The tribal institution of the Aryan-speaking peoples includes a life outside the tribe. That was an outlaw's life, a kinless outcast, whom no tribesman would look upon or assist, whom every tribesman considered as an enemy until he had reduced him to the position of helot or slave, but for whom every tribe had a place in its organisation and a legal status in its constitution. But it was the legal status imposed by the master over the servant, and the kinless included not only the outcast from the tribe, but the conquered aboriginal who had never been within the tribe. It is important to notice this, for it to some extent measures the strength of the tribal organisation. It not only allowed for a special position for all tribesmen, but it allowed for that position to have a definite relationship to persons who were not tribesmen, and it is in the combined forces of tribesmen and non-tribesmen that the tribal organisation which swept over part of Asia and over all Europe obtains its greatest power. There are tribal systems outside the Semitic and the Indo-European, but these do not have the distinctive features that the tribal systems of these two great civilising peoples possess. Like the Semitic and Aryan tribal systems, savage tribes are fashioned for conquest, but, unlike them, they are not fashioned for settlement and resettlement, and perhaps again and again conquest and resettlement. They spent all their power, or most of their power, in their one great effort of conquest, and whether we turn to the American Indian tribes, to the African tribes, or to the Asiatic tribes we find the same facts of frequent dissipation of power after sudden and complete conquest of it. The tribal system which led to civilisation has a different history. It has, too, a different constitution in that to the strength of tribesmen was added the subordination—politically, industrially, and economically—of non-tribesmen. They were the people who, in the terms of the northern poem,

Unfortunately the institution of the tribe has never been properly studied by the great authorities in history, and students are left without guidance in this important matter. And yet in any attempt to get back to the earliest period of history in lands governed by an Aryan-speaking people we must proceed, can only proceed, on the basis of the tribe, and it is the failure to understand this which has made so much early history unsatisfactory and inconclusive and compels us to the conclusion that the master-hand is still needed to rewrite in terms of tribal history all that has been written in terms merely of political history.

If, however, history from the written records is thus at fault, so too is history from the traditional records. No systematic effort has been made to treat the traditional story or the traditional custom and belief as part of the tribal history of our race, and yet in the few cases where it has been so treated the results are obviously satisfactory. I can illustrate the value of this point of view by an example drawn from the period which witnessed the earliest struggles of our race. I think with Mr. Keary that in those German stories "which delight above all things in that portrait of the youngest son of the house—he is the youngest of three—who is left behind despised and neglected when his brothers go forth to seek their fortunes," we have traces of a veritable fact, of an historical condition where the elder sons actually went forth to conquest and to settlement and the youngest son remained in the original home as the hearth-child.[439] The position of hearth-child, surviving as it does in our law of Borough English, is of great significance, and that we can by the aid of tradition reach a state of society which gave birth to it is a point of the greatest importance, even if we could go no further. But there is a stage beyond it. The majority of these youngest-son stories relate to events not to be identified with any particular tribe or people, but which belong to all the tribes and peoples whose course of conquest and settlement took the common form. But if apart from these all-world stories there exist stories, or if there be but one story which has become identified with an episode, a person, or a place belonging to a particular people, we may claim it as part of the history of that particular people. It may be that the general story has become specialised in this one case, or it may be that an entirely new story has sprung out of the special case. But whichever be the origin of such a story attached to a particular people, it must tell us something of that people at a period when its history was being made rather than recorded. What it tells may be very little, may not lead up to anything very great or definite, so far as later history is concerned; but that for the period to which it belongs it relates to an episode worthy to have been kept in the memories of the descendants of the chief actors in the events is the point to bear in mind.

There is one such story which belongs to English history. One of the most famous of these youngest-son stories is that of Childe Rowland, and Mr. Jacobs, on examining its incidents and details, suggests that "our story may have a certain amount of historic basis and give a record which history fails to give of the very earliest conflict of races in these isles."[440] Mr. Jacobs gives good grounds for this conclusion, and shows up a picture of earliest English history which is certainly not contained elsewhere, and we are able by this means to pass from that large group of youngest-son stories, which have brought with them living testimony of an ancient institution of our race in its oldest home, to the narrower but more direct example which comes to us from events which happened just at the dawn of history in our own land. It is not necessary to emphasise the importance of this service to history at the instance of tradition, for it will be obvious to every student that many a struggle must have remained unrecorded and many a hero must have died unnamed in the events which belong to the period of tribal conquest and settlement. And to have still with us the far-off echo of these events is no slight encouragement to an inquiry which has for its object the reconstruction of the conditions under which such events took place.

