attempt them. In this way we started the calling for low score; i.e., trying to make wrong calls. In 225 low-score trials made under this condition he scored only 17, which is 28 below the chance expectation, and this is 6.9 times the p.e. Highly significantly low! This averaged below 2 hits per 25 calls. He produced, for example, when asked to score "high", a 10 in 25, then for "low", a 1 in 25, then a 9 in 25 for "high", and another 1 in 25 for a "low". It seemed purely a matter of choice!
Another new feature introduced into Pearce's work in a half-playful way, and which was also successful immediately, was calling cards down through the pack without removing the cards until the finish of the run. This started off with scores of 8, 8, 12, 6, etc., per 25 for the beginning runs. The first 275 trials yielded 87 hits, a gain of 32 or over 7 times the p.e., and an average of 7.9 per 25. The value of these data is enormous, as we will emphasize later, first in connection with alternate theories of hyperacute sense perception and second in connection with theories of the physical basis of the process of extra-sensory perception; i.e., the subject reads the cards under conditions to which no radiation theory seems applicable and no sensory perception seems adequate.
We repeated on Pearce the sodium amytal experiment made on Linzmayer, using only 6 gr. this time. This would be equal to about half what Linzmayer had taken, allowing for weight differences. Pearce was not incoherent and irrational, but was quite sleepy. He could, however, keep himself awake and could converse intelligently. He made effort several times to re-integrate himself, once even washing his face with cold water. At the beginning his scoring fell off at once, yielding 5, 4, 3 for the first three runs of 25 each. Then he "pulled himself together" and got 10 on the next. The average for 325 trials is 6.1 in 25, as against
[paragraph continues] 9.7 for his regular scoring. This is a very significant drop. This ended the series for the summer, leaving the young minister barely time to "sober up" for his first sermon.
During the summer months Mr. Stuart was encouraged to try again and, to the surprise of every one, he "came back" about as well as ever. During his first 400 he averaged 7.3 per 25, whereas his old average had been 9 in 25 for the first 500 but for the first 1,500 had been only 7.15. In 2,100 trials made by Stuart during the summer, he obtained 575 hits, 155 above chance expectation, which is 12.5 times the p.e. and represents an average per 25 of 6.8. But the same decline set in again with Stuart and after 2,100 trials his results showed that he might as well cease. Stuart carried out, however, one very interesting variation to the regular procedure. He did this independently in Rochester, N. Y., soon after we had done a similar experiment (of which he was ignorant) with Pearce here at Durham. He began giving a "right" and "wrong" call for each card, keeping the records labelled and separate. His results show for the "wrong" calls about the same negative deviation from chance expectation that the "right" calls show for the positive side. That is, when he tries, he can go high or low, in about the same degree. His decline curve based on the "high" column became an incline curve for the "low". (For the full data see Chapter 6.)
In the early fall Mr. Pearce returned to Duke and we went hard at it again. We were particularly desirous of increasing the "D.T." totals. (These are obtained by calling down through the shuffled pack without touching the cards, leaving the pack unbroken until the end.) These soon reached the 1000 mark and the significance rose still farther beyond the range of question. At the "1000" milestone, the deviation was about 12 times the probable error and the average run per 25 was 7.5. One feature of interest here was the apparent difficulty shown in scoring high in the center of the pack, which kept the scoring lower than Pearce's other ("B.T.") work.
Another experiment of interest was the effect of the stimulant drug, caffein. Under the influence of the drug treatments the deviation or gain above chance expectation was doubled over what it had been for the preceding part of the period. The rise of deviation was actually from 44% of chance expectation to 98%. (See Chapter 12.)
The effect of the presence of strangers upon Pearce's scoring was more carefully measured and found to be highly reliable; i.e., at first he would drop to "chance". But in every case he rose again, while the visitor waited, to his original level. He stayed down longest with a magician present, but he rose again before the magician left—and "mystified the mystifier", who, himself, failed to score above the chance average in 75 trials given him.
A fatigue test was run with Mr. Pearce over an 8-hour period, during which he called 900 cards. There were no signs of decline in scoring rate and no special fatigue evidence. The average for the day per 25 was 10.1, which is a little above Pearce's average. About a month earlier Mr. Stuart had made 1,300 calls in one day, 700 for "right" and 600 for "wrong". They were among his highest in rate of success.
There were some interesting experiments with screens with Pearce which will be given in detail later in the report. At the moment I will lump them off, 600 in all, with a yield of 215 or a gain above chance of 95, which is 14 times the probable error. The average per 25 is 9. These are especially interesting on the score of eliminating sensory cues.
But I was at this time pressing Pearce on the point of telepathic perception or thought-transference. I mentioned above that he would drop in score when any one tried to help him by looking at the cards. I then began to work behind the screen so he would not know when I did or did not look. The data came out about the same for a while, whether I looked or not. But all at once he seemed to be utilizing telepathic perception in the unscreened tests made in the same general period, since, when without the screen, I looked at the card, he got very high scores, and fell lower when I ceased. He thereafter showed, even behind the screen, a marked advance in scoring with the added "telepathic" condition. Then I gave him a "pure telepathy" (P.T.) test, in which the agent merely chose at random an image of one of the five symbols on the cards. No cards or objective figures were used. This ruled out clairvoyance as commonly regarded. Pearce began, after some failures, to achieve real success with the "P.T." (pure telepathy) condition, with different agents. His first 950 trials yielded 269 or 79 above. This is 9.6 times the p.e. It is an average of 7.1 in 25, low for Pearce but not unusual for the beginning of a new experiment. This discovery in my best clairvoyant subject of a "pure telepathy" ability also impressed me as of great theoretical importance.
