III. Experiments when agent and percipient were either thirty feet apart in different rooms, with a closed door between, under which circumstances it is incredible that involuntary whispering could have been heard, or thirty miles apart, in which case it is unquestionably impossible that involuntary whispering could have carried. Theory: _Chance coincidence_. This is the only theory left for such experiments, unless conspiracy is charged, and that at different times would have to include not only Mr. and Mrs. Sinclair, but Mr. Irwin, Mrs. Irwin, the Sinclairs’ secretary and Professor McDougall. Refutation of the theory: _The experiments in this class were of such number and had such success both in number and quality as to challenge the production of any such success by guessing though hundreds of series each of an equal number of experiments should be gone through with_. It is credible that the large percentage of Successes and Partial Successes in the first 14 experiments and 24 among the latest ones should have been obtained by one method, that (aside from these) during the earlier months another and quite different method should have been employed, and that (still aside from these) later a third and quite different method should have been resorted to, and yet the whole mass of results be homogeneous? It would certainly be expected that the inauguration of any new method would in some way be reflected in the nature of the results. But the lot produced with intervening distances too great to admit of the involuntary whispering theory melts imperceptibly into the lot produced with the agent and percipient together so that the involuntary whispering process is conceivable, and this in turn melts imperceptibly into the lot where all precautions are discarded, and this again into long-distance experiments and out, without it being possible to detect any changes in the character of the results at the points of junction. Throughout there is homogeneity, some successes being correct literally, some incompletely and partially, some results only suggestive and some entire failures. Throughout we find some corresponding in both shape and meaning, some in idea but not shape, and some in shape only and misinterpreted by the percipient; in fact, all the peculiarities of Mrs. Sinclair’s work are to be found in about equal proportions in all stages. There is perceptible a gradual though irregular tendency to decline in the ratio of success achieved, but in such a manner that the decline cannot be chronologically connected with any of the changes of method. The “peeking” theory cannot be applied to the experiments of Class I. The “involuntary whispering” theory cannot be applied to the experiments of Class II. Neither the “peeking” nor the “involuntary whispering” theory can be applied to experiments of Class III.