This would be all the better understood if we could get a concrete case for illustration, and, fortunately, this is possible by turning to the evidence of India. "What we know of the manner in which the states of Upper India were founded," says Sir Alfred Lyall,

The evidence from India is fact, the evidence from England is tradition, and yet I do not think any student will deny that both fact and tradition are part and parcel of the same conditions of society, the same forces operating upon the same material. The conditions of society in both cases are tribal conditions, and the common factor having thus been discovered, it is possible to determine not only the inter-relationship between fact and tradition, but the means by which we may estimate the value of both.

We cannot, however, stop here. I carry on the same argument from the traditional legend to the traditional custom and belief, and affirm that it is only by their position as part of the tribal system that custom and belief in survival must be tested. If they have descended from early Celtic or Teutonic custom and belief, they have descended from tribal custom and belief, and somewhere in the stages of descent will be found the link which connects them definitely with the tribe. That not all custom and belief has so descended is due to the fact that much of it belongs to the pre-Celtic period, which was not tribal; some of it, no doubt, to comparatively modern times, when, as we have already seen, superstition had taken the place of thought, while some phases of early belief belong to conditions which transcend the division between pre-Aryan and Aryan folk. On this I will say something by way of explanation presently. In the meantime it is an extremely important task to classify survivals into tribal and non-tribal groups. Those which belong to Celtic or Teutonic origins must show their tribal origin, for they could not have come into existence apart from the tribe, and apart from the tribe they could not have survived after the break-up of the tribe consequent upon the development of national and political life. Custom and belief which do not fit into the ancient tribal system, therefore, cannot be recognised as ancient Celtic or ancient Teutonic custom and belief, and contrariwise when it is seen that they naturally fall into this system it may be argued that there we must search for their origin. Our Anglo-Saxon forefathers have left a curious testimony to this view of the question in their word "holy" or wholesome. What is wholesome is so for the whole group. The Anglo-Saxon idea of holiness implies as its chief element relation to the tribal life.[442]

The classification of survivals in folklore into tribal and non-tribal items is a lengthy and intricate process. Some years ago I made a start in a study of fire worship which I presented to the British Association,[443] and I hope shortly to be ready with a volume on Tribal Custom, which will embody a fuller study of fire worship and its accompanying beliefs, together with a complete study of all the remains of traditional custom, rite, and belief, which only as the detritus of the ancient tribal organisation receive adequate explanation of their presence in the midst of modern political and religious institutions. If I leave this part of my subject without further illustration in this present volume, I must add one important note upon the persistence of survivals of both kinless and kinship societies. I have shown that the tribal system of the advanced races included provision for non-tribesmen, provision which kept non-tribesmen outside the tribal bond, and at the same time kept them tied to the tribe by using them as the necessary dependent adjunct of the tribe, using them as bondmen and serfs in point of fact. This extremely important factor in the history of the tribal organisation, which has not been properly noticed by the few authorities who have investigated tribal institutions, receives additional importance when viewed from the standpoint of folklore, for it allows for the preservation of non-tribal cults side by side with tribal cults. Non-tribesmen preserved their custom, belief, and rite simply because they were not admitted to the custom, belief, and rite of the tribe, and this is the explanation of the existence, in survival, of folklore which goes back to pre-Celtic times. Some of this pre-Celtic folklore we have already had before us, and some of it I have studied in my Ethnology in Folklore. Later on I shall have something more to say on the subject. Here it is only necessary to emphasise the importance of having ascertained why it is that the Celtic conquerors of Britain and the earliest tribal conquerors of the Indo-European world generally permitted to live in their midst what in a sense was opposed to all that they believed, to all that they practised, to all that governed them in thought and action.

I think this is a strong position upon which to conduct folklore research. It includes the whole of the historical position; it takes due count of historical facts instead of ignoring them. It is based upon a scientific conception of the meaning of a survival of culture. A survival is that which has been left stranded amidst the development that is going on around. Its future life is not one of development but of decay. We are not dealing with the evolution of society, but with the decaying fragments of a social system which has passed away. We have to trace out its line of decay from the point where it almost vanishes as the mere superstition or practice of a peasant or an outcast, back to phases where it exists in more strenuous fashion, and finally back to its original position as part and parcel of a living social fabric. Moreover, the strength of our position is based upon a scientific conception of the development of the nation or people among whom survivals exist. It is not all parts of the nation which develop at the same rate, at the same time, and for the same period. There are social strata in every country, and it is the observance of these strata which has made it possible for the inquirer of to-day to use the evidence they afford for historical purposes.

The primitive myth does not include all that properly comes within the definition of myth. There must be included the myth formed to explain a rite or ceremony, which originating in most ancient times has been kept up at the instance of a particular religion or cult, but the meaning and intent of which has been forgotten amidst the progress of a later civilisation. Pausanias is the great storehouse of such myths as this, and Mr. Lang has, more than any other scholar, examined and explained the process which has gone on.

