Whoever wishes to prove the reality of telepathy gives any interpretation he likes to a drawing. A book by an anonymous writer, H., that appeared in 1848, contains much interesting information on prophesying. We are told that should the event not agree with the prophecy, we must remember that everything contained in a prophecy is not to be taken literally, so much is only metaphorical. As an example of the ambiguity of prophecies Hennig cites the case of Malachi, who is said to have characterized each Pope by a short descriptive epitheton. Thus Leo XIII. is said to be referred to in the words Lumen de Ccelo, and his successor is characterized by Ignis ardens. Now, Hennig shows to how many cardinals this prophecy might apply if only a little ingenuity were exercised in interpreting it. It would describe that restless being Rampolla accurately. Cardinal Svampa might also have been meant, for svampa is Italian for torch; but it would also apply to Cardinal Gotti, whose coat-of-arms contains a torch; or to Cardinal Capecelatro, an idealist filed with modern ideals. But it would apply even better to Cardinal Hohenlohe, whose name is a literal translation of Ignis ardens.

But when Cardinal Sarto, of whom nobody had ever thought, was made Pope, Malachi's prophecy was interpreted as pointing to him. But as there was nothing fiery about Cardinal Sarto, either intellectually or in his temperament, and, on the other hand, Malachi could not have prophesied falsely, it was discovered that the coat-of-arms of S. Dominic, on whose day (August 4th) Sarto was elected Pope, contains a dog from whose mouth a flame issues, and it was at once said that the prophecy Ignis ardens referred to Cardinal Sarto, and thus Malachi's prophecy was fulfilled. Hennig has done a great service in dissecting this case so thoroughly. It is only by quibbling that so many, many prophecies are made out to have come true.

Spiritists have lately been very fond of quoting the following case as a proof of clairvoyance. On the evening of June 10th, 1903, Charles Richet held a mediumistic seance. The people present wished to hear "raps," and they did. The following words were rapped out: - "Bancalamo" and "Rtguettefamille." Directly Richet read these words he is supposed to have been struck with the idea that they formed a complete sentence with a definite meaning, and he deciphered them as follows: - Banca la mort guette famille. This communication was rapped out between 10.45 P.M. and 11 p.m. On the following day the news arrived that Queen Draga of Servia, together with the king and her family had been murdered overnight, and the sentence was immediately made to refer to that crime. To accomplish this the following method of interpretation was necessary. Banca was converted into Panta. Richet thought it was so easy to mistake a B for a P. The Germans have a way of saying ponne boutcille instead of bonne bouteilie, and the Servian pronunciation of T is best represented by the French C or Z. The name of Queen Draga's father was Panta, and the sentence consequently referred to him and a threat to murder his family.

Richet certainly admitted that the words, "Ma mort guette familie," might apply to any number of families, and that it was consequently necessary to calculate the probability of Banca having been rapped out correctly. This he found to be 1:10,000. Consequently the choice lay between coincidence and clairvoyance. Richet assumed that it was a case of clairvoyance* because a Press Agency in Paris received the news of the murder at 10.45 p.m. the same evening.

I have asked different people what they would read into such a sentence as "Banca la morte guette famille," and most of them replied that the word "Banca" would make them think of "a bank," and that they should never think of turning it into "Panta." At all events, I should have been much more inclined to find the case one of clairvoyance if any great bank had failed and driven one or two people to suicide. The theory of clairvoyance would have applied equally well to either supposition, mine or Richet's. Another man I questioned thought the sentence referred to death on a glacier. He considered "Banca" meant "Bianca," white death, or death on a glacier. I think that any one from the Sunda Islands would have referred it to Banca, an island east of Sumatra. A man, who was interested in ancient arms, said it must mean Bancal, a term for a curved sabre. A lady to whom I showed Richet's lettering, "bancalamo ruguettefamille," said it made her at once think of calamus, as if something had been done with calmus. Another man said it reminded him of a Ban of Croatia, because "ban" was the French for it.

I looked this up in the dictionary, and can only say that "ban" has so many meanings that it might be made to supply a dozen - nay, hundreds and even thousands of interpretations of that prophecy. But Richet was determined to have it that the murder of the Servian Royal Family was conveyed to Paris by means of clairvoyance, and so he proved it.

Richet's method of turning Panta into Banca and Banca into Panta reminds one of those philological tricks against which we were warned in our school-days - the little dodges by means of which one word could be derived from another by changing a letter; always, of course, on apparently scientific grounds.

6. The probability of chance success. As many experiments fail, it should be considered whether the number of successful ones exceeds probability. Preyer, however, is inclined to think that for our objects statistics of probability have little significance. But if such calculations are to be taken into consideration they must be made correctly. Richet's case shows how little this is thought of. He calculated the probability of "Banca" having been correctly given at 1 to 10,000. But he quite overlooked the fact that " Banca " was not spelled as one word; it was wrapped out as "Bancalamo." He also overlooked the fact that if the sentence had not been made to fit in with the murder of the Servian king, it might equally well have been made to refer to a thousand other episodes, as I showed in my explanation of fictitious interpretations. Moreover, he did not take into account the frequency of false prophecy, or, more especially, how often • things have been communicated at his own seances which were never substantiated. It is only when all these matters are taken into account that we can ascribe any reasonable importance to probability.