In 1894 swarms of locusts, for the first time in thirty years, came down on the Shiré Highlands and consumed all the crops in the native gardens—even attacking at last the white men’s coffee-trees. Fresh broods kept succeeding each other throughout the dry season, and as the time for the rains drew near, the villagers became anxious. What was the good of sowing their maize if the dzombe were there, ready to eat up the young shoots as soon as they appeared above ground? Great discussions went on among the elders in the bwalo as to the source of this visitation—if one could only conjecture its reason, it might be possible to find a remedy. Chesinka, an old head-man on Mlanje mountain, had a dream, one night in October, which, at any rate, suggested a solution. His old friend, Chipoka, dead some four years, appeared to him, and told him that it was he himself who had sent the locusts, as a hint to his people that they were not treating him properly; it was a long time since they had given him any beer, and he was very thirsty in the spirit-world. So Chesinka sent word to Chipoka’s son, who at once took steps for repairing the omission.

Chipoka had been ‘a person of importance in his day’; he was the principal chief on Mlanje in Livingstone’s time, and, when he died in 1890, ‘had, with the consent of all his sub-chiefs and subjects, transferred the sovereign rights of his country to the Queen, in order to pledge the British Government to the protection of the indigenous Nyanja people against Yao attacks.’[7] His son, of course, does not occupy anything like the same position; but the village, when I saw it, must have been in its old place—or very near it—on the bank of the Mloza, a clear stream coming down out of the heart of Mlanje, between the two peaks of Chinga and Manga. Chipoka’s grave, with some huge bamboos growing on it, was within a short distance of the huts.

I had heard that a ceremony was to take place for the purpose of propitiating the old chief’s spirit; and when I walked over, on the morning of October 29, I found a sort of subdued stir, the people very busy, but all looking extremely solemn. Young Chipoka—a man of about thirty—and some other men were seated under the eaves of a hut, while the women moved in and out of the huts with pots of beer, and other people were busied about a group of neat miniature huts, made of grass, about two feet high. The roofs of these huts, which had been finished separately, were not yet put on, and I could see that a couple of earthen jars were sunk in the ground inside each. These jars were now filled with beer, and then the roof was lifted on. Chipoka, draped in his blue calico, came forward very courteously to greet me, and explained that the houses were ‘for Mulungu.’

Now ‘Mulungu’ is the word which is generally translated ‘God,’ and it really does sometimes seem to denote a supreme Deity; but here it clearly meant Chipoka’s spirit. Mr. Duff Macdonald has made it clear that the ‘gods’ of the Yaos—or, at least, those most definitely thought of as such, are the spirits of the dead—of a man’s father or grandfather, or the chief of his village,—sometimes of a great chief, who ruled over a large extent of country, like Kangomba of Sochi. When such an one lived long ago, people are apt to forget who he was when alive, and to think of him as a spirit only. Such spirits are often associated with particular hills, as is the case with Sochi and Ndirande, and might easily be mistaken by inquirers for genuine nature-powers.

I have more than once seen these little spirit-huts in villages, though I never on any other occasion received so straightforward an explanation of them. Once, the children who were with me objected to my approaching the hut, saying there was a chirombo there. Chirombo is a comprehensive word which may mean an insect, objectionable or otherwise, a lion, or other beast of prey, a mythical monster or bogy, or, simply, any animal or plant not good to eat. They may have meant the uncanny Something which was believed to have its abode there, or they may have been trying to keep out unauthorised intrusion by the fiction of a palpable chirombo with claws and teeth. Whether or not they consciously think of the dead as little shadowy figures, a few inches high[8] (like the representation of the soul as it issues from the body, on some Greek vases), such was evidently the thought that suggested the erection of these miniature dwellings.