There is also included in this secondary class of myth, the myths upon which are founded the great systems of mythology. The Hindu mythology, in spite of all that has been done to place it on the pedestal of primitive original thought, is definitely relegated to the secondary position by its best exponents. The Vedic religion is tribal in form, and in the pre-mythological stage.[205] In the RÁmÁyanÁ and MahÁbhÁrata, on the contrary, "we trace unequivocal indications of a departure from the elemental worship of the Vedas, and the origin or elaboration of legends which form the great body of the mythological religion of the Hindus."[206] The pre-mythological and the mythological stages of Hindu religion, therefore, are both discoverable from the traditional literature which has descended from both ages, and this fact is important in the classification of the various phases of tradition. When once it is admitted that the beginnings of mythology are to be traced in one section of the people who are supposed to derive a common system of mythology from a common home, future research will hesitate to interpret, as Kuhn and Max MÜller and their school have done, the traditions of Celts, Teutons, and Scandinavians as the detritus of ancient mythologies instead of the beginnings of what, under favourable conditions, might have grown into mythologies. Mythological tradition is essentially a secondary not a primary stage. This fact is overlooked by many authorities, and I have noted some of the unfortunate results. It is not overlooked by those who study the principles of their subject as well as the details. Thus, as Robertson-Smith has so well explained, "mythology was no essential part of ancient religion, for it had no sacred sanction and no binding force on the worshippers.... Belief in a certain series of myths was neither obligatory as a part of true religion, nor was it supposed that by believing a man acquired religious merit and conciliated the favour of the gods. What was obligatory or meritorious was the exact performance of certain sacred acts prescribed by religious tradition. This being so, it follows that mythology ought not to take the prominent place that is too often assigned to it in the scientific study of ancient faiths."[207] This is exactly the position, and all that I have advanced for the purpose of aiming at a classification of the various kinds of tradition is in accord with this view.

All that I am anxious to prove, all that it is possible to prove, from these considerations of the position occupied by myth, is that myths constitute a part of the serious life of the people. They belong to the men and women, perhaps some of them to the men only and others to the women only, but essentially to the life of the people.

I do not think that even Mr. Hartland in his special study of the subject has quite understood this. He begins at a later period in the history of tradition, the period of story-telling proper, when myths have become folk-tales,[208] and he treats this period as the earliest instead of the secondary stage of myth. In this stage something has happened to push myth back from the centre of the people's life to a lesser position—a new religious influence, a new civilisation, a new home, any one of the many influences, or any combination of influences, which have affected peoples and sent them along the paths of evolution and progress.

It is in this way that we come upon the folk-tale. The folk-tale is secondary to the myth. It is the primitive myth dislodged from its primitive place. It has become a part of the life of the people, independently of its primary form and object and in a different sense. The mythic or historic fact has been obscured, or has been displaced from the life of the people. But the myth lives on through the affections of the people for the traditions of their older life. They love to tell the story which their ancestors revered as myth even though it has lost its oldest and most impressive significance. The artistic setting of it, born of the years through which it has lived, fashioned by the minds which have handed it down and embellished it through the generations, has helped its life. It has become the fairy tale or the nursery tale. It is told to grown-up people, not as belief but as what was once believed; it is told to children, not to men; to lovers of romance, not to worshippers of the unknown; it is told by mothers and nurses, not by philosophers or priestesses; in the gathering ground of home life, or in the nursery, not in the hushed sanctity of a great wonder.[209] The influence of changing conditions upon the position of mythic tradition is well illustrated by Dr. Rivers in his account of the Todas. This people, he says, "are rapidly forgetting their folk-tales and the legends of their gods [that is, their myths], while their ceremonial remains to a large extent intact and seems likely to continue so for some time." Dr. Rivers attributes this to the effect of intercourse with other people. This intercourse has had no missionary results and has not therefore affected their religious rites and ceremonies, but has shown itself largely in the form of loss of interest in the stories of the past.[210] In other words, and in accordance with the definitions I am suggesting, the primitive myths of the Todas have definitely assumed a secondary position as folk-tale, and not a strong position at that, while religion has clung to rite and formula.

Primitive myth dislodged in this fashion is sometimes preserved in a special manner and for religious purposes in its ancient setting as a belief, or as a tradition belonging to sacred places and appertaining to sacred things. This is what has happened to the Genesis myth of the Hebrews; it has also happened to some of the sacred myths of the Hindus, and perhaps to some of the sacred myths of the Greeks. In this position the myth may even be reduced to writing, and where this happens all the sacredness appertaining to tradition is transferred to the written instrument.

