There is a kind of hallucinations—­namely, Phantasms of the Dead—­about which it seems better to say nothing in this place.  If such phantasms are seen by savages when awake, they will doubtless greatly corroborate that belief in the endurance of the soul after death, which is undeniably suggested to the early reasoner by the phenomena of dreaming.  But, while it is easy enough to produce evidence to recognised phantasms of the dead in civilised life, it would be very difficult indeed to discover many good examples in what we know about savages.  Some Fijian instances are given by Mr. Fison in his and Mr. Howitt’s ‘Kamilaroi and Kurnai,’ Others occur in the narrative of John Tanner, a captive from childhood among the Red Indians.  But the circumstance, already noted, that an Australian lad became a wizard on the strength of having seen a phantasm of his dead mother, proves that such experiences are not common; and Australian black fellows have admitted that they, for their part, never did see a ghost, but only heard of ghosts from their old men.  Mr. David Leslie, previously cited, gives some first-hand Zulu evidence about a haunted wood, where the Esemkofu, or ghosts of persons killed by a tyrannical chief, were heard and felt by his native informant; the percipient was also pelted with stones, as by the European Poltergeist.  The Zulu who dies commonly becomes an Ihlozi, and receives his share of sacrifice.  The Esemkofu on the other hand, are disturbed and haunting spirits[1].

As a rule, so far as our information goes, it is not recognised phantasms of the dead, in waking vision, which corroborate the savage belief in the persistence of the spirit of the departed.  The savage reasoner rather rests his faith on the alleged phenomena of noises and physical movements of objects apparently untouched, which cause so many houses in civilised society to be shut up, or shunned, as ‘haunted.’  Such disturbances the savage naturally ascribes to ‘spirits.’  Our evidence, therefore, for recognised phantasms of the savage dead is very meagre, so it is unnecessary to examine the much more copious civilised evidence.  The facts attested may, of course, be theoretically explained as the result of telepathy from a mind no longer incarnate; and, were the evidence as copious as that for coincidental hallucinations of the living, or dying, it would be of extreme importance.  But it is not so copious, and, granting even that it is accurate, various explanations not involving anything so distasteful to science as the action of a discarnate intelligence may be, and have been, put forward.

We turn, therefore, from a theme in which civilised testimony is more bulky than that derived from savage life, to a topic in which savage evidence is much more full than modern civilised records.  This topic is the so-called Demoniacal Possession.