Duff Macdonald says, ‘The spirit of every deceased man and woman, with the solitary exception of wizards and witches, becomes an object of religious homage.’ Of course, no one can worship all, and the chief of a village worships his immediate predecessor as the representative of all the people who have lived in the village in past times, and the whole line of his ancestors. In presenting his offering, he will say, ‘“Oh, father, I do not know all your relatives, you know them all, invite them to feast with you.” The offering is not simply for himself, but for himself and all his relatives.’ Sometimes a man approaches his deceased relatives on his own behalf; but, as a rule, it is the chief who prays and sacrifices on behalf of the village. As all the people living there are usually related to the chief, the deceased chief is the one to whom they would naturally pray; but any immigrants from elsewhere would probably pray to their own ancestors on matters connected with their own private concerns, while joining on public occasions in the recognised worship of the village god.

Naturally it is difficult for an outsider to gain any exact information as to religious practices, and, still more, religious beliefs. The Rev. Duff Macdonald, whom we have just quoted, enjoyed almost unequalled advantages, as regards the Yaos—spending three years in their country at a time when it was still virtually untouched by European influences, and being able to acquire a sufficient knowledge of the language to obtain the people’s own account of things at first hand. Great patience and tact, it is needless to say, are necessary for inquiries of this sort; and, even if one knows the language, it is not, as a rule, much use asking direct questions—unless of natives with whom you are fairly well acquainted. Of things which the stranger can see for himself in passing through the villages, the most noticeable are the little spirit-houses already mentioned, where sacrifices are presented from time to time. Sometimes these offerings are seen under trees, either in the village, or away from it—in fact, Mr. Macdonald says that the huts are erected, if there is no tree handy, close to the dead man’s house. (The house itself, as we shall see, is usually either pulled down, or shut up and left to decay.) If the tree is quite outside the village, the site may have been shifted, as often happens; or perhaps the spirit may be one of the ‘old gods of the land.’ This is possibly the case with the tree in the illustration, which is on Ndirande mountain, a few miles from Blantyre, though I am not sure whether this particular tree is close to a village or not. Ndirande, like Sochi, has a spirit of its own; and I suppose this is the reason why the boy who was accompanying me in the ascent of Nambanga (an isolated peak or knob at the northern end of the mountain) showed a sudden reluctance to go on. I thought he was tired, and told him to rest, and I would go on alone; but this seemed equally objectionable, and he was evidently making up his mind to go with me, as the lesser evil, when I decided to avoid the risk of inhumanity by turning back. As I could by no manner of means induce him to explain, I suspected the spirit might have something to do with the matter.

In Mr. Macdonald’s time, the chiefs of the Blantyre and Zomba villages were all Yaos, and their canonised predecessors therefore belonged to the same tribe; but a certain amount of reverence was also paid to ‘the old gods of the land’—i.e. the spirits of dead Nyanja chiefs who haunted the principal mountains, and were specially appealed to for rain. We have already alluded to Kangomba of Sochi. The Rev. H. Rowley, when at Magomero, in 1861, saw Kangomba in the flesh; he was then ‘about forty years of age, had a frank, open countenance and a good head, and was altogether a very manly fellow.’ Apparently he did not live to be old. Mr. Macdonald says: ‘One tradition concerning him is this—When the Wayao were driving the Wanyasa[9] out of the country, Kangomba, a Wanyasa chief, saw that defence was hopeless, and entered a great cave on the mountain-side. Out of this cave he never returned; “he died unconquered in his own land.” The Wayao made the old tribe retire before them, but the chief, Kangomba, kept his place, and the new comers are glad to invoke his aid to this day. Their supplication for rain takes the form Ku Sochi, kwa Kangomba ula jijise, “Oh, Kangomba of Sochi, send us rain!” The Wayao chief, Kapeni, often asks some of the Wanyasa tribe that can trace connection with Kangomba to help him in these offerings and supplications.’

