The Kikuyu, Anyanja and Kamtchadal customs are evidently the same. The penalty for non-observance, though not expressed in our account of the Kikuyu custom, is, it will hardly be questioned, the same as in the other two; it is the death of the second husband. What I venture to suggest, in view of the other Bantu customs already laid before the reader, is that the risk actually run by marrying or cohabiting with the widow is that of death from the posthumous jealousy of the deceased. The fact that the deceased is still supposed to desire a continuance of conjugal relations renders it natural to think that, even where the extreme terrors that torture the widow in some places are not shown, the belief may linger in his jealousy of other men, who do what he perhaps is no longer conceived capable of. The suggestion derives support from the rites performed about the Lower Congo. There, it will be remembered, the terror of a dead husband is excessive. The elaborate purification of the widow, every incident of which points to the desire to rid her of his attentions, culminates in his brother stepping over her outstretched legs.
Now the act of stepping over another is everywhere regarded as one of great rudeness, if not insult. More than this, it is regarded as likely to communicate some mysterious injury or to take away some luck, good quality or advantage from the person who is thus treated. Hence it is often hotly resented. Without going beyond the Congo region, it is enough to note that in Loango to step over a child is to interfere with his development—in other words, to stop his growth; to step over an adult is to transfer to him every evil from which one may be suffering. It is not good even to reach, or to throw anything, across him; and if such a thing be done, the action must be repeated in the contrary direction, in order, it would seem, to reverse the spell.[224.1] But in the ceremony of purification an act in ordinary circumstances so injurious is ritually performed for some beneficial effect. This may be simply to take away some evil; it may be to dispossess finally and for ever the tenacious ghost, in case all the previous efforts have proved unavailing. But the difference of the performer’s sex means surely more than this. When a widow is purified, her husband’s brother performs the final rite; when it is a widower, his wife’s sister. Now it will be observed these persons belong to the precise class from which the next spouse is to be taken. When the husband dies the widow becomes the wife of one of his brothers; when the wife dies the husband demands another wife from her family.[224.2] The act of stepping over the widow or widower, therefore, seems to symbolize the taking of possession by one of the persons entitled to do so.
Let us turn to the Baganda, a Bantu people on the highest pitch of civilization to which any branch of the race had attained prior to the coming of the white man. There we find that jumping over a woman, or stepping over her legs, is regarded as “equivalent to, or instead of, having sexual connection with her.” “For a woman to sit with her legs straight in front of her, or apart, was looked upon as unbecoming; and for any man to step over her legs was equivalent to having intercourse with her. The mere fact of stepping over a wife, or over some of her clothing, was a method frequently followed to end a taboo which necessitated intercourse” (scil., to end it).[225.1] The act is performed on a variety of ceremonial occasions when coition would be inconvenient. To mention only one, the king had an officer of the court, a relative of his own, called the Kauzumu, whose duty it was to fulfil certain rites and taboos for him, and thus to save him from inconvenience. Among others, it was said that in former times it was his duty to take the women who were to become the king’s wives for one night to his bed. To the latest period of national independence, when one of the king’s wives died and her clan sent another in her place, the Kauzumu jumped over her before presenting her to the king.[225.2] By this act the taboo was removed, the danger, whatever it was, was diverted from the king to the Kauzumu. We may safely infer that the act of stepping over the legs of the surviving spouse in the Congo region is a symbolic coition. It is performed by one who belongs to the class of prospective spouses, though probably after the performance of such a ceremony not the one who would actually wed the survivor; for its object, if I am right, is to take on the shoulders of the performer the consequences which would otherwise light on the new consort.[226.1]
Though not by any means so conclusively as the practices just discussed, the delay between the ending of a first marriage by the death of one of the parties and the second marriage of the other party seems to indicate the same dread of the deceased. The mourning ceremonies must usually all be accomplished before the survivor, at all events if a woman, can be married again. With the completion of the mourning ceremonies it is a common belief, of which we have had more than one instance, that the deceased is finally despatched to the society of the dead; in any case, he is at rest, and is very often speedily forgotten. Among the Namib-Bushmen of South-West Africa, a mongrel people, the result of the intermingling of probably many broken tribes with an original stock of Hottentots, marriage is monogamous. That does not mean that death only can separate the married pair. If separation between two living spouses take place, either can marry again forthwith. But if the separation be caused by death, something like half a year (that is to say, either a rainy or a dry season) must elapse before re-marriage; for, we are told, the belief prevails that, for example, the woman whom a widower marries without waiting will soon herself die. And when the widower does venture on a new marriage, it must be to a sister of his deceased wife if there be one at liberty.[227.1] So in the Togo district of Atakpame, a woman who is left by her husband, even if only for a time, may be taken by another man; and this, it seems, under German rule renders it difficult to get men to work at the making of roads, for his absence will cost the labourer his wife. But when a man dies his widow must wait for three farm-years (a farm-year equals ten months) before she marries again. The delay seems not to be popular. In the western portion of the district, among the Akposso, the time has been shortened to two years, while in the eastern portion, where the people are less purely Akposso, the widow marries after eleven months. The modern Akposso widow holds it silly to wait longer: she has not murdered her first husband, and therefore she ought, she thinks, to be able to marry sooner. But she must wait these eleven months, else she would die. If, however, her husband has been hanged for murder or some other reason, she may marry at the end of two or three months. The ground alleged for this shortening of the period of delay is economic. There seems to be no rule requiring the widow to be taken by a surviving member of the husband’s family, though the children resulting from a second marriage belong to it. If she cannot find another husband, therefore, she must go back to her own family and she may become a burden upon it. On marrying she must work in her husband’s fields, for the benefit of himself and his family. But she is entitled to a small field for herself, out of which she may make her own profit; and there is now plenty of money in the country, and a general desire to earn as much of it and as soon as possible.[228.1] All this may be very true; but seeing that the widow who marries prematurely must die, she will not be likely to run so serious a risk for the chance of making a little money. If we are right in supposing that the death-penalty is one exacted by the deceased, the conjecture will not be deemed unreasonable that hanging is believed to inflict such injuries on the ghost that its interference is not to be dreaded. The mode of death and the mutilations (if any) inflicted on the corpse are widely held to affect the departed in the future life.