Thus in Arkadia, Pausanias tells us, was a temple of Demeter, and every second year, when they were celebrating what they called the greater mysteries, they took out certain writings which bore on the mysteries, and having read them in the hearing of the initiated, put them back in their place that same night.[211] In India examples occur of land being held for telling stories at the UchÁos or festivals of the goddess DÉvi.[212] The colleges of Rome, composed of men specially skilled in religious lore, and charged with the preservation of traditional rules regarding the more general religious observances, the proper fulfilment of which implied a certain amount of information, and rendered it necessary for the state in its own interest to provide the faithful transmission of that information, have been described by Mommsen.[213]

I pass to the third class of tradition, namely, the legend, and this need not detain us long. We have already illustrated it by the notes on history and folklore, and by its very nature it belongs essentially to the historic age. In dealing with legend, there is first to determine whether its characters are historical, or are unknown to history. If the former, there is next to disengage those parts of the tradition which, by their parallels to other traditions, or by their nature, may be safely certified as not belonging to the historical hero or to the period of the historical hero. If the latter, the details must be analysed to see what elements of culture are contained therein. In both cases tradition will have served a purpose, and that purpose must be sought. Tradition does not attach itself to an historical personage without cause. There is necessity for it, and in the case of Hereward the necessity was proved to have been the great gap in the history of a national hero. Tradition does not preserve details of primitive culture-history without cause, and in the examples already quoted it has been shown that this cause rests upon the indissoluble links which the uncultured peasant of to-day has with the pre-cultured past of his race. He will have forgotten all about his tribal life and its consequences, but will retain legends which are founded upon tribal life. He will have lost touch with ideas which proclaim that man or woman not of his tribe is an enemy to be feared or attacked, but will gladly relate legends which deal with events growing out of a state of perpetual strife among the ancestors of people now in friendship. He will not understand the personal tie of ancient times, but will listen to the legends attached to places in such strange fashion as to make places seem to possess a personal life full of events and happenings. He will know nothing of giants and ogres, but will love the legends which tell of heroes meeting and conquering such beings. The history of the school books is nothing to him, but the history unknowingly contained in the legends is very real, and is applied over and over again to such later events as by force of circumstances become stamped upon the popular mind and thus succeed in displacing the original. It would be an important contribution to history to have these legends collected and examined by a competent authority. They would be beacon lights of national history preserved in legend. It will be readily conceded, I think, that in attempting these definitions of the various classes of tradition, and in illustrating them from the records of man's life in various parts of the world, it has been impossible for me to deal with certain points in the problem before us. In particular I have not considered the favourite subject of the diffusion of folk-tales. I do not believe in a general system of diffusion, such a system, I mean, as would suffice to account for the parallels to be found in almost all countries.[214] I think diffusion occupies a very small part indeed of the problem, and that it only takes place in late historical times. It is a large subject, and I have virtually stated my answer to the theory of diffusion in the definitions and classifications which I have ventured to put forward. It may be considered by some that other facts in the conditions of myth, folk-tale, and legend would not confirm the general outline I have given of the three classes of tradition to which I have applied these terms; and of course there are many side issues in so great a problem. I would not urge the correctness of the views I have put forward as applicable to every part of the world, or to every phase in the history of tradition; but I would urge that in the great centres of traditional life they are practically the only means of arriving at the position occupied by tradition, and that in all cases they form a working hypothesis upon which future inquirers may well base their researches.

II

Of late years there have been placed alongside of the traditional myth, folk-tale, and legend many other products of tradition—customs, ceremonies, practices, and beliefs, and it has been argued, and argued strongly and convincingly, that the tradition which has brought down the saga and song as far-off echoes of an otherwise unrecorded past has also brought down these other elements which must also belong to the same distant past. This argument is now no longer seriously disputed. But there still remains open for discussion the exact kind of evidence which these elements of tradition supply, the particular period or people from which they have descended, the particular department of history to which they relate. All this is highly disputed.

Folklore has in this department been greatly aided by Dr. Tylor's impressive terminology, whereby the custom, ceremony, practice, and belief which have come down by tradition are classed as "survivals." This term implies an ancient origin, and the necessary work of the student is to get back to the original. Until very lately the fact of survival has carried with it the presumption of ancient origin, but Mr. Crawley has raised an objection which I think it is well to meet. He urges that "the history of religious phenomena exemplifies in the most striking manner the continuity of modern and primitive culture; but there is a tendency on the part of students to underestimate this continuity, and, by explaining it away on a theory of survivals, to lose the only opportunity we have of deducing the permanent elements of human nature."

This sentence at once prepares us for much that follows; but Mr. Crawley leaves the point itself untouched, except by implication, until he is in the middle of his book, and then we have his dictum that "it may be finally asserted that nothing which has to do with human needs ever survives as a mere survival."[215] It will at once be seen that we have here a new estimate of the force which survivals play in the evidence of human progress. They prove the continuity of modern and primitive culture. They are part and parcel of modern life, filling a vacuum which has not been filled by modern thought, carrying on, therefore, the standard of religious belief and religious ideal from point to point until they can be replaced by newer ideas and concepts. This definition of survivals is very bold. It answers Mr. Crawley's purpose and argument in a way which no other fact in human history, so far as we can judge, could answer it. It is the basis upon which his whole argument is founded. Occupying such an important place, it should have received explicit investigation, instead of being treated as a sort of side issue of incidental importance.

