We have already seen that the relations between the shaman and the spirits may be defined as those between «master» and «servants». However, such are the relations only when the shaman is really a good and strong person, while a weak person may easily become the prey and an instrument of the spirits.
There is no doubt that from the point of view of the shaman's psychology there are two main types of spirits. There are spirits which actually are various mental and psychological conditions which the shaman needs for his performance of social functions and for his own use, when ordinary methods of thinking are not sufficient for solving various problems. There are also spirits which symbolize conditions of psychic disfunction, and which ought to be «mastered» by the shaman, if he is not to become susceptible to the harmful conditions of various psychoses.
In the description of the history of shamans we have seen that the shaman may begin his life career with a psychosis, but he cannot carry on his functions, if he does not master spirits, or, as has been shown, does not master himself.
By such an approach to shamanism, in so far as the psychology of the shaman is concerned, I leave aside all earlier explanations as to the shaman's psychic condition, which were mostly the result of aprioristic reasoning based on the principle of analogy. These are theories of no interest, except that they depict the European complex. However, there is a new theory which deserves special attention, because of its recent and scientific, in appearance, form. Here I have in view L. Sternberg's theory of the sexual character of the shamanistic complex, generalised up to covering shamanism as a whole and disclosing similarity between ethnographically most remote phenomena. L. Sternberg's Divine Election in Primitive Religion was first promoted by the observation of a Goldi shaman who confessed about his sexual dreams to L. Sternberg who says: «the sexual motive of Election in shamanism, a matter so simple and natural to the mind of primitive man, never occurred to me before having met the Goldi Shamans» (p. 481). So we can start from the original point, i.e. the Goldi.
According to L. Sternberg there are assistant spirits and supreme spirit who elect the shaman, whence another question arises as to «the intimate relation between the elected person and the electing spirit, the problem of the motives, why a certain spirit fixes his choice upon a certain individual and becomes his patron and helper» (op. cit. p. 475). It is evident that a distinction between assistant spirit and supreme spirit is introduced, a distinction which contains an answer — the character of the spirit — «supreme» is presumed, and thence another conclusion as to «choice» is made, further modified and identified with «election», where the conception of «patron and helper» is only natural. All these elements, as «election», «supreme», and «patron», are absolutely alien to the Tungus conception, and the actual key of the theory resides in the first quotation, when a whole construction is built up. A guess and scaffolding of facts supposed to support it!
The fact recorded is a case of a young male shaman who becomes lover of one of the spirits, which also become the principal spirit of the shaman. L. Sternberg especially stresses that the spirit is called ajami, while other spirits are called sywen (I have some doubt as to the transcription; cf. supra). However, there is nothing particular in this fact. What is designated by ajami has been shown by P. Simkevich and I. A. Lopatin as the chief spirit which is the head of groups of spirits, i.e. exactly what we have seen in the Manchu complex. Naturally, in every particular case of individual shamans, all depends on which group of spirits is mastered by the shaman. The investigators of Goldi have shown that any group may be so mastered and there will be at the head of a group — usually a series of shaman manifestations — an ajami (cf. P. Simkevich, op. cit. Materials). Therefore there may be different ajami. Moreover, the Goldi term ajami is not applied only to these spirits (cf. e.g. I. A. Lopatin, op. cit. p. 814 and many others) — ajami fon'alko figures as a placing for the spirit of a deceased person. Exactly the same situation is observed among the Manchus and Tungus. The first mastered spirit, which helps to master other spirits, naturally holds a special position in the group of mastered spirits. The term ajami, in which m'i is a suffix, is perhaps not unknown in other Northern Tungus dialects in which it appears as aja(n)~haja. However, I do not insist upon the parallels. In Manchu such a spirit would be called dalaxa vochko, and it is not a «supreme spirit» and «patron».