The offerings usually consist of native beer and maize flour (ufa), sometimes also calico, as seen in the illustration. It is torn into strips lest it should be appropriated by some needy and unscrupulous passer-by, or perhaps because each offerer only feels it incumbent on him to present a mere shred as a symbolic gift, since spirits, properly speaking, have no use for calico. Mr. Macdonald quotes the native reasoning on the subject. Spirits intimate their desire for various things—in dreams, or by means of the oracle, and, if their request be at all reasonable, it is granted. But, ‘if a spirit were to come, saying, “I want calico,” his friends would “just say that he was mad,” and would not give it. “Why should he want calico? What would he do with it? There was calico buried with him when he died, and he cannot need more again.”’ Food and tobacco, and even houses, are, it would appear, quite another matter.

Perhaps this opinion as to calico has been modified since the above was written; certainly I have seen at least one tree covered with strips of calico, and that within a mile or two of a mission station. That on the Ndirande tree is a special sacrifice for rain. Usually the stones at the foot of the tree support one or more pots of moa (native beer made of millet), and there is either a little basket of flour, or some is poured in a heap on the ground.

According to Mr. Macdonald, men would often sacrifice to their own particular ancestor, by putting down a little flour inside the hut, at the head of their sleeping-mat. Omens were drawn from the shape of this little heap of flour—whether it fell so as to form a neat cone, or otherwise. Beer, also, was made to furnish omens; it was poured out on the ground, and if it sank in, the god accepted the offering.

The same writer says that ‘when a deceased smoker wants tobacco, his worshippers put it on a plate and set fire to it.’ Matope, the chief of a village near Blantyre, died in 1893. In the following year, the Rev. J. A. Smith (now of Mlanje), happening to be at the village, saw, as he told me, the head-men take out the dead chief’s snuff-box, fill it with snuff, and place his stool in a certain spot, sprinkling snuff all round it. ‘This is done from time to time’ (I quote from my diary) ‘during the first year or two after a death—after this time the spirit ceases to haunt the place.’

In the Upper Shiré district, I was not very successful in gleaning information, but have a note that a girl told me the ‘Angoni’ a-pempera Mulungu (‘pray to God’) after the following fashion: ‘The people here sometimes sacrifice (kwisula) a goat; it is done by the head-man, and the people all stand round nda! nda! nda! (i.e. in a row or circle); they eat the meat afterwards.’ Here too, evidently, the head-man acts on behalf of the village; and though it is not clear whether Mulungu here means the spirit of a dead chief, or a Supreme Deity, this is, for the moment, immaterial. I do not know whether the word kwisula is Nyanja or not, as I never heard it on any other occasion, and have hitherto been unable to trace it.

We have already seen that Mulungu is a name applied to the spirits of the dead—the amadhlozi or amatongo of the Zulus (we shall come back to these presently), and as local deities seem to be in many instances (perhaps in all, if we could trace them) identical with deceased chiefs, it looks as if the religion of the Bantu consisted, in the main, of ancestor-worship. However, other ideas, though dim and vague, seem to attach to the word Mulungu. Originally, perhaps, it meant no more than ‘the great ancestor’—the Zulu Unkulunkulu. This name, literally ‘the Great, Great One,’ seems to have been used by the Zulus as if conveying the notion of a supreme being and creator; but some of them, on being questioned, stated that he was the first man, the common ancestor, at any rate, of themselves and those tribes whom they acknowledged as akin to them. No special worship, however, was offered to him, as he had lived so long ago that no family could now trace its descent to him; and worship is (as with the ancient Romans) a family, or at most a tribal matter. The word um-lungu, in Zulu, means ‘a European’; it is used in no other sense, but seems originally to have been bestowed under the idea that white men were supernatural beings of some sort.

One might feel inclined to doubt the above etymology, which is Bleek’s, since, in some languages, as in Nyanja, the word Mulungu and the adjective -kulu (‘great’) exist side by side. But against this we may set the possibility of the former being borrowed from other tribes. Mr. Rowley, writing of the time when the Yaos were only beginning to settle in the Shiré Highlands, says expressly that they used the name Mulungu where the Anyanja spoke of Mpambe. Speaking from my own observation, I should say that the former had quite displaced the latter throughout the Blantyre and Upper Shiré districts. Now in Yao, precisely, the word for ‘great’ is not -kulu, but -kulungwa.

