386 BUREAU OF AMERICAN ETHNOLOGY [Bull. 175 



more thrilling than the real affair had been. Thus, after describing 

 his ghostly sex partner's charms in considerable detail, he wistfully 

 concluded : "Ghosts stay young." ^'^ Finally — and this is the point 

 which is of the greatest importance in the present context — Hivsu: 

 Tuporma made it quite clear that his ghostly mistress enjoyed 

 their dream intercourse as much as he did and was as enamored of him 

 as he was of her. He himself yelled in dream with delight, but when 

 his wife asked him why he yelled he said : "I dreamed that someone 

 came to kill me." *^ This lie is probably an unconscious allusion to 

 the fatal enticements of the dead (pt. 4, pp. 128-186). 



In some instances the Mohave actually take it for granted that if 

 the victim does not reveal his magical killer's name, this proves that 

 the bewitched person is a willing victim. Thus, even though Tcuhum 

 contemptuously rejected the advances of Sahaykwisa:, because she 

 had been formerly a lesbian, the fact that he did not reveal her name 

 when she bewitched him was said to prove that he did love her after 

 all and wished to become her victim (Devereux, 1937 b, and Case 105). 



The next point to stress is that the witch's victim knows that, sooner 

 or later, the witch will succumb to the temptation of being forever 

 with his beloved victims, and will therefore commit vicarious suicide 

 (pt. 7, pp. 387-426) . Hence, the linked "suicides" of the victim and of 

 the witch resemble in many respects the suicide pacts of lovers,^* in 

 which the stronger of the two first kills the more timid partner and 

 then commits suicide. The similarity between such suicide pacts and 

 the linked voluntary deaths of the victim and the witch is further 

 increased by the fact that, just as suicide pacts are made by frustrated 

 lovers, so these linked witch-and-victim "suicides" usually involve 

 persons who — being related to each other, or at least of the same 

 gens (Case 105) — cannot marry, nor even have an affair, without 

 being socially penalized for it. 



As regards the real cause of the willing victim's death, it is more 

 likely to fit the "vagus death" pattern (Richter, 1957) than the fear- 

 and-rage type of psychic death described by Cannon (1942). 



Given the intimate link between the death of the willing victim and 

 the vicarious suicide of the witch, the latter form of "suicide" will be 

 discussed in the next section. 



« Presumably the witch's captive ghosts are meant, since ordinary ghosts are supposed 

 to grow old and to die In the land of the dead (Devereux, 1937 a). It la permissible to 

 suppose that captive ghosts stay young precisely because their killers remcmher them as 

 they were. By contrast, all other deceased persons are supposed to be forgotten as soon 

 as possible (Kroeber, 1925 a), which explains why the survivors can believe that they do 

 not remain forever as they were when they died. In fact, one may even suggest that 

 beliefs concerning the aging and ultimate "second death" of ghosts can occur only In 

 societies where the survivors seek to forget the deceased as soon as possible. 



" A complete account of Hivsu : Tupo :ma's confession was published elsewhere 

 (Devereux, 1948 1; and Appendix, pp. 505-548). 



•* Suicide pacta were cogently discussed by Jones (1951). 