Now the question is: whether the first spirit is always of an opposite sex or not? L. Sternberg wants it to be so and he gathers facts from various sources, and even from different groups. In the first place he quotes my earlier publication in Russian (Es-say, etc.) in which I have given a brief account of the «election» of a new shaman, already related in details in the preceding chapter (vide supra Section 131). As a matter of fact, the female candidate was rejected exactly on the ground that she was possessed by the spirit of her lover, but did not master it, and she was considered as an abnormal person who cannot become shaman. L. Sternberg has omitted this important condition, but he puts emphasis on sexual symptoms of the hysterical fit described. Then, putting stress on the statement of a shaman (p. 478), who was a beginner and did not speak good Russian [662] and who was evidently affected by some sexual trouble, L. Sternberg asserts that another shaman had a female spirit too, and proceeds to another generalization: the movements of a shaman have a sexual character when he is dancing (p. 481). Really such an enquiry is not easy when it is carried out in a foreign language and during a short stay with the people [663]. L. Sternberg, being carried away by his imagination, could not see the simple fact that without moving the low part of the trunk to and from one cannot produce sounds of the trinkets, but these movements can have a sexual meaning only in an mind affected by an idee fixe. Practically all of these evidences were gathered among the Manchus and the Goldi. As to other facts gathered among the Yakuts and Buriats, they need very careful scrutinizing before being used as evidence [664] and naturally, first of all, a detailed investigation on the spot is needed. I leave aside other evidences brought forth by L. Sternberg, for they are selected as regards their nature and thus a strictly scientific generalization is not possible [665]. However, I am far from the idea of denying a possibility of such occurrences in some ethnical groups, as a general phenomenon, and as exceptional cases among the Manchus and Tungus. For instance, in the case related by me, the female candidate might have become shaman, and thus she would be possessed by a male spirit (not mastering him) on a perfectly sexual ground. However, in so far as the Tungus and Manchus are concerned, no generalization of this kind is possible and no condition of «election by the spirits» is observed among these shamans. So the new theory ought not to be extended to the groups here discussed.
* * *
In the previous description (vide Section 128) we have seen which are the preliminary conditions for becoming shaman. Such ones may be a mass psychoses which affect a great number of clansmen, or a limited number of them; as a matter of fact, if there is a delay in the «election» of a new shaman, the psychosis may affect a very great number of clansmen. Yet, the extent of psychosis depends also upon the character of the group connected with the spirit. There are cases when the spirit is not recognised by all clansmen but only by a group, even a family (i.g. the case of the wujala clan; vide supra Section 131), so that only a very limited number of persons may be affected by a psychosis. Moreover, there may be individual cases of psychosis in which a person is usually affected by a newly discovered spirit, very often of a foreign origin (dona among the Tungus of Manchuria). Indeed, such a foreign and individual spirit may become effective on other clansmen as well, and, as shown, it may even become a clan spirit. Lastly, there may be no psychosis sharply manifested, but a fear of its occurrence, which may stimulate «election» of a new shaman; moreover, there may be the need of a new shaman, as a regulator of the psychic life of the group (clan or territorial unit), and this will be sufficient to begin the «teaching» of a candidate for shamanship.
It is thus evident that from the very beginning different initial relations may be established between the spirits and the candidate to shamanship. There is no uniformity.
In case of mass psychosis, the candidates will, at least for a while, be mastered by the spirits and all of them, except one, will not be able to master the spirits, which is brought about by the idea of the existence of the clan shaman and by the idea that, being mastered by a person, the spirit will not harm other people.
In case of individual psychosis, according to the theory, the person is possessed by a spirit, and thus the mastering of the latter is one of the means of curing the candidate.
In case of mastering spirits for preventing them from doing harm to the candidate and other people, also for curing the person affected by a spirit, the candidate becomes at once a master of spirits.
One thing is evident: in all these cases, when the candidate becomes shaman he functions as «master» of spirits, which is expressed by the term ejin-ejen of the Tungus dialects and in Manchu, and which can only be interpreted as «master». These relations are expressed in Manchu as ejimbe k'ichalemb'e (ejen be kichelembi, Manchu Writ.) i.e. the spirits are working for their master [666].
Indeed, in case of psychic troubles, due to unmastered spirits, the shaman actually masters himself, regulates his own psycho-mental complex, after which he is no more affected by the condition which was the initial cause of his becoming a shaman. The more he practises his art of self-control, the more spirits he masters; the more skill he acquires in the process of his practising, the more spirits (psychic conditions, or imaginary dangers) he masters.