When explicit investigation is undertaken, Mr. Crawley's case must, I think, break down. Survivals are carried along the stream of time by people whose culture-status is on a level with the culture in which the survivals originated. It matters not that these people are placed in the midst of a higher civilisation or alongside of a higher civilisation. When once the higher civilisation penetrates to them, the survival is lost. There is not continuity between modern and primitive thought here, but, on the contrary, there is strong antagonism, ending with the defeat and death of the primitive survival. This is the evidence wherever survivals can be studied, whether in the midst of our own civilisation, or even of primitive civilisations, which constantly exhibit traces of older beliefs and ideas being pushed out of existence by newer. It is, indeed, a mistake to suppose, as some authorities apparently do, that survivals can only be studied when they are embedded in a high civilisation. It is almost a more fruitful method to study them when they appear in the lower strata; and even in such a case as the Australian aborigines I think that it is the neglect of observing survivals that has led to some of the erroneous theories which have recently been advanced against Messrs. Spencer and Gillen's conclusions.

For the purpose of examining survivals in custom, rite, and belief, we have nothing more than a series of notes of customs and beliefs obtaining among the lower and lowest classes of the people, and not being the direct teaching of any religious or academic body. These notes are very unequal in value, owing to the manner in which they have been made. They are often accidental, they are seldom if ever the result of trained observation, and they are often mixed up with theories as to their origin and relationship to modern society and modern religious beliefs. To a great extent the two first of these apparent defects are real safeguards, for they certify to the genuineness of the record, a certificate which is more needed in this branch of inquiry than perhaps in any other. But with regard to the third defect there is considerable danger. An inquirer with an object is so apt to find what he wishes to find, either by the exercise of his own credulity or the ingenuous extension of inquiry into answer; whereas the inquirer who is content to note with the simplicity of those who occupy themselves by collecting what others have not collected, may be deficient in the details he gives, but is seldom wrong or violently wrong in what he has recorded. In every direction, however, great caution is needed, and especially where any section of custom and belief has already been the subject of inquiry. It is indeed almost safe to say that all research into custom and belief, even that of such masters as Tylor, Lang, Hartland, Frazer, and others, needs re-examination before we can finally and unreservedly accept the conclusions which have been arrived at.

Such an examination must be directed towards obtaining some necessary points in the life-history of each custom, rite, and belief. We have to approach this part of our work guided by the fact that folklore cannot by any possibility develop. The doctrine of evolution is so strong upon us that we are apt to apply its leading idea insensibly to almost every branch of human history. But folklore being what it is, namely the survival of traditional ideas or practices among a people whose principal members have passed beyond the stage of civilisation which those ideas and practices once represented, it is impossible for it to have any development. When the original ideas and practices which it represents were current as the standard form of culture, their future history was then to be looked for along the lines of development. But so soon as they dropped back behind the standard of culture, whatever the cause and whenever the event happened, then their future history could only be traced along the lines of decay and disintegration. We are acquainted with some of the laws which mark the development of primitive culture, but we have paid no attention to the influences which mark the existence of survivals in culture. For this purpose we must first ascertain what are the component parts of each custom or superstition; secondly, we must classify the various elements in each example; and thirdly, we must group the various examples into classes which associate with each other in motif and character.

By this treble process we shall have before us examples of the changes in folklore, and demonstrably they are changes of decay, not of development. By grouping and arranging these changes it may be possible to ascertain and set down the laws of change—for that there are laws I am nearly certain. It is these laws which must be discovered before we can go very far forward in our studies. Every item of custom and superstition must be tested by analysis to find out under which power it lives on in survival, and according to the result in each case, so may we hope to find out something about the original from which the survival has descended.

Each folklore item, in point of fact, has a life history of its own, and a place in relationship to other items. Just as the biography of each separate word in our language has been investigated in order to get at Aryan speech as the interpretation of Aryan thought, so must the biography of each custom, superstition, or story be investigated in order to get at Aryan belief or something older than Aryan belief. We must try to ascertain whether each item represents primitive belief by direct descent, by symbolisation, or by changes which may be discovered by some law equivalent to Grimm's law in the study of language.

Analysis of each custom, rite, or belief will show it to consist of three distinct parts, which I would distinguish by the following names:—

1. The formula.

2. The purpose.

3. The penalty or result.

It will be found that these three component parts are not equally tenacious of their original form in all examples. In one example we may find the formula either actually or symbolically perfect, while the purpose and penalty may not be easily distinguishable. Or it may happen that the formula remains fairly perfect; the purpose may be set down to the desire of doing what has always been done, and the penalty may be given as luck or ill-luck. Of course, further variations are possible, but these are usually the more general forms.

