That to the primitive mind they thus present themselves^ is shown by facts. Southey, writing of the Abipones, says that ''what became of the Lokal [spirit of the dead] they knew not, but they fear it, and believe that the echo was its voice.'' Goncemitig the Indians of Cnmana (Central America), Herrera tells us that they '' believed the soul to be immortal, that it did eat and drink in a plain where it resided, and that the echo was its answer to him that spoke or called.'' And, narrating his voyage down the Niger, Lander says that '' from time to time, as we came to a turn in the creek, the captain of the canoe halloed to the fetish, and where an echo was returned, half-a-glass of rum, and a piece of yam and fish, were thrown into the water. When asked why, he said — ' Did you not hear the fetish r" Here, as before, I must ask the reader to ignore these special interpretations, acceptance of which forestalls the argument. Attention is now drawn to this evidence simply as confirming the inference that, in the absence of physical explanation, an echo is conceived as the voice of some one who avoids being seen. So that once more we have duality implied — of an invisible as well as a visible state.
PBimnyx idias. 135 § 59. To a mind nnfnniislied witH any ideas Bave those of its own gatherings soironnding nature thus presents mnltitadinoDB cases of seemingly-arbitrary change — now slight and slow, now gradual and great, now sudden and extreme. In the sky and on the earth, things make their appearance and disappear; and there is nothing to show why they do so. Here on the surface and there deeply imbedded in the ground, are things that have been transmuted in substance -^chai^ed from flesh to stone, from wood to flint. Living bodies on all sides exemplify metamorphosis in ways murellous enough to the instructed, and to the primitive man quite incomprehensible. And this protean character which so many things around him exhibit, and which familiarize him with the notion that there are two or more inter-changeable states of existence, is again impressed on him by such phenomena as shadows, reflections, and echoes.
Did we not thoughtlessly accept as innate, the conceptions slowly elaborated during civilization and acquired insensibly during our early days, we should at once see that these ideas which the primitive man forms, are inevitably tormei. The laws of mental association necessitate these primitive notions of transmutation, of metamorphosis, of duality; and, until experiences have been systematized^ no limits or restraints are known. With the eyes of de* veloped knowledge we look at the snow as a particular form of crystallized water, and at hail as drops of rain which congealed as they fell. When these become fluid we say they have thawed — thinking of the change as a physical effect of heat; and, similarly, when the hoar frost, fringing the sprays turns into hanging drops, or when the surface of the pool solidifies and again liquifies. But looked at with the eyes of absolute ignorance, these are transmutations of substance— passings from one kind of existence into another kind of existence. And in like ways are necessarily conceived all the changes above enumerated.
Let us now ask what happens in the primitive mind when 136 THE DATA OF SOCIOLOOY.
there has been accamulated this heterogeneous assemblage of crude ideas^ having, amid their differences, certain resemblances. In conformity with the law of eyolution, every aggregate tends to integrate, and to differentiate while it integrates. The aggregate of primitive ideas must do this. After what manner will it do it? At the outset, these multitudinous vague notions form a loose mass with* out order. They slowly segregate, like cohering with like, and BO forming inde&iitely-marked groups. When these groups begin to form a consolidated whole, constituting a general conception of the way in which things at large go on, they must do it in the same way: such coherence of the groups as arises, must be due to some likeness among the members of all the groups. We have seen that there is such a likeness — ^this common trait of duality joined with this aptitude for passing from one mode of existence to another. Integration must commence by the recognition of some conspicuous typical case. It is a truth perpetually illustrated, that accumulated facts lying in dis« order, begin to assume some order if an hypothesis is thrown among them. When into a chaos of detached observations, is introduced an observation akin to them in which a causal relation is discernible, it forthwith commences assimilating to itself from this heap of observations, those which are congruous; and tends even to coerce into union those of which the congruity is not manifest. One may say that as the protoplasm forming an unfertilized germ, remains inert until the matter of a sperm-cell is joined with it, but begins to organize when this addition is made; so a loose aggregate of observations continues unsystematized in the absence of an hypothesis, but under the stimulus of an hypothesis undergoes changes bringing about a coherent systematic doctrine. What particular example, then, of this prevalent duality, plays the part of an organizing principle to the aggregate of primitive ideas? We must not look for an hypothesis properly so caUed: an hypothesis ?iinitnTB IDEAS. 137 is an implement of inquiry not to be framed by the primitire mind. We mnst look for some experience in wliich this duality ia forcibly thrast on the attention. As a oonsdonsly-held hypothesis is habitually based on some obtmsive instance of a relation^ which other instances are inspected to be like; so the particular primitiye notion which is to serve as an unconscious hypothesis^ setting up organization in this aggregate of primitive notions^ must be one conspicuously exemplifying their common trait.