Among the Tungus groups, at the beginning of a shaman's career, the shaman may have a very limited number of mastered spirits. Among them the first spirit, which comes into the candidate and which later on is used for taking control over other spirits, comes sometimes very rarely during the shaman's life and perhaps even only once in his life. For instance, among the Khingan Tungus a female shaman became shaman at the age of fifteen years when a female spirit introduced itself into her [667]. After that time, for twenty years, this spirit did not come, and perhaps, I was told, it would never come again. However, this is the greatest spirit of the shaman and it receives from time to time regular sacrifices [668]. Among the Kumarchen such a spirit is usually called ogdinga seveng), i.e. the greatest shamanistic spirit. However, ogdinga may also be referred to «good», «strong» spirits in general; the shaman will thus be called ogdinga saman. The same refers to the Birarchen. Among them I observed that the greater part of the shamans begin by possessing a complex malu which, as shown, is sexless. The female shamans in the beginning usually have only animal manifestations of this complex, namely, a lizard, a snake and a turtle which are simple manifestations used for hiding and travelling to distant places (an exterioration of soul, i.e. «sublimation of the spirit», liberation etc.). The other manifestations of malu come later. For instance, as I have shown, a shaman beginner had only three spirits: a malu, a kadarn'i and a dona Manchu spirit. With the age and practice the number of spirits may increase up to more than thirty and, besides the clan spirits, there may be various dona spirits. Among the latter, as a rule, there are Yakut (joko) spirits (the shaman is then supposed to speak the Yakut language), Tungus (teya), and seldom Manchu, Dahur, and even Chinese spirits. There may also be individual spirits which may easily leave the shaman (the clan spirits cannot do it).
Among the Manchus the situation is slightly different. Since the list of spirits, owing to the existence of writing, is known and cannot be very much altered, the shamans believe that they must master first two spirits, namely mafa saman and mama saman (whose names I could not find out) who are supposed to be the most important spirits and who undoubtedly originated among the shamans. Naturally there are «ancestors» in every clan, and they come, at the same time, as chief helpers of the shaman [669]. However, they do not become dalaxa vochko (leading spirits); such a spirit may be gradually formed out of existing spirits, or as one of the new spirits individually incorporated by the shaman. This is the spirit with which the man works more than with other spirits and which takes charge of all spirits. Among the shamans the choice is subject to great variations. There are usually present some Tungus (kilin), Chinese (both n'ikan and jergin), Mongol and Dahur spirits which incidentally may become dalaxa. The number of spirits mastered at the beginning or the career is small, but gradually it may increase. The shamans usually have at their disposal from twenty to sixty spirits, but not all of them are used. However, theoretically all of the spirits of the list are supposed to be mastered. There are the individual lists of spirits actually used in different clans are not alike. Together with this the ritualism connected with the spirits is also different.
It may be generalized as follows: the shaman is the master of spirits; the first spirit mastered may have, but not always necessarily, a special position; it may be one of the clan spirits and it may be a foreign spirit; there is no definite tendency as to the sex of the shaman and that of the first spirit, and of the spirits of importance (dalaxa in Manchu, ogdinga in Tungus); in general the number of spirits is subject to variations conditioned by the personal ability of the shaman, his experience, and thus by the period of shamanistic activity, as well as by a gradual decrease of the number of spirits, owing to the changes which occur in the shaman's activity.
* * *
Although the mastering of spirits is a characteristic feature of shamanism, the relations between the shaman and the spirits are not like those of a master and his slaves, deprived of their will and rights. The shamans, to a certain degree, transfer human relations to those between themselves and the spirits. Since the Tungus and Manchu complex recognizes the necessity of considering the character of persons whom they master, e.g. junior relatives, slaves (Manchus), workmen, subordinate soldiers and officials etc., they also consider the character of the spirits. The shaman knows that these spirits may have various desires in accordance with their particular character; he knows that there are bad-natured, wicked spirits, spirits which are not to be trusted, even for a moment spirits which may be trusted; there are spirits anxious to have benefits from their work for the shaman, and spirits which are not very greedy. To be brief — the characters of spirits are the reproduction of those of human beings, but in addition some spirits are more powerful than men, so the managing of them presents great difficulties. The shaman must be ready for all sorts of eventualities and he must carry on quite a complicated policy, for he may sometimes greatly suffer from the spirits. The question is what kind of policy should be adopted? If the shaman shows a weak heart by serving sacrifices too frequently, the spirits may form the idea that the shaman is afraid of them and they will not work for him, but will require more and more sacrifices. So he must not abuse this method. There are spirits which must be kept in strong hands and even they must be badly treated by the shaman; as long as the shaman manifests no fear, he is safe, but if the spirits see him weakening they may attack him immediately or may leave him alone and so he will be attacked by other spirits [670]. If at any moment the shaman loses his control, they may lead him to death. I have already quoted a case of a young shaman who committed suicide. Although rare, such cases do occur from time to time. Psychologically this case is clear: the shaman loses control of himself and the idea of persecution and self-suggestion may bring him to a fatal issue. In order to neutralize dangerous spirits he may create a kind of conflict among the spirits, so that, for a time at least, he will not be harmed. Thus, even the methods of controlling the spirits are a reproduction of the methods used among living people.