I will give an example or two of these phases of change or degradation in folklore. First, then, where the formula is complete, or nearly so, and the purpose and penalty have both disappeared. At Carrickfergus it was formerly the custom for mothers, when giving their child the breast for the last time, to put an egg in its hand and sit on the threshold of the outer door with a leg on each side, and this ceremony was usually done on a Sunday. Undoubtedly I think we have here a very nearly perfect formula; but what is its purpose, and what is the penalty for non-observance? Upon both these latter points the example is silent, and before they can be restored we must search among the other fragments of threshold customs and see whether they exist either separately from the formula or with a less perfect example. Secondly, where the formula has disappeared and the purpose and penalty remain, nearly the whole range of those floating beliefs and superstitions which occupy so largely the collections of folklore would supply examples. But I will select one example which will be to the point. When the Manx cottager looks for the traces of a foot in the ashes of his firegrate for the purpose of seeing in what direction the toes point, the penalty being that, if they point to the door, a death will occur, if to the fireplace, a birth,[216] there is no trace of the ancient formula. It is true we may find the missing formula in other lands; for instance, among some of the Indian tribes of Bombay. There the formula is elaborate and complete, while the purpose and the penalty are exactly the same as in the Isle of Man. But this hasty travelling to other lands is not, I contend, legitimate in the first place. We must begin by seeing whether there is not some other item of folklore, perhaps now not even connected with the house-fire group of customs and superstitions, whose true place is that of the lost formula of this interesting Manx custom. And when once we have taught ourselves the way to restore these lost formulÆ to their rightful places, the explanation of the mere waifs and strays of folklore will be attended with some approach to scientific accuracy, and we shall then be in a position to get rid of that shibboleth so dear to the non-folklore critic, that all these things we deal with are "mere superstitions."

Thirdly, when the formula is complete, or nearly so, and the purpose and penalty become generalised. At St. Edmundsbury a white bull, which enjoyed full ease and plenty in the fields, and was never yoked to the plough or employed in any service, was led in procession in the chief streets of the town to the principal gate of the monastery, attended by all the monks singing and a shouting crowd. Knowing what Grimm has collected concerning the worship of the white bull, knowing what is performed in India to this day, there is no doubt that this formula of the white bull at St. Edmundsbury has been preserved in very good condition. The purpose of it was, however, not so satisfactory. It is said to have taken place whenever a married woman wished to have a child; and the penalty is lost in the obvious generalisation that not to perform the ceremony is not to obtain the desired end.[217]

The second process, that of classification of the various elements in each example, will reveal some characteristics of folklore, which, so far as I know, have never yet been taken count of. One very important characteristic is the prevalence of a particular belief attached to different objects in different places. Thus Sir John Rhys in his examination of Manx folklore stopped short in his explanation of the superstition of the first-foot, because he had heard that, while in the Isle of Man it was attached to a dark man, elsewhere it was attached to a fair man. Of the examples where, on New Year's morning, it is held to be unlucky to meet a dark person, I may mention Lincolnshire, Durham, Yorkshire, and Northumberland. It is, on the contrary, lucky to meet, as first-foot, a dark-haired man in Lancashire, the Isle of Man, and Aberdeenshire.[218] In these cases we get the element of "dark" or "fair" as the varying factor of the superstition; but instances occur in Sutherlandshire, the West of Scotland, and in Durham, where the varying factor rests upon sex—a man being lucky and a woman being unlucky.

Similarly of the well-known superstition about telling the bees of the death of their owner, in Berkshire, Bucks, Cheshire, Cornwall, Cumberland, Lincolnshire, Lancashire, Monmouthshire, Notts, Northumberland, Shropshire, Somersetshire, Suffolk, Surrey, Sussex, Wilts, Worcestershire, it appears that a relative may perform the ceremony, or sometimes a servant merely, while in Derbyshire, Hants, Northants, Rutland, and Yorkshire it must be the heir or successor of the deceased owner. Again, while in the above places the death of the owner is told to the bees, in other places it is told to the cattle, and in Cornwall to the trees;[219] and, in other places, marriages as well as death are told to the bees.[220]

In some cases the transfer from one object to another of a particular superstition is a matter of absolute observation. Thus, the labourers in Norfolk considered it a presage of death to miss a "bout" in corn or seed sowing. The superstition is now transferred to the drill, which has only been invented for a century. Again, in Ireland, it is now considered unlucky to give any one a light for his pipe on May-day—a very modern superstition, apparently. But the pipe in this case has been the means of preserving the old superstition found in many places of not giving a light from the homestead fire.