However that may be, some Yaos, at any rate, think of Mulungu as ‘the great spirit of all men, a spirit formed by adding all the departed spirits together.’ This might almost seem too abstract a conception to be a genuine native view, but it was clearly stated to Mr. Macdonald, and is confirmed by Dr. Hetherwick, who has had many years’ experience of the Yaos. This writer also states the view (which Mr. Macdonald hesitated to accept), that the form of the word, or rather its plural (which shows it to belong, not to the first, or ‘personal’ class, but to the second, including things without a separate life of their own, such as parts of the body, trees, etc.), points to Mulungu not being regarded as a person. Dr. Hetherwick was once trying to convey to an old head-man the idea of the personality of God. The old man, as soon as he began to grasp what was meant, talked of Che Mulungu, ‘Mr. God!’

I have noted down some uses of the word I have come across, which I think could not possibly be set down to missionary influence. On two occasions, people told me that their dead friends or relatives had ‘gone to Mulungu’; on another, a mother said that ‘Mulungu had taken away’ the little sick girl I was inquiring after. On hearing thunder, at the beginning of the rainy season, another woman remarked, ‘Mulungu anena’—‘Mulungu is speaking.’ This is very suggestive of the theory on the subject of earthquakes held, according to the Rev. A. G. MacAlpine, by the Atonga of Lake Nyasa, viz., ‘that it was the voice of God calling to inquire if his people were all there. When the rumble of the earthquake was heard, they all shouted in answer, “Ye, ye,” and some went to the flour mortars and beat on them with the pounding sticks.’ Any person who failed thus to answer ‘Adsum’ would be sure (it was believed) to die before long. The name used, in this case, was not Mulungu, but Chiuta.

We have mentioned the name Mpambe as used by the Anyanja. Livingstone, on his first visit to the Shiré Highlands, notes, ‘They believe in the existence of a supreme being called Mpambe and also Morungo’ (Mulungu). Mr. Rowley gives an interesting account of supplications addressed to Mpambe for rain. The principal part was taken by a woman—the chief’s sister. She began by dropping ufa on the ground, slowly and carefully, till it formed a cone, and in doing this called out in a high-pitched voice, ‘Imva Mpambe! Adza mvula’ (‘Hear thou, O God, and send rain!’), and the assembled people responded, clapping their hands softly and intoning—they always intone their prayers—‘Imva Mpambe.’ The beer was then poured out as a libation, and the people, following the example of the woman, threw themselves on their backs and clapped their hands (a form of salutation to superiors), and, finally, danced round the chief where he sat on the ground. The ceremony concluded with a rain-charm; but as this is rather magic than religion (the previous proceedings, as being distinctly an appeal to unseen and superior powers, come under the latter head), it will be more convenient to return to it later on. In this very neighbourhood, I heard, one sultry afternoon in September 1894, weird, shrill cries, which I was told were ‘the people shouting for rain’ on Mpingwe—one of the mountains between Blantyre and Magomero. Distant peals of thunder had been heard before the crying began, and the rain came before morning.

It is worth noting that, here, Mpambe is thought of as sending rain, while, in some parts, as on Nyasa, the word means ‘thunder.’ The connection between rain and thunder is obvious, especially where, as is the case in these countries, the latter always heralds the breaking-up of the dry season. Sometimes the word is said to mean ‘lightning’—which comes to much the same thing as far as the idea is concerned.

This certainly looks like a personification of nature-powers, which seems more probable than the suggestion that mpambe only came to mean thunder or lightning because these were the work of the being to whom the name was originally applied.

It is worth noticing that in Yao the rainbow is called Mulungu, or ukunje wa Mulungu (‘the bow of Mulungu’) and an earthquake chilungu, which is the same word with another prefix.