It a remarkable that in a great number of cases the spirits may require of the shaman the performance of a shamanizing and that the shaman cannot refuse to do it — this is the moment when the shaman wants to shamanize. Conflicts among the spirits, created by the shaman, are a remarkable reflection of the psychic life of the shaman, in which there are conflicts of contradictory elements of the shaman's complex, the balancing of which requires additional spirits for fighting other spirits. This method is not considered (by Birarchen) as a harmless one, for since the shaman comes into a conflict with his own servants, the spirits, he may lose his control altogether. When the shaman is involved in such a fight, he usually finishes his life by an accident, by disease, exhaustion, and, as far as I know, by insanity caused, according to the Tungus, either by his own spirits or by other unmastered spirits.
The shaman's tragedy is that he cannot dismiss clan spirits himself, he can dismiss only those spirits which are foreign, or his personal spirits. But even this task is sometimes beyond the shaman's art, and the spirits may return again and again. Generally speaking, the dismissal of spirits is not a very frequent practice at the beginning of the shaman's career; but an experienced shaman, after a perfect mastering or spirits, can do it. An experienced shaman may only keep his eye on the clan spirits and they will not bother him; they would even require of him no shamanizing or sacrifices. In rare cases the spirits may leave the shaman alone, even before a very advanced age, or even before he becomes very experienced in handling spirits. However, in this case, he will not be able to carry on his function of a safety valve — some other person may be affected by the spirits, and the latter may be mastered by a new shaman. These cases exactly correspond to the changes in the psychic condition of the shamans when with the advance of age conflicting conditions of the psychomental complex are eliminated by balancing or by dropping of some elements (e.g. the sexual complex etc.). However, in case the shaman loses the confidence of the clansmen, the spirits will affect other members of the clan, and the shaman would cease to function as a shaman. When the shaman himself is affected by the spirits, another, but always a stronger and more experienced, shaman may come to his assistance. However, such an interference of another shaman is not liked by the shamans, for there is a danger for the shaman who is helping another shaman that the spirits of the affected shaman may mix up with the spirits of the helping shaman.
Such complex relations, greatly «anthropomorphized» by the Tungus, may look like a very naive and artificial construction of an imaginary world of spirits. However, if we carefully analyse the character of various spirits and all hypotheses connected with them, we shall find that these spirits are merely a symbolization of various psychomental elements and complexes. I have also pointed out that the Tungus do not represent these spirits as anthropomorphic beings, but they are figured, in the Tungus mind, as immaterial (in Tungus sense) «beings», which are satisfied with an immateriality of sacrifice, human actions etc. To describe the Tungus attitude towards these spirits, as it may be inferred from the sacrifice, picturing «ugly looking horrible idols», «queer shaman's costumes», «frightening spirits» and spirits protecting the shaman from other spirits, is but to show a perfect ignorance of this complex problem [671]. These spirits are hypotheses, some of which are admitted by the European complex as well, hypotheses which formulate observations of psychic life of the people and particularly that of the shaman, and which are quite helpful in the regulation of the psychomental complex to which the Tungus have come after a long period of adaptation.
Let us consider some situations. The shaman must offer from time to time a sacrifice to his spirits; if he does it too frequently, they may disobey; if he neglects them, they may harm him, avenge themselves. If we replace «spirit» by a «certain psychomental element», the regular sacrifice will be understood as a practical means to have a constant check on the shaman's complex by the shaman himself. After the sacrifice the shaman feels sure that he is in control of the spirits. If the shaman is afraid of the spirits, it merely means that there is some essential change in his complex, which he cannot master, and he tries (according to the Tungus conception) to satisfy the spirits with a pleasant sacrifice — he is no more in control of the spirits and they may affect other people who will no longer have confidence in the shaman. Therefore, it is recommended to the shaman not allow himself to be influenced by elements of his own complex. Lastly, a neglect of the spirits may put the shaman under the condition of an accidental growth of one of the psychic elements, which may result in the mastering of the shaman by this spirit, or in disabling him for acting as a safety valve of the group. In all these situations we have a mere symbolization of the complex conditions of psychic life of the shaman and a perfectly adapted system of regulating mechanism created after long observations and experiments. In reality these symbols are based upon hypotheses, but their functioning does not suffer from it. The phenomenon of psychic life is not understood in the same form as the modern science would understand it, but it is regulated, and its components are perhaps better analysed (in spirit-symbols) than it is done by psychologists who operate with such conceptions as «instincts» and «complexes».