First identifying this typical notion^ we shall afterwards have to enter on a surrey of the general conceptions which resolt. It will be needful to pursue various lines of inquiry and exposition not manifestly relevant to our subject; and it will also be needful to consider the meaning of much evidence furnished by men who have advanced beyond the savage state. But this discursive treatment is unavoidable. Until we can figure to ourselves with approximate truth the primitive system of thought, we cannot fully understand primitive conduct; and rightly to conceive the primitive system of thought, we must compare the systems found in many societies: helping ourselves by obsenring its developed forms, to verify our conclusions respecting its undeveloped form.* * The reader who is surprised to find in the suooeeding chapters so mooh •pwe dsvtited to the genesis of those ''superstitions," as we caU them, which e(nstitiite the primitiTe man's Theoiy of Things, will get a cine on taming to the first part of the Essay on " Huiners and Fashion," originally puhlished in 1S54 (iM SmauSf Ae^ VoL L). The conception, there hriefly indicated, of the vay in which social organization is affected by his beliefs, I hare been, since that da^ slowly deteloping; and the following chapters present it in a complete form. Beyond pnhiishing an article on ** The Origin of Animal* Worship " hi May, 1870^ I have done nothing toward setting forth these dereloped views — other sabjeets having had prior claims. In the meantime the important works of Mr. ''Tylor and Sir John Lubbock, have established, by abnndant evidence, views in some respects like them. It will be seen, however, that, while coinciding in several of their special condosions, I differ in lespeot to the order of genesis and mode of dependence of primitive superstitions.
CHAPTER DL THE IDEAS OF THE ANIMATE AND THE INANIMATE.
§ 60. At first sight, the difference between an animal and a plant seems greater than the difference between a plant and a lifeless object. Its frequent movements distinguish a quadruped or a bird from inert things; but a plant, inert in most respects, is not thus distinguished. Onlj to beings capable of making those comparisons between past and present by which growth is detected and the cycle of reproductive changes traced, can it become manifest that plants are allied with animals more than with the rest of things. The earliest classification, then, puts animals into one group and the rest of things into another.
Hence, in considering how there arises in consciousness the distinction between the living and the not-living, we may^ for a while, neglect the phenomena of plant-life and consider only those of animal-life.
Fully to understand the nature of this distinction as conceived by the primitive man, we must observe the development of it through lower forms of consciousness.
§ 61. When wandering some sunny day on the sea-shore, among masses of rock covered with '' acorn-shells, '^ and occasionally standing still to examine something, a feeble hiss may be heard. On investigation, it will be found that this sound proceeds from the acorn-shells. During the absence of the tide they commonly remain with their valves THE IBBAS Of THl AHIUATB AKD nTANIHATB. 139^ not quite shut; bat those on which ' a shadow is saddenly Gist begin to close^ and hy simultaneous closure of the grest numbers covered by the shadow, this faint sound is produced. Here the fact to be observed is that these cirrhipeds^ which are transformed crostaceans having aborted eyes imbedded in their bodies, and vision which suffices only to discriminate light from darkness, draw to Uie doors of their cells when there is a sudden obscuration, (^^dinarily^ something alive casts the shadow — ^there is an sdjacent source of danger. But as the shadow may be cast by a sharp-edged cloud, which obscures the sun with adeqoate suddenness, an adjacent living body is frequently not tile cause: the test is a very imperfect one. Still, we see tiiat deep down among creatures thus unintelligent, there is a vague general response to an indication of adjacent life: ^ indication being a change that implies a moving body.
Various inferior types whose lives are carried on by reflex actions only, display no very marked advance on this mode ^ discriminating the living from the not-living, as visually presented. Further along the shore, in the tide-pools, are Bhrnnps, which suddenly dart hither and thither when a ^rge body comes near; and when decaying sea-weed is dis-^bod, the sea-fleas jump, whatever may have caused the ^^sWbance. So in the neighbouring fields, the insects, not ^^stinguishing the shapes of moving objects or their kinds of motion, fly or leap when sudden great changes of visual Impression are made on them — each such change usually ^plying a living body near at hand. In these cases, as in the cases of caterpillars that roll themselves up when touched, the action is automatic. After the vivid nervous Btunulus there comes the strong motor discharge, resulting m flight or in a convulsive contraction of the muscles.
