So far as the strength and weakness of the arguments for mental telegraphy depend upon the perspective of value attached to the various data and to the conditions under which these have been gathered, I have presented my estimate and indicated the burden of my conclusions. But I am aware that I may have laid myself open to the charge--which will be brought not by the advocates of telepathy, but by its most emphatic opponents--of a neglect of consideration of the general logical status of telepathy as a germane and legitimate hypothesis. That the hypothesis of telepathy when carefully interpreted is capable, if not of explaining the data, at least of being fitted without undue straining to a large portion of the data, may be claimed with some plausibility; that I regard the hypothesis as unwarranted and unnecessary has been made sufficiently clear. But what if the hypothesis is not a legitimate one, not one which the methods and spirit of science can properly or profitably consider? If this be the case, it would seem superfluous to consider whether the hypothesis is warranted by the data or capable of explaining them. That it is the policy of science to allow the utmost latitude of opinion and theory and to interpret the possible in an unprejudiced and liberal spirit will readily be conceded. That it is equally the policy of science to demand of all claimants for recognition authentic credentials framed in accordance with the laws of logic and the principles of evidence and probability, is sometimes overlooked. Science cannot possibly consider all hypotheses, but only legitimate ones. To explain coincidences and the success of experiments in thought-transference by assuming that there is a demon, whose special business it is to make people have uncanny feelings when their relatives in distant places are dying or in danger, and to suggest to the guesser what is in the mind of the party of the second part in the experiment, is certainly not an hypothesis worthy of consideration by science; and incidentally be it noted that this hypothesis may be successfully shaped to fit the facts, and cannot be definitely disproved. Some absurd hypotheses may be readily disproved and others not; but are scientists really called upon to disprove them? There recently fell under my observation a claim for the theory that when persons felt an unaccountable aversion for one another, either at once or after a time of friendship, it was due to their opposite horoscopic natures, and it would be found that their birthdays were not far from six months apart, that is, nearly as far apart as they possibly could be. Divorces, breaches of promise, family feuds, and antipathies at first sight could thus be accounted for. Now, it would involve no very burdensome undertaking to disprove this theory; but I should not expect a cordial approval of my efforts on the part of my colleagues if I carried through the investigation.