January 7, 1887.] 



SCIENCE. 



19 



make them accurate ; learning, of the solidest 

 sort, in discussing them ; in theorizing, subtlety 

 and originality, and, above all, fairness, for the 

 work absolutely reeks with candor, — this com- 

 bination of characters is assuredly not found in 

 every bit of so-called, scientific research that is 

 published in our day. 



The book hardly admits of detailed criticism, 

 so much depends on the minutiae of the special 

 cases reported : so I will give a broad sketch of 

 its contents. The title, ' Phantasms of the living,' 

 expresses a theory on which the recorded facts are 

 strong, but of which the latter are of course inde- 

 pendent. The ' facts ' are instances of what are 

 commonly called 'apparitions.' Collected for the 

 Society of psychical research, their sifting and cata- 

 loguing is a laborious piece of work which has a 

 substantive value, whatever their definitive expla- 

 nation may prove to be. Very roughly speaking, 

 there are reported in the book about seven hun- 

 dred cases of sensorial phantasms which seem 

 vaguely or closely connected with shme distant 

 contemporaneous event. The event, in about 

 one-half of the cases, was some one's death. In 

 addition to these cases, Mr. Gurney has collected 

 about six hundred of hallucinations seemingly ir- 

 relevant to any actual event, and thus has cer- 

 tainly a wider material to work upon than any 

 one who has yet studied the subject of phantasms. 

 Of course, the rationalistic way of interpreting the 

 coincidence of so large a number with a death or 

 other event, is to call it chance. Such a large 

 number of ' veridical ' phantasms occurring by 

 chance would, however, imply an enormous total 

 number of miscellaneous phantasms occurring all 

 the while in the community. Mr. Gurney finds 

 (to take the visual cases alone) that among 5,705 

 persons, interrogated at random, only 23 visual 

 hallucinations had occurred in the last twelve 

 years. And combining by the calculus of proba- 

 bilities such data as the population drawn upon 

 for the coincidence-cases, the adult population of 

 the country, the number of deaths in the country 

 vrithin twelve years, etc., he comes to the conclu- 

 sion that the odds against the chance occurrence 

 of as many first-hand and well-attested veridical 

 visual phantasms as his collection embraces, is as 

 a trillion of trillions of trillions to 1. Of course, 

 the data are extremely rough ; and, in particular, 

 the census of phantasms occurring at large in the 

 community ought to be much wider than it is. 

 But the veridical phantasms have, furthermore, 

 many peculiarities. They are more apt to be 

 visual than auditory. Casual hallucinations are 

 oftener auditory. The person appearing is almost 

 always recognized ; not so in casual hallucina- 

 tions. They tend to coincide with a particular 



form of outward event, viz., death. These and 

 other features seem to make of them a natural 

 group of phenomena. 



The next best rationalistic explanation of them 

 is that they are fictions, wilful or innocent ; and 

 that Messrs. Gurney, Myers, and Podmore are vic- 

 tims, partly of the tendency to hoax, but mainly 

 of the false memories and mythopoietic instincts 

 of mankind. These possibilities do not escape our 

 authors, but receive ample consideration at their 

 hands. Nothing, in fact, is more striking than 

 the zeal with which they cross-examine the wit- 

 nesses ; nothing more admirable than the labor 

 they spend in testing the accuracy of the stories, 

 so far as can be done by ransacking old newspa- 

 pers for obituaries and the like. If a story con- 

 tains a fire burning in a grate — presto the Green- 

 wich records are searched to see whether the 

 thermometer warranted a fire on that day ; if it 

 contains a medical practitioner, the medical regis- 

 ter is consulted to make sure he is correct ; etc. 

 But obviously a hoax might keep all such acces- 

 sories true, and a story true as to the main point 

 might have grown false as to dates and accesso- 

 ries. It therefore comes back essentially to the 

 investigator's instinct, or nose, as one might call 

 it, for good and bad evidence. A born dupe will 

 go astray, with every precaution ; a born judge 

 will keep the path, with few. Saturday revieivers 

 will dispose of the work in the simplest possible 

 way by treating the authors as born dupes. ' Sci- 

 entists ' who prefer offhand methods will do the 

 same. Other readers will be baffled, many con- 

 vinced. The present writer finds that some of 

 the cases accounted strong by the authors strike 

 him in the reading as weak, while scruples shown 

 by them in other cases seem to him fanciful. 

 This is the pivot of the whole matter ; for I sup- 

 pose the improbability of the phantasms being 

 veridical by chance, will, if the stories are true, 

 be felt by every one. Meanwhile it must be re- 

 membered, that, so far as expertness in judging 

 of truth comes from training, no reader can pos- 

 sibly be as expert as the authors. The way to be- 

 come expert in a mafter is to get lots of experience 

 of that particular matter. Neither a specialist in 

 nervous diseases, nor a criminal lawyer, will be 

 expert in dealing with these stories until he has had 

 Messrs. Gurney's, Myers's, and Podmore's special 

 education. Then his pathology, or his familiarity 

 with false evidence, may also serve him in good 

 stead. But in him, or in them, ' gumption ' will, 

 after all, be the basis of superiority. How much 

 of it the authors have, the future alone can decide. 



One argument against the value of the evidence 

 they rely on is drawn from the history of witch- 

 craft. Nowhere, it is said (as by Mr. Lecky in his 