Chiuta (which is treated by the Rev. D. C. Scott as synonymous with Mulungu and Mpambe) is perhaps derived from the Nyanja word uta, ‘a bow,’ and connected with the rainbow (called in this language uta wa Lezi). Lezi, or Leza, meaning ‘lightning’ in the Kotakota dialect, is another synonym.[10] I have never heard it used except in the above compound. Chiuta is the word used by the Atonga, and, according to the Rev. A. G. MacAlpine, ‘is very difficult to derive with certainty, but whatever its root may be, it now denotes one who inspires wonder and awe.’ If, however (as is quite possible), the name was borrowed by the Atonga from the Anyanja, this may be a secondary meaning imported into it.

We shall see, later on, that a distinction seems to be made between deaths ‘by the act of Mpambe,’ and from other causes.

Before leaving this part of the subject, we may note that, according to the Yao belief, Mulungu ‘arranges the spirits of the dead in rows or tiers,’ and some mysterious beings called ‘the people of Mulungu’ figure in their fragmentary legends of creation.

Besides the above, the Yaos seem to have had special deities of their own, connected with the country whence they came, and, therefore, probably, ancient chiefs of theirs, as ‘the old gods of the country’ were of the Anyanja. Such was Mtanga, supposed to dwell on Mount Mangoche, lying midway between Lake Chiuta and the Shiré, and the old home of the Mangoche tribe. It was believed that the voice of Mtanga could still be heard on Mangoche. Some said (twenty or thirty years ago) that ‘Mtanga was never a man,’ and the name is only ‘another word for Mulungu.’ However, both meanings would seem to have been lost sight of in more recent times, since, in Dr. Hetherwick’s Yao vocabulary we only find ‘Mtanga, a hobgoblin.’ This definition would also suit Chitowe (or Siluwi), who is enumerated along with Mtanga by Dr. Macdonald, but figures in fairy tales as a kind of monstrous being, with only one arm, one leg, one eye, etc., the rest of his body being made of wax. ‘He is associated with famine.... He is invoked by the women, on the day of initiating their fields ... when the new crop has begun to grow.’ Chitowe may become a child or a young woman. In this disguise he visits villages and tells whether the coming year will bring food or famine. He receives their hospitality, but throws the food over his shoulder without eating it. Chitowe is a child or subject of Mtanga, and some speak of several Chitowe who are messengers of Mtanga. The Nyanja bogy, Chiruwi (the word is translated in Scott’s Dictionary, ‘a mysterious thing’), is probably the same as Chitowe; he is constructed as above described, and, in addition, carries an axe. He is in the habit of meeting travellers and wrestling with them in lonely places; if the traveller falls, ‘he returns no more to his village—he dies.’ If, on the contrary, he throws Chiruwi, he says, ‘I will kill you, Chiruwi,’ and Chiruwi entreats for his life, promising to show the man ‘lots of medicines.’ Then the man lets him go, ‘and Chiruwi goes on before and says this tree is for such a disease, and that tree is for such a disease’—in short, gives him a lecture on materia medica, which proves exceedingly useful.

