158 BUREAU OF AMERICAN ETHNOLOGY [Bull. 175 



the patient is too weak for me to hurt him. Then the stomach gas goes out 

 either way (eructation or flatus).'^' I blow on the solar plexus first and 

 then immediately press my hand on the same spot, to keep my breath on it. 

 I may also blow on the patient, starting from the upper part of his chest 

 and progressing toward his abdomen, to remove the pains. If the pains are 

 strong, I will also boil the bark of the white mesquite tree, or make a willow 

 bark tea and give it to the patient to drink. 



Summing up, at first the child's body consists only of menstrual blood which 

 is kneaded into human shape by the father's sperm. If the pregnant woman 

 then cohabits with some other man or men, the semen which she receives 

 molds the ctiild to the likeness of the new partner (s).'^ The ghost of the dead 

 child sometimes returns to its mother's womb and, without help from human 

 semen, makes a body for itself out of the menstrual blood which accumulates 

 between (two or more) menstrual periods. This disease kills the mother. "We 

 believe that the ghost child kills its mother and takes her soul to the land 

 of the ghosts. 



Curing songs: The final version of the following songs approximates rather 

 closely the actual text of Harav He:ya's songs, which varied considerably as 

 he repeated them. Harav He :ya's singing was interspersed with copious throat 

 clearings, stage whispers, explanations given in a kind of recitative which 

 could barely be told apart from the songs themselves, and even with extra 

 syllables, allegedly caused by "shortness" (sic) of breath, (e.g., oh-eyam, in- 

 stead of eyam ) . The entire performance of this toothless, stubborn, but willing 

 old man, who lost his bearings and had to start all over again whenever he 

 was asked to sing more slowly, caused the interpreter to exclaim, "If we 

 hadn't had Hivsu: Tuporma along, who had heard these songs before, I 

 couldn't have grasped a single word for it." In fact, even Harav He : ya's alert 

 old wife and his highly cooperative nephew, who had heard him sing these 

 songs many times, were unable to catch certain words well enough to repeat 

 them for me. 



There are four songs altogether. The first one had to be recorded three 

 times and the second and third twice. The fourth was recorded only once 

 because, by that time, Harav He :ya had finally learned to sing more slowly. 

 The differences between the various recordings suggest that the actual text 

 of the songs is flexible. This supposition is wholly compatible with the fact 

 that, unlike the ritual formulas of obsessive-compulsive ethnic groups, the 

 effectiveness of Mohave songs does not depend on their letter-perfect reproduc- 

 tion, but on the singer's possession of the necessary powers. In fact, several 

 informants stated that when a certain illness is not treated by anyone for a 

 number of years, because no one had received suitable powers in dream, the 

 text of the traditional songs may be lost forever, so that the next person to 

 receive the necessary powers will simply recite the (condensed?) prose text 

 of the relevant myth, instead of making up new songs." The variability of 

 Mohave ritual texts is, both from the anthropological and from the psycho- 

 logical point of view, too complex a problem to be discussed in this context 

 (Devereux, 1957 b). 



"Jones' (1951) analysis of Hindu pneuma-theorles Is Illuminating In this context (see 

 also part 1, pp. 9-17). 



" Note the nexus between this theory of modifiable paternity (Devereux, 1949 c) and 

 theories of ghostly pseudocyesis. 



" It may be assumed, however, that even such narratives will approximate the singing 

 style and will consist primarily of a series of Ijey words. 