I will just refer to one other example, the well-known custom of offering rags at sacred wells. Sir John Rhys thought that the object of these scraps of clothing being placed at the well was for transferring the disease from the sick person to some one else. But I ventured to oppose this idea, and considered that they were offerings, pure and simple, to the spirit of the well, and referred to examples in confirmation. Among other items, I have come across an account of an Irish "station," as it is called, at a sacred well, the details of which fully bear out my view as to the nature of the rags deposited at the shrine being offerings to the local deity. One of the devotees, in true Irish fashion, made his offering accompanied by the following words: "To St. Columbkill—I offer up this button, a bit o' the waistband o' my own breeches, an' a taste o' my wife's petticoat, in remimbrance of us havin' made this holy station; an' may they rise up in glory to prove it for us in the last day."[221] I shall not attempt to account for the presence of the usual Irish humour in this, to the devotee, most solemn offering; but I point out the undoubted nature of the offerings and their service in the identification of their owners—a service which implies their power to bear witness in spirit-land to the pilgrimage of those who deposited them during lifetime at the sacred well.[222]

Now, in all these cases there is an original and a secondary, or derivative, form of the superstition, and it is our object to trace out which is which. Do the rags deposited at wells symbolise offerings to the local deity? If so, they bring us within measurable distance of a cult which rests upon faith in the power of natural objects to harm or render aid to human beings. Does the question of first-foot rest upon the colour of the hair or upon the sex of the person? I think, looking at all the examples I have been able to examine, that colour is really the older basis of the superstition, and, if so, ethnological considerations are doubtless the root of it. Again, if the eldest son of the deceased owner of bees appears in the earliest form of the death-telling ceremony, we have an interesting fragment of the primitive house-ritual of our ancestors.

When, however, we come upon the worship of local deities, when we can suggest ethnological elements in folklore, and when we can speak of the house-father, and can see that duties are imposed upon him by traditional custom, unknown to any rules of civilised society, we are in the presence of facts older than those of historic times. It is thus that folklore so frequently points back to the past before the age of history. Over and over again we pause before the facts of folklore, which, however explained, always lead us back to some unexplored epoch of history, some undated period, which has not revealed its heroes, but which has left us a heritage of its mental strivings.

The method of using these notes of custom, rite, and belief for scientific purposes is therefore a very important matter. It is essential that each single item should be treated definitely and separately from all other items, and, further, that the exact wording of the original note upon each separate item should be kept intact. There must be no juggling with the record, no emendations such as students of early literary work are so fond of attempting. Whatever the record, it must be accepted. The original account of every custom and belief is a corpus, not to be tampered with except for the purpose of scientific analysis, and then after that purpose has been effected all the parts must be put together again, and the original restored to its form.

The handling of each custom or belief and of its separate parts in this way enables us, in the first place, to disentangle it from the particular personal or social stratum in which it happens to have been preserved. It may have become attached to a place, an object, a season, a class of persons, a rule of life, and may have been preserved by means of this attachment. But because every item of folklore of the same nature is not attached to the same agent wherever that particular item has been preserved, it is important not to stereotype an accidental association as a permanent one. Moreover, the modern association is not necessarily the ancient association, and there is the further difficulty created by writers on folklore classifying into chapters of their own creation the items they collect or discuss.[223] In the second place, we are enabled to prepare each item of folklore for the place to which it may ultimately be found to belong. The first step in this preparation is to get together all the examples of any one custom, rite, or belief which have been preserved, and to compare these examples with each other, first as to common features of likeness, secondly as to features of unlikeness. By this process we are able to restore what may be deficient from the insufficiency of any particular record—and such a restoration is above all things essential—and to present for examination not an isolated specimen but a series of specimens, each of which helps to bring back to observation some portion of the original. The reconstruction of the original is thus brought within sight.

Generally, it may be stated that the points of likeness determine and classify all the examples of one custom or belief; the points of unlikeness indicate the line of decay inherent in survivals. This partial equation and partial divergence between different examples of the same custom or belief allows a very important point to be made in the study of survivals. We can estimate the value of the elements which equate in any number of examples, and the value of the elements which diverge; and by noting how these values differ in the various examples we shall discover the extent of the overlapping of example with example, which is of the utmost importance. A given custom consists, say, of six elements, which by their constancy among all the examples and by their special characteristics may be considered as primary elements, in the form in which the custom has survived. Let us call these primary elements by algebraical signs, a, b, c, d, e, f. A second example of the same custom has four of these elements, a, b, c, d, and two divergences, which may be considered as secondary elements, and which we will call by the signs g, h. A third example has elements a, b, and divergences g, h, i, k. A further example has none of the primary elements, but only divergences g, h, i, l, m. Then the statement of the case is reduced to the following:—