Meanwhile, the results were piling up at a rapid rate. With Pearce alone there were over 10,000, with Stuart another 10,000 and there were six others. Scepticism among colleagues was abating. They had only to come and see. And several students were becoming interested. During the fall of 1932 a number of girl students very kindly tested themselves for extra-sensory perception and, out of perhaps 10 or 12, there were 3 that stood out strikingly. These were Miss Sara Ownbey, a graduate student in our Department, Miss May Frances Turner and Miss June Bailey. All three have since done dependably high scoring over long series.
Miss Bailey was first encouraged to try the tests because she had had parapsychical experiences in her childhood. She has now made 3,900
trials, at an average of 7.8 per 25. The value of the deviation over the p.e. is as high as 26. She does well both with "pure telepathy" (P.T., no cards, only mental images) and "pure clairvoyance" (P.C.; i.e., cards only—no agent). Her P.T. score (average 9.4) is somewhat higher than her P.C. (average 8.0) at present.
Miss May Frances Turner, during the academic year 1932-33, made the large total of 5,125 trials, mostly unwitnessed, with the average per 25 of 8.4 hits. This gives her work a value of deviation from chance expectation divided by p.e. of 35, which is a most remarkable value statistically. Her ability, too, lies in both phases of extra-sensory perception, the telepathic and the clairvoyant. In fact, she has about the same standing for the two, an average of 9.0 for "P.C." and one of 9.4 for "P.T." She, like Miss Bailey, does not yet do so well at the D.T. calling (solid pack, untouched) and this has lowered both general averages somewhat.
The best subject on the D.T. work is Miss Sara Ownbey. Her average on 1,425 trials under D.T. conditions is 7.8 hits per 25 calls. On the total for the year of 1,975 trials, P.C. and P.T., she averages 8.9 per 25, and her positive gain (+307 ±12) is 26 times the probable error,—again, a tremendous value. Miss Ownbey also does well in both branches of this mode of perception. Her P.T. average is 8.1, her D.T. 7.9 and her other P.C. work averages 11.6 in 25. (P.C. includes both D.T. and B.T.; in B.T. the card is removed from the top of the pack after being called.) She has hardly gotten well started on P.T. work as yet, however, and most probably will improve her P.T. score, bringing it nearer the P.C., as others have done. Her own telepathic perception has really been neglected, due to the fact that she is an excellent agent and has served mostly with others in that capacity.
Again in the spring of ’33, Linzmayer kindly gave us an opportunity to work with him further. But this time he fell still lower in his rate of scoring, averaging now only 5.9 per 25 on the 3,000 trials he took part in. This systematic decline is surprising, all the more so since he is so keen to improve in his E.S.P. ability. Even so his results are significant, the 3,000 trials yielding a positive deviation 7.5 times the p.e.
But the most interesting feature was the fact that Linzmayer yielded significant results on the P.T. condition also, averaging 6. per 25 for 1,000 trials. To find in this declining "pure clairvoyant" subject a "pure telepathic" perception approximately equal to the already weakened P.C. was a very stimulating discovery, since the possible relationships between P.T. and P.C. were becoming more and more fascinating.
The keenest interest then lay in getting at possible tests differentiating between P.C. and P.T., with distance. For instance, pure telepathy might possibly work at a greater distance than P.C. or vice versa. A long
illness thwarted the plans for a time but in June, 1933, Stuart started working with Pearce on distance effect comparisons between P.T. and P.C. But somehow, to our surprise, Pearce did not do well even under ordinary circumstances and about all we got was the fact that, at short distances, Pearce scored the same on P.C. as on P.T., 6.3 in 25 on the one, 6.4 in 25 on the latter, both very low for him. This is in itself another link in the chain of comparisons that is seeming to unite these two phases of extra-sensory perception. One other curious feature showed up in this Stuart-Pearce experiment of June. Pearce did best on D.T., as did also Linzmayer, when he fell to his lowest. This would superficially seem to be the hardest, under ordinary conditions, and yet with a general decline it seems not to drop so easily. The data are as a whole, however, very significant, the total 5,400 trials giving a positive deviation more than 11 times the probable error.