Let us take another example. In order to neutralize a harmful activity of the spirits, the shaman may make spirits fight among themselves, but this is not a recommendable method. In substituting the elements of the shaman's psychomental complex for the spirits, we obtain a picture of a badly balanced psychic condition, in which some elements may attain a dominant character, overshadowing other elements. On the whole the shaman would be psychomentally altogether disabled. It is stated that the shaman may even perish, which, in our terminology, would be styled as a maniacal state. According to the Tungus, the shaman must not allow himself to be tempted by experiments with spirits, and the other shaman may come to his assistance. Perhaps herein the Tungus are more right than the European psychiatrists who often underestimate purely psychic and mental conditions in the regulation of psychomental troubles. In reality, hysteria can be easily regulated. If some better methods are found there will perhaps be no need to look for a justification of failures, having recourse to hypothetical pathologo-anatomical and physiologico-chemical explanations of some psychomental conditions now supposed to be incurable. The internal conflict of the shaman with himself puts him out of practical use as a safety valve. The Tungus came to those conclusions after covering with their keen ability of observation a great number of facts. The Tungus hypothesis is wrong, from our point of view, but as a practical solution of the problem it works perfectly well.
Forms of repression of spirits («bad treatment»), for their wickedness, greediness etc. are mere symbolizations. They may much better be understood in their functioning than as pictures of «idols» and quotations of prayers or addresses.
Such are the «relations» between the spirits and the shamans. I shall now give some instances of practical managing of spirits with the help of sacrifices, but first I want to emphasize that the forms in which sacrifices are made are of no great importance for understanding the actual nature of shamanism. They are interesting as any other ethnographical phenomena which offer some material for ethnology.
SACRIFICES BY THE SHAMAN TO HIS SPIRITS are not
offered among all groups with the same regularity. Among the Reindeer Tungus of
Transbaikalia a sacrifice is offered, when the shaman feels that it should be
done. So it may be delayed for several years. Among the Tungus of Manchuria the
sacrifice, consisting of various animals (the more, the better), is given
periodically, no less than once every three years. In main lines it is the same
as the first shamanizing of the shaman who calls in, one after another, all
spirits mastered by him, to show that he possesses them. The shaman addresses
to the spirits his request that they remain near him. After the performance the
shaman climbs up to the turu, where he remains for a while. The shaman
assisting the performer carefully checks up all moments of the performance and
a competent person makes special marks on the wooden device. During a shaman's
life he may offer up to six periodical sacrifices. There is no difference
between the sacrifice carried out by the clan shaman (mokun'i) and a
free shaman [672]. Among the Manchus this operation is greatly complicated as
compared with the Tungus [673]. The ritual is carried out on the second day of
the first moon of the Chinese calendar. All families that were assisted by the
shaman [as has been shown they are called jarumbo (jarun bao)] bring
with them various sacrifices which are put in front of the placings for spirits
exhibited on the large stove-bed (in front of the entrance once I observed the
whole adult pig!). The tajar'i (chief assistant) sits on this stove-bed,
on the right of a small table with sacrifices and placings; the shaman is in
front of the sacrifice; elderly persons occupy places of honour other people
stand behind the shaman, while the women with children sit on the stove beds
behind the seniors. The performance begins at eight o clock in the evening. The
shaman puts on what is required by the ritual — conforming to which of the
spirits is to come. For calling in the spirits the shaman goes out into the
yard and returns. The assistant must recognize which spirit is introduced. If he
falls the shaman once more returns to the yard and comes back again [674]. The
spirits are recognized by the ritual including words. After four or five
spirits are introduced the shaman is asked to introduce some more. If he is
tired being, old or psychically weak he may decline it; but if he is a strong
man he will shamanize throughout the whole night. During the performance I once
observed besides tajar'i three more assistants who were helping with
drumming. Lastly, tajar'i recited by heart all spirits (cholo) while
the shaman was only confirming by his monotonous keku! (widely used in
shamanism as refrain and symbol of confirmation.)
662. A great number of shamanistic conceptions cannot be translated into Russian, even by myself, while L. Sternberg did not know the Goldi language, which I know to be the fact, and which is moreover clear from his questions.