Jealousy is a passion by no means confined to the male sex. In the lower culture it is believed to continue to inspire women as well as men, even after death. If not so prominent a superstition as the belief in masculine jealousy, this is perhaps to be accounted for partly by the generally dominant position in the household and in society of the man, partly to the widespread habit of polygamy, and partly to the physiological rule that the decay of sexual impulse makes its appearance earlier in women than in men. We have already found examples of customs pointing to posthumous feminine jealousy among various races in both hemispheres. One or two other examples may be given.
Among the Bantu tribes of North-Eastern Rhodesia, when a married woman dies her husband sends to her father for another wife in her stead. A sister or some such relative of the deceased is the proper person to take her place. If there be none unmarried and therefore available, a married sister of the deceased must spend one or two nights with the widower “to take the death from off his body.”[228.2] This is so much in harmony with the Bantu practices previously considered, that we can only suppose the deceased would acquiesce in her sister’s succession to her rights, but would feel jealous and angry at the intrusion of another woman. The women of these tribes are high-spirited and jealous, those of the Awemba being proverbially fierce. Among most of the tribes it is imperative for the bridegroom to move into the bride’s parents’ village, where he will always be under the eye of his mother-in-law, and where, no doubt, in case of the wife’s death, a sister or other relation would be conveniently handy to step into her shoes. Indeed, among the Awemba the wife often relinquishes her position voluntarily to another member of the family. When she “has presented her husband with two or three children she considers that she has fulfilled her marriage obligations towards him. With his consent, which as a rule is not difficult to obtain, she hands over her niece as a substitute. The niece inherits her aunt’s position, and cares for her children, while the aunt retires to the peace of a single life, or very often finds a new partner.”[229.1] Thus it may be presumed that a dead wife, if she returned and found a relative of her own in her husband’s arms, would submit to a relationship between them such as she herself might in a year or two have voluntarily initiated, satisfied that her children, if any, would be looked after. How long that relationship would last (and divorce is quite an everyday occurrence) would be no concern of hers, and she would not trouble herself further in the matter. The next woman whom her husband married would therefore be rendered safe: the “death” would have been taken off his body.
In India the widows of the higher castes are not allowed to marry again. Among the aboriginal tribes widows who do so often marry “under cover of darkness, and with the aid of certain ceremonies commonly relied upon as a protection against spirits.” If in spite of these precautions the woman should be troubled by her departed husband, he is placated, as among the Kolīs of Ahmadnagar, by a tiny image of himself worn in a copper case round her neck, or set among the household gods, and by the expenditure of money in charity.[230.1] But the spirit of a dead wife is more troublesome. Her character for jealousy and malice is perhaps partly due to confusion with a woman dead in childbirth or pregnancy, or who has never had a child. A ghost of this kind is greatly dreaded.[230.2] The ghosts of ordinary married women are, however, not to be despised. They perform all sorts of unpleasant tricks on the survivors, not the least of which are the attacks they make on their successors in their husbands’ affections. Happily these attacks can be warded off by a judicious homage to the departed or a pretended identification of the deceased with her successor. Thus among the Gaddis, a Hindu sect in the Panjab, when the first wife dies the second wears a silver plate called saukan mora, or crown of the rival wife. This plate represents the deceased, and is propitiated to avert her hostility.[230.3] Indeed, the practice is not confined to one sect: it is common to several of the castes. The widower hangs a miniature portrait of his former wife, or even her name engraved on a silver or gold plate, about the new bride’s neck. The object, it has been suggested, is to humour the spirit of the first wife by identifying her with the second, thus proving the husband’s fidelity. In the Central Panjab, on the other hand, the bride is dressed as a milkmaid, or a flower-seller, and given a servile nickname, apparently to convince the spirit of the deceased that the girl being married is not a real wife, but a slave-girl. When the death of the second wife shows that the device has been unsuccessful, a mock-marriage of the bridegroom to a tree or a sheep, adorned like a bride, is resorted to before his marriage with a woman. “It is interesting to watch the bedecked sheep sitting on the khárás (reversed baskets) with a bridegroom and being led by him round the sacrificial fire, while the real bride sits by.” Here it is doubtless intended to fix the attention of the deceased upon the tree or the sheep, and so leave the real wife free from her jealousy. After the death of the third wife the evil influence of the first is deemed to be exhausted.[231.1] The Lets of Bengal, who are also Hindus in religion, permit their widows to marry again, though not by the rite for a virgin-bride. The second husband is usually a widower; and he places the iron bangle of his former wife on his new wife’s arm.[231.2] So in Baroda a widower who marries again has to present to his new wife a neck-ornament with marks to represent the feet of his first wife. She will wear this to prevent the ghost of the latter from troubling her.[231.3]