Mbona of Tyolo seems to have been one of these local deities. Mr. Rowley says that he was supposed to be the supreme ruler of the Anyanja, superior to the Rundo, or Paramount Chief, who consulted him in all special emergencies. This was done through the medium of Mbona’s wife—a woman chosen for the purpose, who lived in a solitary hut on Tyolo Mountain—or elsewhere, for ‘he was thought to be ubiquitous,’ and huts on other mountains were consecrated to his service. ‘He was spoken of as having a visible presence, but no one could say they had actually seen him or heard him.... If the Rundo wished for Bona’s advice, he or his deputies would proceed to the top of the mountain, with horn-blowing and shouting, to make the bride of Bona know of his approach. She then retires to the seclusion of her hut, hears without seeing those who come to her, seeks and finds Bona in sleep, receives from him, in this condition, that which he wishes her to declare, and when she awakes she declares to the expectant people the message Bona has given her to deliver.’ As Mbona’s wife was thus condemned to solitude for the rest of her days, the position was naturally not much coveted, and the Rundo usually had women kidnapped to fill it. Mr. Macdonald, some twenty years later, merely refers to this spirit in passing, as a ‘local deity.’ The word appears now to be used as a common noun, in the following senses: (1) ‘A wonder,’ (2) ‘something desirable’, (3) ‘one who looks after people or things’—as the overseer at the namwali ceremonies, which will be referred to later on, and (4) ‘a witness.’ It may be connected with the verb bona, meaning ‘to look at’ (in Nyanja, ona is ‘to see’); and possibly (3) is or was used as the name of the spirit. I cannot help wondering whether a story I once heard, of an old woman living on Morambala, who kept a number of spirits shut up in her house, has any reference to this tradition. I once, at Blantyre, questioned some Chikundas from the River about this old woman, and they said that they had heard of her, and that her name was Mbonda. I knew nothing about Mbona at the time, and perhaps misheard Mbonda instead of it; the sounds are not unlike, if both o‘s are pronounced open. The application of the spirit’s name to the woman is just the sort of confusion that might arise when a tradition is falling into oblivion. About the same time (in 1894) I heard of another old woman living in a cave on Malabvi—a mountain a few miles east of Blantyre. No European had been able to acquire the land in the immediate neighbourhood, as the people refused to sell it during her lifetime. It has since occurred to me that she might have been one of ‘Mbona’s wives.’

I do not think that, as a rule, the Bantu have much notion of anything that can be called a devil, or, indeed, of evil spirits as such; the spirits of the dead seem to be thought of as beneficent or hostile, according to their dispositions when alive, or the behaviour of the survivors. Dr. Hetherwick says, ‘While there is no trace of a devil in the Mang’anja faith, they have the ziwanda, who are ... the mizimu of men and women, but who work only ill. They are always feared. Their nature is that of the other mizimu, but they have only the wish to do harm.’ The Wankonde, however, according to Dr. Kerr Cross, ‘firmly believe in a spirit of evil’—Mbasi.

‘In one place Mbasi is a person, an old man, who exercises extraordinary power. He only speaks at night, and to the head-men of the tribe, and during the interview every other voice must be silent and every light extinguished. In Wundale the people believe in such a person as having the power to make lions, and being able to send them off as messengers of evil ... against whom he chooses. His house is surrounded with long grass, in which he keeps his lions as other men keep dogs.’

Coming back to the spirits of the dead, we find that the Yaos use the word lisoka to express that part of a man which survives when he dies; when it is an object of worship, it is called mulungu. These spirits, as we have already seen, are frequently prayed to, and may give evidence of their existence in three ways—by answering the prayers addressed to them, in dreams, and through the prophetess. There are also various means of divination (such as that of the flour already mentioned) and omens, which may be consulted and interpreted by professional diviners.

Every village has its ‘prayer-tree,’ under which the sacrifices are offered. It stands (usually) in the bwalo, the open space which Mr. Macdonald calls the ‘forum,’ and is, sometimes, at any rate, a wild fig-tree.[11] The Wankonde offer prayers—at least their priests (waputi) do—in the sacred groves where the dead are buried. The nearest approach to a temple among the Anyanja is a small hut called kachisi, which is sometimes built near the house of the village chief, if not actually under his eaves—sometimes in the bush, at a short distance from the village. ‘The chief utters the prayers in the house himself, alone, while the people answer by chanted accompaniments and clapping of hands at the door.’ The same sort of ritual was observed in the prayers for rain described by Mr. Rowley. The shower which fell on that occasion was, of course, accepted as an answer to the people’s supplications.