663. Cf. J. P. Alkor (Koshkin) [L. J. Sternberg as Tungusologist, in At the memory of L J. Sternberg 1861-1927 published by the Academy of Sciences, Sketches on history of knowledge. Fasc. Vll, 1930] who has used the diary of L. Sternberg, relates that this investigator saw the first shaman only once and «pressed with questions (in Russian it is still stronger: don'imal- «vexed», «pressed»!!) during long hours, in the result of which he succeeded in clearing up the problem of the relations of the shaman with his spirits» (p. 141). Not longer were his meetings with another shaman who fooled him, refusing to sing into the phonograph, saying that «the ajami may kill me for doing it», which was absurd, since the spirit was mastered. L. Sternberg accepted this as bona fide (L. Sternberg, op. cit. p. 479). Probably the greater part of the time this shaman was drunk, for L. Sternberg says «he was always drunk» (ibid.). Indeed, a inquiry lasting several hours and carried out by an official (such was in the eyes of the shaman L Sternberg's position) is not a sure method, even if the investigator does not suggest the answers in his questions. Still less reliable is this method of getting information from a drunken man who in such a state may tell anything which is required of him. Indeed, it was «required», for L. Sternberg by that time already possessed his discovery of sexual motives conceived after the inquiry of the first shaman.
664. However, apparently, after an insistent inquiry, L. Sternberg obtained some information from an old Yakut woman who decades ago had left her own people, had married a Russian political exile and investigator of the Yakuts, and had been living over twenty years in St. Petersburg with her husband. This woman was used as a source of information about things she had known in her youth. She was, doubtless, influenced by the eminent colleagues of her husband and might unintentionally have given a description, as she had been desired to give. First of all, it is necessary to examine carefully the question whether, in the Yakut complex, the spirits are «mastered» by the shaman, or the latter is «mastered» by them. Secondly, the kinds of spirits must be distinguished and they must not be mixed up. The spirits abassy, which can be both male and female ones, are by no means shamanistic spirits, while manarik means only «psychically abnormal» people (as well as their souls). The only spirits of shamanistic complex are the amagat, which can be of different sexes or of no sex, and which may even belong to distinct ethnical groups (cf. E Pekarskij, Dictionary, «amagat», e.g. Tungus) it is really impossible to rely upon the statement of a old, half «disethnizised» woman. Still more it is dangerous to rely upon a statement of a Buriat informer who, for winning the sympathy of his eminent protectors, pretended to be a shaman, and, being a real «arriviste», knew perfectly well what his protectors expected from him — the confirmation of their theories. G. D. Sanzeev (cf. Darxaty. Ethnographical Report on a Journey in Mongolia in 1927, Ac. S. 1930 St. Petersburg; also Weltanschauung und Schamanismus der Alaren-Burjaten, in Anthropos, Vol. XIII, pp. 538-560) gives some facts showing that the sexual moment may and may not play a role in the fate of a candidate to shamanship. Sometimes candidates are also simply «elected» by the lamas (Darxats, p 457), who use the Shaman's art for themselves.
665. I want to point out that W G. Bogoras has found among the Chukchis that the shamans are called enenelit, which exactly means «those possessing a shaman spirit», which presumes no election. After quoting W. Jochelson as to the love relations with the spirits, L. Sternberg remarks, «unfortunately Jochelson does not give us any information, as to whether this spirit is identical with the shaman's beloved spirit of the female sex» (op. cit., p. 489), by which he leaves an impression in the reader's mind that there is such a «beloved spirit». The fact itself is not yet demonstrative of «election» on a sexual ground, but it only points to the liability of the shamans to sexual dreams.
666. The theory of a guardian spirit, proposed by R. F. Benedict (The concept of the Guardian Spirit in North America, Memoirs of the Amer. Anthr. Assoc, No 29) and generalized with respect to Siberian groups, naturally cannot be applied to the Tungus and Manchus.
667. I knew all the circumstances of this case, for the shaman approached me for making a picture of this spirit, and she had to produce all possible details, before I could apply my «iconographic art» in shamanism.
668. It should be noted that the female shaman had a female spirit. Her device with twenty-six marks of spirits' visits was supplied with carved female genitalia and breast.
669. It should be noted that no sexual relations are possible.
670. It should be noted that this is not a function of a «guardian» spirit. It would be quite fallacious if one should take this attitude in respect to the Tungus and, I should say, in respect to other «savages».
671. I need not quote a multitude of names of authors who have taken this attitude in reference to the Tungus and, I should say, in reference to other «savages».
672. I myself, had no occasion to observe a periodical sacrifice. The above remarks are the result of my numerous inquiries.
673. I have twice observed the ceremony.
674. Once a shaman assisted by a man of little experience, had to repeat this operation four times.