It need hardly be said that the rites and tales here discussed involve something more than the belief in the survival of death by a bare human personality. They could not have come into existence without the belief that what remained after the catastrophe was still in some degree a sentient and powerful being. It is difficult for mankind to acquiesce in the reality of death. The imagination refuses to harbour the thought of the cessation of conscious existence. The dead, although they appear to respond no longer to the physical and social stimuli hitherto effective, must still be. Vanished from our ordinary ken, they must be living somewhere, somehow; they cannot have been annihilated. If they still live, they must live under conditions analogous to those we know, and with affections, desires, appetites, aversions, similar to those they had in this life, though the objects may conceivably be changed. Hence every sort of activity known to man is ascribed to them. In dreams, in trances, and in the phenomena of possession they are observed to carry on these activities. But because they are observed to carry them on only under such mysterious conditions they repel the survivors and fill them with an undefinable horror and awe. Among the affections and appetites of mankind those connected with the sexual life are almost the strongest. It is natural therefore to imagine that the dead will still endeavour, by social and even by fleshly intercourse with the survivors, to gratify them. For if the belief in their activity in other directions be conceded, this cannot be denied; and perhaps the very horror that would be inspired by the ultimate stages of such a possibility serves to attract the thoughts, and therefore the belief, of peoples on the lower planes of civilization.
Ill weeds grow rank in such a soil. Not only are sexual appetites and affections almost the strongest; but, as is now well recognized, in the complex attitude of mankind to the invisible world they have no inconsiderable share. Marriage is a status hemmed about by many taboos and easily lending itself to mystical presentation. It is natural that the demise of one of the parties should cause a shock to the survivor corresponding to the change of condition involved, and should powerfully affect the sexual impulses. Even apart from marriage they are an object on which the thought of mankind, whether savage or civilized, broods with a persistent endeavour to solve the mystery surrounding them. Hints are seen in them for the solution of many a problem other than that of the propagation of the species. The ideas suggested by their contemplation are utilized alike for practical and economical purposes (for example, in agricultural magic), and for speculation on the transcendent themes of life and death and the constitution of the universe. The tales and superstitions I have brought together exhibit these impulses emphasizing the terror of the dead. Many of the phenomena of dreams, and the phenomena of hysteria and various forms of delusion are due to the sexual impulses. The sexual impulses in turn are in many, if not all, of the cases referred to above greatly aided by the conditions imposed on the patient. No less than luxury and over-feeding, fasting and abstinence from the ordinary and reasonable gratification of animal appetites, and from social intercourse, are the parents of nightmare and delusion. Both extremes lead through physical disturbances to disorders of the imagination and the reason. Hermits and celibates have probably suffered even more than rakes and gluttons and drunkards. The annals of monasticism with tiresome and pitiful iteration record the disastrous effects of asceticism upon body and mind. Among savage and barbarous peoples these effects are as a rule less noxious, because the conditions inducing and accelerating them are less continuous. But it is precisely at the time when sexual relations are put out of normal gear, and the sexual impulses are left without their normal gratification, that exclusion from society and abstinence, sometimes from sleep, always from food usual in quality and amount, are imposed. The imagination, already stimulated by the shock of the death (rarely attributed to natural causes), the possibility of an accusation of witchcraft hanging over the widow’s head, the certainty in many cases (as among the tribes of British Columbia) of a long period of hardship and even torture, all combine with the exclusion and enforced abstinence to produce a state of unnatural excitement, amounting at times to terror and delusion—in any case liable powerfully to affect the dreams, and to produce hysteria.
It must be remembered that the entire community shares to the full the underlying beliefs, and participates in varying degrees in the fears of the bereaved spouse. We are familiar even in Europe with the effect upon a crowd—nay, upon an entire nation—of emotions felt in common. They are intensified. The atmosphere becomes electrical, and a spark suffices to produce a conflagration. When the object of the emotion is strange, imperfectly apprehended, mysterious, the effect is heightened; it becomes a species of insanity. Things are imagined which do not exist; eyewitnesses attest what a dispassionate observer knows to be impossibilities. The most appalling of the tales and the most vivid of the rites we have been considering are thus easily accounted for. The marvel is that they are not more widely distributed over the world than our present information enables us to affirm.