Speaking generally, we may say that in such cases the motion which implies life is confounded with the motion ^bich does not. The kind of mental act is like that occuri^Qg in ourselves when some large object suddenly passes close in firont. An involuntary start results^ before there is time to decide whether the object is alive or dead, a sonrce of danger or not. Here the primary suggestion with us, as with these lower creatures, is that the motion implies life; but whereas with us conscious obseryation instantly disproyes or verifies this suggestion, with them it does not.
§ 62. What is the first specialization of this primary consciousness? How do superior creatures begin to qualify this association between motion and life, in such way as to exclude from the class of living things a number which move but are not living 7 Where intelligence rises beyond the merely automatic, the motion implying life begins to be distinguished from other motion by its spontaneity. Without being struck or pushed by anything external, bodies which are alive suddenly change from rest to movement, or from movement to rest. Books watching doubtfully as you pass in the distance, rise into the air if you stop; or, not doing this, do it when you walk on, or even when, without changing place, you move your arms.
That the spontaneity of the motion serves as a test, is clearly shoTtm by the beliaviour of tame animab, and even of wild animals, in presence of a railway train. In the early days of railways they showed great alarm; but after a time, familiarized with the roar and the swift motion of this something which, appearing in the distance rushed by and disappeared in the distance, they became regardless of it: the cattle now continue to graze, and even the partridges on the embankment-slope scarcely raise their heads.
Converse evidence is yielded by the behaviour of a dog mentioned by Mr. Darwin. Like others of his kind, and like superior animals generally, he was regardless of the swaying flowers and the leaves occasionally rustled by the summer breeze. But there happened to be on the lawn an opened parasol. From time to time the breeze stirred this; and when it did so, the dog growled fiercely and ^ THB IDEAS OF TH£ AITDCATE AND tKANIMATE. 14l barked. Conscioas^ as his experiences had long ago made him^ that the familiar agency which he felt raising his own hair, sufficed also to move the leaves about, and that consequently their motion was not self -produced, he had not observed so large a thing as a parasol thus moved. Hence arose the idea of some living power — an intruder.
Again, appearances which at first vividly suggest life, are presently classed as not alive if spontaneity is absent. The behaviour of a dog before a looking-glass proves this. At first conceiving the reflected image to be another dog, he is excited; and if the back of the looking-glass is ao* cessible, makes attempts to reach the supposed stranger. When, however, the glass is so placed, say in a chiffonier, as to show him the image very frequently, he becomes indifferent to it. For what reason? The appearance does not spontaneously move. While he is still, it remains still; and any motion in it follows motion in himself.
§ 63* Tet a further test used by intelligent animals to discriminate the living from the not-living, is the adaptor iUm of motion to ends. Amusing herself with a mouse she bas caught, the cat, if it remains long stationary, touches it with her paw to make it run. Obviously the thought is that a living thing disturbed will try to escape, and so bring a renewal of the chase. Not only is it expected that there will be motion which is spontaneous, in the sense that it is self -produced; but it is expected that this motion will be away from danger. Habitually it is observable of animals that when failing to decide by the odour whether something smelt at is a living creature or not, there is an anticipation that disturbance will cause it to run away if it is alive. And even the behaviour of some gregarious birds when one of their number has been shot, shows that the absence of response to the cries and movements of the flock, leads to the impression that their companion is no longer one of that class of objects known as animated.
142 THB DATA OF SOCIOLOOT.
§ 64. Thos, ascending from low to high types of creatures^ we see that the power of distingaishing the animate from the inanimate increases. At first extremely yagae, the discriminations become gradually more definite; at first very general^ they become more special; and the consequent classings become less frequently erroneous. First motion, then spontaneous motion, then adapted spontaneous motion, are the successiye tests used as intelligence progresses.
Doubtless other traits aid. Sniffing the air, a deer perceives by something in it' the proximity of an enemy; and a camivore often follows prey by the scent it has left. But odours, though concomitants of life in certain adjacent objects, are not used as tests of life; for when found, an object from which an odour proceeds is not regarded as living if it exhibits none of the expected motions. Sounds, too, serve as indications; but these, when caused by ani* mals, are the results of spontaneous motions, and are taken to imply life only because they accompany other spontaneous motions.