The natives say, ‘A man complains, and the spirits can hear him, but they can have no intercourse with man except in dreams, and in the silent care which they can exercise, having power to lead men, and to watch over them with favour, or when a man is going into danger to turn him back.’ If more explicit communication is needed, they inspire some person, and make him rave (bwebweta); his words are not directly intelligible, but some one is found to interpret them; ‘one man is laid hold of by the spirits that he may tell all people and they may hear.’ The person thus inspired may be a man or a woman—among the Yaos perhaps more frequently the latter.

The dead may manifest themselves in the shape of animals; but this does not happen so often as among the Zulus, who quite expect their deceased relatives to come back, like Cadmus and Harmonia, as ‘bright and aged snakes,’ and are very glad to see them when they do. The Yao theory seems to be that none of the departed will do this, unless they mean to be nasty. ‘If a dead man wants to frighten his wife, he may persist in coming as a serpent. The only remedy for this is to kill the serpent’—which no Zulu in his senses would dream of doing. However, the accidental killing of ‘a serpent belonging to a spirit’ seems to demand some sort of apology. ‘A great hunter generally takes the form of a lion or a leopard; and all witches seem to like the form of a hyena.’ But witches often turn into hyenas without dying first—which belongs to another part of the subject. The Makanga, in the angle between the Shiré and Zambezi, refrain from killing lions, believing that the spirits of dead chiefs enter into them.

There seems to be some difference of opinion with regard to the degree to which the spirits will make communications in dreams. An old Nyanja chief said to Livingstone, ‘Sometimes the dead do come back, and appear to us in dreams; but they never speak nor tell us where they have gone, nor how they fare.’ On the other hand, as we have seen, communications in dreams are expected as a regular thing. The Anyanja and Makololo (says Mr. Macdonald) were inclined to lay more stress on dreams than on the oracle of the ufa cone. They argued that, if you put the flour down carefully enough, it will always assume the proper pyramidal shape, and if you cover it with an upturned pot overnight (the usual practice), it will keep it—unless the rats overturn the pot. Perhaps this was due to the rationalising influence of the Makololo, who had been longer in contact with white men, and (like other natives in like case) were always burning to assert the superiority which this gave them.

If the dreams are not sufficiently explicit, we must fall back on the prophet or prophetess. Macdonald’s account tallies with the description of the woman set apart for the service of Mbona, except that he speaks of one living in a village with her family, who may arouse the neighbours with her shrieks in the dead of night. The people assemble to hear the message delivered by the spirits, and then return to bed; ‘or there may be a great meeting in the morning, when the prophetess appears, her head encircled with Indian hemp, and her arms cut as if for new tatus.’

But the Kubwebweta inspiration may come on any one, at any time, or in any place. Thus, one of a party of carriers on a journey suddenly finds himself ‘possessed,’ and his companions listen to his ravings with the greatest reverence. These utterances of possessed people always require some one to interpret them, and ‘an old woman or other skilled person’ is usually found. Macdonald says nothing of this, in the case of the prophetess, but if, as is probably the case, she is more or less of a professional,[12] she will have the necessary skill herself. The messages are not, as a rule, of a very recondite character—either the deceased chief wishes to help his people by warning them of war, or telling them where game may be found; or he feels himself neglected (like old Chipoka), and demands such and such offerings. Namzuruwa, an ancient chief of the district below the Murchison Cataracts, occasionally inspires people in this way.

The dead sometimes appear in visible form, as a native told the Rev. D. C. Scott: ‘People sometimes see those who have died and are dead walking outside in the gardens.’[13] I have never had the luck to hear a ghost story at first hand myself—the ‘night fears’ of the small boys whom I found objecting to go out after dark were connected, not with ghosts, but with wizards, of whom more presently.

There are haunted places in the Bush, where spirits are supposed to be heard, but not seen. M. Junod was told, over and over again, in the Delagoa Bay country, of people who had heard the drumming and singing of the spirits. These haunted spots were the burial-places of the chiefs, and no doubt this is so in other cases. The Anyanja have a curious account of ‘the spirits’ hill,’ where people who go, carrying pots on their heads, have them lifted off by the baboons, and hear a sound, ‘as of people answering.’ They also speak of the ‘spirit-drums’—the small ones sounding piye! piye! and the big ones, pi! pi! as though a dance were going on, and so far as one can gather, these spirits seem to be thought of as in sight like ordinary men and women—not the little εἴδωλα who dwell in the spirit-houses.