Another excellent subject, one whom I had tested out over a year previously and who had gotten 27 correct in the first 50 trials, appeared again on the campus in July, 1933, Mr. T. Coleman Cooper. He made 4,850 trials in about 2 weeks, with Miss Ownbey as agent and observer. His average per 25 was 8.6 hits, for P.C., and 8.1 for P.T. in the same room with the agent. He, like the others, seemed to be able to perceive both telepathically and clairvoyantly in about the same measure. Moreover, in his daily fluctuations, which were quite marked, he went up and down in P.C. and P.T., both together. Like Pearce, Cooper did not achieve any contribution to the distance question, now uppermost in mind, valuable as the results are otherwise. One other point of considerable interest came out clearly in this set of data. Mr. Cooper worked with two agents in his P.T. work, Miss Ownbey and Miss Parsons. The results with Miss Ownbey averaged 10.9 in 25, those with Miss Parsons, 6.5 per 25. This is a point of value as to the role of the agent. Miss Ownbey has herself marked E.S.P. ability. Miss Parsons has shown none as yet.
Then in July, ’33, right in our midst arose a new and very successful subject, Mr. George Zirkle, another graduate assistant in our Department. Miss Ownbey, his fiancée, discovered his ability to perceive telepathically. Working with her as agent, he has already got most striking results, averaging in P.T. through 3,400 trials, 11.0 hits per 25. Another peculiarity is his inability to work clairvoyantly as yet but, of course, this may develop slowly in him at a later period, as telepathic capacity did in Pearce. Zirkle's scoring is sometimes phenomenal. He several times has gotten 22 in 25 correct. Once in calling 50 in a series, 26 were found correct in unbroken succession. This was about equalled only once before in our experience with E.S.P. and that was by Pearce, who, in my presence, ran 25 straight hits by pure clairvoyance.
Still more to the point is the fact that Zirkle does quite as well 12 and 25 feet away, with walls between him and the agent, as he does across the table from her. In fact, the averages are better with the distance and screening. The average per 25 in the same room when he is physically well is 14, at 8-12 feet away, 14.6 and at 28-30 feet, 16. Also, the conditions include the noise of an electric fan to exclude possible "unconscious whispering" and thus give quite good conditions for safety. A telegraph key is used for signalling.
Under these conditions, fan going, wall serving as a screen, telegraph key as a signal, a constantly changing system of choosing imaginal figures, we carried out another drug experiment with Zirkle as subject. He was given 5 gr. of caffein in capsule, not knowing whether it was caffein or sodium amytal, and when tested an hour later, rose from his pre-drug level of 12.5 per 25 to 14.7 per 25 for 12 series, 300 trials. The next day, he was given sodium amytal (5 gr.) and dropped from 13.5 per 25 to 7.8 per 25, an hour after. Three hours afterward he dropped to 6.2. Then he was given 5 gr. of caffein and came back up to 9.5. He did not know which drug he had until after the amytal made him very sleepy and he did not know what effect on E.S.P. to expect from either drug.
But the climax to our story, so far, consists of long distance E.S.P. conducted by Miss Turner as percipient, at Lake Junaluska, N.C., and Miss Ownbey as agent, here at Durham, a distance well over 250 miles. They arranged to run series of 25 trials at certain hours, the trials five minutes apart. The first eight series run as follows, per 25: 19, 16, 16, 7, 7, 8, 6, 2. This is an average of 10.1 per 25, with a positive deviation of 41, which is 10.8 times the probable error. The first three series are enormously significant, as is obvious, reaching a value of well over twenty times the probable error for chance expectation based on 75 trials.
What now about distance and P.C.? This has become a most crucial sort of question, indeed. Later on in the summer, Miss Turner and Miss Ownbey arranged for a long distance P.C. experiment (300 miles), Miss Ownbey simply selecting the card, isolating it on a table, but not looking at it. The cards are changed at regular known time intervals as were the images in the P.T. tests. These were to alternate daily with distance P.T. runs, conducted as those described above. The distance P.T. fell to chance average and stayed there for the 4 runs attempted. All we can say in explanation is that Miss Ownbey did not feel well for part of the period and was expecting to be married within a few weeks. What disturbance this might give to the necessary act of "concentration" we can only conjecture. At the same time, Miss Ownbey was acting as agent in distance-P.T. tests with Zirkle 165 miles away and these, too, yielded only 5.5 per 25 for 10 runs, which is not significant scoring. But in the
distance-P.C. Miss Turner was presumably alone responsible. She began with 4 in 25, 4 in the second run, then rose to 7, 8, 7. At this promising point they had to stop.
But just then Pearce began on distance-P.C. Not so far away from the cards, it is true. But for some features 100 yards are as good as 100 miles. He began with a distance of 100 yards, from the Duke Library Building to the Physics Laboratory. He got an average of 9.9 for 12 runs, a deviation of 59 above chance in 300 trials, which is 12.2 times the p.e. He did better at the distance than he did in the same room, with all other conditions equal. As I write this he is just beginning to score high at 250 yards, after an initial adjustment period which he always requires.
If distance is no barrier, either to P.C. or P.T., what then, about time? Can these E.S.P. subjects evade that dimensional factor also? Only that factor itself has kept us from attempting to find out but it cannot long do so now.
Footnotes
35:1 Rhine, J. B. and Rhine, Louisa E. An investigation of a "mind-reading" horse. Jour. Abn. and Soc. Psy. 23: pp. 449-466, 1929, and the sequel "Second Report on Lady", same journal, 24: pp. 289-292, 1929.
36:1 See Appendix to Chapter 2, page 32.