It should be added that the ability thus to class with approximate correctness the animate and the inanimate, is inevitably developed in the course of evolution; since it is an essential means to self-preservation. Under penalties of death by starvation or destruction, there has been a constant cultivation of the power to discriminate the two, and a consequent increase of it* § 66. Shall we say that the primitive man is less intelligent than the lower mammals, less intelligent than birds and reptiles, less intelligent even than insects 7 Unless we say this, we must say tiiat the primitive man disting^shes the living from the not-living; and if we credit him with intelligence higher than that of brutes, we must infer that he distinguishes the living from the not-living better than brutes do. The tests which other creatures use, and which the superior among them rightly use in nearly all cases, he THE ID1SAS &t THE ANIMATE AND INANIXATB. 143 also most use: the only difference being tliat occasional errors of classing into which the most developed among other creatores fall^ he ayoids.
It is true that the uncivilized man as we now find him, oonuaonlj errs in his classification when shown certain products of civilized art, having traits of structure or behaviour like those of living things« By the Esquimaux, Boss's vessels were thought alive — ^moving as they did without oars; and Thomson says of the New Zealandersy that " when Cook's ship hove in sight, the people took her for a whale with wings/' Andersson tells us that by the Bushmen, a waggon was supposed to be animated, and to want grass: its complexity, its symmetry, and its moving wheels, being irreconcilable with their experiences of inanimate things. " It is alive," said an Arawak to Brett, on seeing a pocket-compass. That a watch is taken by savages for a living creature, is a fact frequently noted. And we have, again, the story told by one of the Arctic voyagers of the Esquimaux, who, ascribing life to a musical box and a barrel-organ, regarded the one as the child of the other. But automatic instruments emitting various sounds, are in that respect strikingly like many animate bodies.' The motions of a watch, apparently uncaused by anything external, seem spontaneous; and hence the ascription of life is quite natural. We must exclude these mistakes made in classing things which advanced arts have made to simulate living things; since such things mislead the primitive man in ways unlike those in which he can be misled, by the natural objects around him. Limiting ourselves to his conceptions of these natural objects, we cannot but conclude that his classification of thp into animate and inanimate, is substantially correct.
Concluding this, we are obliged to diverge at the outset from certain interpretations currently given of his t/ superstitions. The assumption, tacit or avowed, that the pximitive man tends to ascribe life to things which are. not 144 THIS DATA OV SOCIOLOQT.
living, is clearly an nntenable assnmption. Oonscioosness of the difference between the two, g^wing ever more dejSnite as intelligence evolyes, mnst be in him more definite V than in all lower creatures. To suppose that without cause he begins to confound them, is to suppose the process of evolution is inyerted.
§ 66. It is, indeed, said that undeyeloped human intelligence daily shows a tendency to confound them. Certain &cts are referred to as implying that children fail in the discrimination. Were not this evidence vitiated by the suggestions of adults, it would have weight. But on remem* bering that when trying to pacify a child that has hurt itself against some inanimate object, a mother or nurse will affect to take the child's part against this object, perhaps saying, " Naughty chair to hurt baby — ^beat it 1'' we shall suspect that the notion does not originate with the child but is taught to it. The habitual 'behaviour of children to surrounding things implies no such confusion. Unless an inanimate object BO far resembles an animate one as to suggest the idea that it may be a motionless living creature which will presently move, a child shows no fear of it. True, if an inanimate thing moves without a perceived external force^ alarm results. Unlike as a thing may be to living things, yet if it displays this spontaneity characteristic of living things, the idea of life is aroused, and a scream may be caused. But otherwise, life is no more ascribed by a child than by a puppy or a kitten. Should it be said that, given as it is to dramatizing, an older child, endowing its playthings with personalities, speaks of them and fondles them as though they were living; the reply is that this is not belief but deliberate fiction. Though pretending that the things are alive, the child does not really think them so. Were its doll to bite, it would be no less astounded than an adult would be. To secure that pleasurable action of un« used faculties called play, many intelligent creatures thu|i THE IDEAS OF THS ANDtATS AND INANIMATE. 145 dramatise; and^ lacking the required living objects^ will accept as representing them, non-liying objects— especially if these can be made to simulate life. But the dog pursuing a stick does not think it alive. If he gnaws it after catching it, he does but carry out his dramatized chase: did he think the stick alive, he would bite it as eagerly before it was thrown as after. It is further alleged that even the grown man sometimes betrays a lurking tendency to think of inanimate objects as animate. Made angry by resistance to his efforts, he may in a fit of rage sweeur at some senseless thing, or dash it on the ground, or kick it. But the obvious interpretation is that anger, like every strong emotion, tends to discharge itself in violent muscular actions, which must take some direction or other; that when, as mostly happens, the cause of the anger is a living object, the xnuscalar actions have been directed towards the injury of this object; and that the established association directs the muscular discharges in the same way when the object is not Uving, if there is nothing to determine them in any other way. But the man who thus vents his fury cannot be said to think the thing is alive, though this mode of dis« charging his irritation makes him seem to think so.