The notion of a connection between religion and morality comes comparatively late in human development; but we can perhaps see traces of it in the idea of the chirope. This means that, when a man has killed another man, he will either be ill, or be seized by a sort of murderous madness till he has performed some expiatory ceremony. The accounts I have before me are somewhat different, but are not, perhaps, inconsistent with one another. Among the Yaos, in Macdonald’s time, it seems to have been a condition that the victim should not be an enemy (towards whom no obligations were recognised), or even a person of the same tribe, whose kinsfolk could take up the feud and demand compensation. But, if a Yao killed his own slave (or, apparently, his child, his younger brother, or any one under his charge), he feared that he ‘would pine away, lose his eyesight, and die miserably, unless he went to the chief, paid him a certain fee, and said, “Give me a charm, for I have slain a man.”’ The Angoni, like the Zulus, apply the notion to killing a man in battle, and think that, unless they gash the bodies of the slain, so as to let out the air from the intestines, and prevent the corpse from swelling, they will be attacked by a mysterious disease which causes their own bodies to swell up. (This precise symptom is not given in the accounts before us, but is believed in by the Zulus, and probably by the Angoni.) The Angoni afterwards dance a war-dance ‘to throw off the chirope.’ The word appears to be connected with mlopa, ‘blood,’ used particularly of blood shed in killing—as of animals in hunting—and the idea is that the spirit of the slain enters the body of the slayer. This is even the case with animals; and hence it is the custom for the hunter to cut off a small piece of the meat as soon as he has shot any animal, throw it on the fire, and eat it, ‘because of the spirit of the beast that enters into one if one does not.’

The Angoni and various other tribes west of the Lake have a belief that there is a distinct relation between smallpox and morality; that, if the disease attacks a village where the moral tone is good, all the patients will recover; whereas, in a place given, as the native statement puts it, ‘to adultery and other sins,’ every one who sickens, young or old, will die. The locality, and various other circumstances, make it unlikely that this is an imported notion.

It is generally believed that the Eastern Bantu have no ‘idols’ properly so called; and their charms, to which we shall come back later, do not usually take the form of human figures. But the Tonga chiefs used to carry about with them little wooden images called angoza—representing men, women, or animals. Sometimes they were only sticks with a little head carved at one end. The Rev. A. G. MacAlpine, who seems puzzled what to make of them, does not state whether any are now in existence. ‘Long ago they used to be owned by chiefs only, and were lodged in the house of the head-wife.... They were not displayed except on special occasions. In the talking of important cases, they are said to have been brought out and planted in the ground at some little distance from the chief, and when he went on a journey they might be carried along with him, both of which uses would suggest their being an emblem of authority.... Often people came asking to see them, when they would be brought out covered up and not exposed till some gift had been made.’ We find that the Achewa have articles described as ‘fetiches’ and consisting ‘of a few short pieces of wood the size of one’s forefinger, bound together with a scrap of calico into the figure of a child’s doll. Inside the calico is concealed a tiny box made of the handle of a gourd-cup, ... [and] supposed to contain the spirit of some dead ancestor.’ Spirits wandering homeless in the bush are apt to annoy the living in various ways, till captured by a ‘doctor’ and confined in one of these receptacles.

The Yao children play with dolls bearing about as much resemblance to the human figure as a ninepin, but evidently intended to represent it. If games are survivals of religious ceremonies, they may originally have been teraphim, or fetiches of some sort. The ‘ugly images’ found by Livingstone near Lake Mweru, in ‘huts built for them,’ which were used in rain-making and cases of illness, seem to have been somewhat different from the angoza of the Atonga.[14]