578 The Outline of Science 



the hypothesis of telepathy from the dead is regarded favourably 

 by some of those competent to judge. 



The standard classical instance of such an occurrence, as 

 narrated by the poets, is the appearance of the drowned Ceyx 

 to his beloved wife Alcyone, and her consequent veridical con- 

 viction of his fate. The story is beautifully told, with full cir- 

 cumstance and vividness, in the eleventh book of Ovid's Meta- 

 morphoses. But it is noticeable that in this instance Ceyx had 

 been dead for some days when the phantom appears, and the 

 poet takes it to be a messenger from the gods, assuming the form 

 and voice of the dead man in order to convince Alcyone of the 

 truth. 



And yet telepathy, though wide in its range, does not cover 

 all the ground. It has to be stretched considerably in order to 

 account for many apparitions, and especially for what is called 

 the "fixed local" ghost, that is to say, an apparition said to be 

 encountered in association with certain houses or places with the 

 reputation of being haunted; any stranger being said to be 

 able to see the apparition at suitable times, even if he were 

 ignorant of the legend and unacquainted with the traditional 

 haunting. 



The first thing is to make sure that the facts are as described, 

 and that such persistent haunting is a reality. It seems wisest 

 to preserve an open mind on that subject ; for the evidence, though 

 noteworthy, is not yet considered as crucial as that for the other 

 class of phantasms the class more readily conceived of as due 

 to transmitted mental impression. There appears to be a cer- 

 tain objectivity about this fixed local ghost: there seems no 

 obvious agent to whom to attribute a telepathic impulse; and 

 besides, the things the ghost is sometimes said to do are hardly 

 consistent with a mere mental impression though certainly the 

 hypothesis of a vivid dream on the part of the percipient must be 

 allowed all the benefit of any doubt there may be, and the 

 burden of proof that there is anything objective in the experi- 