None of these facts, then, imply any real confusion between the animate and the inanimate. The power to distinguish between the two, which is one of the first powers vaguely shown even by creatures devoid of special senses, which goes on increasing as intelligence evolves, and which becomes complete in the civilized man, must be regarded as approaching completeness in the uncivilized man. It ca^nnot be admitted that he confuses ideas which, through all lower forms of mind, have been growing clear.
§ 6*7. '* How, then, are we to explain his superstitions?'' it will be asked. " That these habitually imply the ascrip* tion of life to things not alive, is undeniable. If the primitive man has no proclivity to this confusion, how is it L 14G THB DATA OP SOCIOLOGT.
possible to oxplain the extreme prevalence^ if not the nniyersality, of beliefs which give personalities^ and tacitly ascribe animation, to multitudes of inanimate things? " The reply is, that these cannot be primary beliefs, but must be secondary belie& into which the primitive man is betrayed during his early attempts to understand the surrounding world. The incipiently-speculative stage must come after a stage in which there is no speculation — ^a stage in which there yet exists no sufficient language for carrying on speculation. During this stage, the primitive man no more tends to confound animate with inanimate than inferior creatures do. If in his first efforts -at interpretation, he forms conceptions inconsistent with this preestablished distinction between animate and inanimate, it must be that some striking experience misleads him — introduces a germ of error which develops into an erroneous set of interpretations.
What is the germinal error? We may fitly seek for it amid those experiences which mask the distinction between animate and inanimate. There are continually-recurring states in which living things simulate things not alive; and in certain attendant phenomena we shall find the seed of that system of superstitions which the primitive man forms.
CHAPTER X.
THK IDEAS OF SLEEP ANQ DSEAMS.
§ 68. A CONCIPTION whicli is made so &miliar to ns aoring education that we mistake it for an original and necessary one^ is the conception of Mind, as an internal existence distinct from body. The hypothesis of a sentient, thinking entity, dweUing within a corporeal framework, is now BO deeply woven into our beliefs and into our language, that we can scarcely imagine it to be one which the primi* tire man did not entertain, and could not entertain.
Yet we have but to ask what is giyen in experience to the untaught human being, to see that there is nothing to tell him of any such existence. From moment to moment he sees things around, touches them, handles them, moyes them hither and thither. He knows nothing of sensations and ideas — ^has no words for them. Still less has he any such highly-abstract word or conception as consciousness.
Be does not think about thought: neither his faculties nor his language suffice for this. During early stages he merely thinks without observing that he thinks; and therefore never asks how he thinks, and what it is which thinks. His senses make him conversant only with things externally existing, and with his own body; and he transcends his senses only far enough to draw concrete inferences respecting the actions of these things. An invisible, intangible entity, such as Mind is inferred to be, is a High abstraction unthinkable by him, and inexpressible by his vocabulary.
148 TilE DATA OF SOCIOLOGY.
This, which is obvious a priori, is verified a posteriori. The savage cannot speak of internal intuition except in terms of external intuition. We ourselves, indeed, when saying that we see something that has been clearly explained, or grasp an argument palpably true, still express mental acts by words originally used to express bodily acts. And this use of words implying vision and touch, which with us is metaphorical, is, with the savage, not distinguished from literal. He symbolizes his mind by his eye. (See Principles of Psychology, § 404.)
But until there is a conception of Mind as an internal principle of activity, there can be no such conception of dreams as that which we have. To interpret the sights and sayings and doings we are conscious of during sleep, as activities of the thinking entity which go on while the senses are closed, is impossible until the existence of this thinking entity is recognized. Hence arises the inquiry — What explanation is given of dreams before the conception of Mind exists.
§ 69. Hunger and repletion, both very common with the primitive man, excite dreams of great vividness. Now, after a bootless chase and a long fast, he lies exhausted; and, while slumbering, goes through a successful huntkills, skins, and cooki his prey, and suddenly wakes when about to taste the first morsel. To suppose him saying to himself — '' It was all a dream,'^ is to suppose.him already in possession of that hypothesis which we see he cannot have. He takes the facts as they occur. With perfect distinctness he recalls the things he saw and the actions he performed; and he accepts undoubtingly the evidence of memory. True, he all at once finds himself lying still. He does not understand how the change took place; but, as we have lately seen, the surrounding world familiarizes h^m with unaccountable appearances and disappearances, and why shoxdd not this be one f If, at another time, lying gorged THB IDEAS OP 3LKSP AND DREAMS. 149 irritli food^ the disturbance of His circulation produces nightnrnro — if, trying to escape and being unable, he fancies himself in the clutches of a bear, and wakes with a shriek; why should he conclude that the shriek was not caused by an actual danger? Though his squaw is there to tell him that she saw no bear, yet she heard his shriek; and like him has not the remotest notion that a mere subjective state can produce such an effect — has, indeed, no terms in which to frame such a notion.
This interpretation of a dream as an actual experience, is confirmed by narration of it in imperfect langfuage. We are apt to forget that discriminations easy to us, are impossible to those who have but few words, all concrete in their meanings, and only rude prepositional forms in which to combine these words. When we read that in the language of so advanced a people as the ancient Peruvians, the word hucica meant " idol, temple, sacred place, tomb, figures of men, animals, etc., hill,^' we may judge how extremely indefinite must be the statements which the vocabularies of the rudest men enable them to make; and when we read of an existing South American tribe, that the proposition — " I am an Abipone,'* is expressible only in the vague way — '*I, Abipone '/* we cannot but infer that by such undeveloped grammatical structures, only the simplest thoughts can be rightly conveyed. When, further, we learn that among the lowest men the inadequate words indefinitely combined are also imperfectly pronounced, as, for instance, among the Akka, whose speech struck Schweinfurth by its inarticulateness, we recognize a third cause of confusion. And thus prepared, we need feel no surprise on being told that the Zuni Indians require "much facial contortion and bodily gesticulation to make their sentences perfectly intelligible /^ that the language of the Bushmen needs so many signs to eke out its meaning, that " they are unintelligible in the dark;'' and that the Arapahos '' can hardly converse with one another in the dark." If, now, remembering all 150 THE DATA OF BOCIOLOOT.
tUs^ we ask what moat happen when a dream is narrated by a savage^ we shall see that eyen supposing he snspects some distinction between ideal actions and real actions^ he cannot express it. His language does not enable him to say — ** I dreamt that I saw/* instead of — '* I saw/* Hence each relates his dreams as though they were realities; and thus strengthens in erery other, the belief that his own dreams are realities.
What then is the resulting notion? The sleeper has been visibly at rest. On awaking he recalls various occurrences^ and repeats them to others. He thinks he has been elsewhere; witnesses say he has not; and their testimony is verified by suddenly finding himseU where he was when he went to sleep. The simple course is that of believing both that he has remained and that he has been away — ^that he has two individualities, one of which leaves the other and presently comes back. He, too, has a double existence, like many other things.
§ 70. From all quarters there come proofs that this is the conception actually formed of dreams by savages — a conception which continues to be held after considerable advances in civilization have been made. Here are a few of the testimonies.
Schoolcraft teUs us that the North American Indians in general, think '' there are duplicate souls, one of which remains with the body, while the other is free to depart on excursions during sleep; '' and, according to Orantz, the Greenlanders believe '' that the soul can forsake the body during the interval of sleep.'' Thomson says the New Zeal&nders believed ''that during sleep the mind left the body, and that dreams are the objects seen during its wanderings;'' and in Fiji, '' it is behoved that the spirit of a man who still lives will leave the body to trouble other people when asleep.'' Similarly in Borneo. It is, according to St. John, the conviction of the Dyaks that '' the soul / THl IDIAS or 8LXSP AND DBIAMS. 151