The first conclusion to be drawn from this is that the overlapping of the several examples (No. 1 overlapping No. 2 at a, b, c, d, No. 2 overlapping No. 3 at a, b + g, h, No. 3 overlapping No. 4 at + g, h, i) shows all these several examples to be but variations of one original custom, example No. 4, though possessing none of the elements of example No. 1, being the same custom as example No. 1. Secondly, the divergences g to m mark the line of decay which this particular custom has undergone since it ceased to belong to the dominant culture of the people, and dropped back into the position of a survival from a former culture preserved only by a fragment of the people.[224]

The first of these conclusions is not affected by the order in which the examples are arranged; whether we begin with No. 4 or with No. 1, the relationship of each example to the others, thus proved to be in intimate association, is the same. The second conclusion is necessarily dependent upon what we take to be "primary elements" and "secondary elements;" and the question is how can these be determined? As a rule it will be found that the primary elements are the most constant parts of the whole group of examples, appearing more frequently, possessing greater adherence to a common form, changing (when they do change) with slighter variations; while the secondary elements, on the other hand, assume many different varieties of form, are by no means of constant occurrence, and do not even amongst themselves tend to a common form. The primary elements, therefore, constitute the form of the custom which represents the oldest part of the survival. They alone will help us to determine the origin of the custom, whether by features represented in the elements thus brought together or by comparison with ancient custom elsewhere or with survivals elsewhere similarly reconstituted. Altogether these elements, thus linked together by the tie of common attributes, are parts of one organic whole, and it is on this reconstructed organism we have to rely for the evidence from tradition.

When any given custom or belief has undergone this double process of analysis of its component parts and classification of its several elements, another process has to be undertaken, namely, to ascertain its association with other customs or beliefs, in the same country or among the same people, each of which customs or beliefs, being treated in exactly the same manner, is found to exhibit some degree of relationship in origin, condition, or purpose to the whole group under examination. In this way classification, analysis, and association go hand in hand as the necessary methods of studying survivals. Without analysis we cannot properly arrive at a classification; without classification we cannot work out the association of survivals.

The process is perhaps highly technical and complicated. It may not be of interest to all to discuss the process by which results are attained when what is most desired are the results themselves. But in truth the two parts of this study cannot well be separated. To judge of the validity of the results one must know what the process has been, and too often results are jumped at without warrant; items of custom and usage or of belief and myth are docketed as belonging to a given phase of culture, a given group of people, when they have no right to such a place in the history of man. It is not only distasteful to the inquirer, but almost impossible to dislodge any item of folklore once so placed, and thus much of the value of the material supplied by folklore is lost or discounted.

Custom, rite, and belief treated in this fashion become veritable monuments of history—a history too ancient to have been recorded in script, too much an essential part of the folk-life to have been lost to tradition. We may hope to restore therefrom the surviving mosaic of ancient institutions, ancient law, and ancient religion, and we may further hope, with this mosaic to work upon, to restore much of the entire fabric which has been lying so many centuries beneath the accumulated and accumulating mass of new developments representing the civilisation of the Western world.

III

It is only here that we can discover the point where we may properly commence the work of comparative folklore. An item of folklore which stands isolated is practically of no use for scientific investigation. It may be, as we have seen, that the myth is in its primary stage as a sacred belief among primitive people, in its secondary or folk-tale stage as a sacred memory of what was once believed, in its final or legendary stage when it does duty in preserving the memory of a hero or a place of abiding interest. It may be, as we have seen, that the custom, rite, or belief is a mere formula without purpose or result, a mere traditional expression of a purpose without formula or result, a mere statement of result without formula or purpose. We must know the exact position of each item before we begin to compare, or we may be comparing absolutely unlike things. The exact position of each item of folklore is not to be found from one isolated example. It has first to be restored to its association with all the known examples of its kind, so that the earliest and most complete form may be recorded. That is the true position to which it has been reduced as a survival. This restored and complete example is then in a position to be compared either with similar survivals in other countries on the same level of culture, or within the same ethnological or political sphere of influence, or with living customs, rites, or beliefs of peoples of a more backward state of culture or in a savage state of culture. Comparison of this kind is of value. Comparison of a less technical or comprehensive kind may be of value in the hands of a great master; but it is often not only valueless but mischievous in the hands of less experienced writers, who think that comparison is justified wherever similarity is discovered.

Similarity in form, however, does not necessarily mean similarity in origin. It does not mean similarity in motive. Customs and rites which are alike in practice can be shown to have originated from quite different causes, to express quite different motifs, and cannot therefore be held to belong to a common class, the elements of which are comparable. Thus to take a very considerable custom, to be found both in folk-tales and in usage, the succession of the youngest son, it is pretty clear that among European peoples it originated in the tribal practice of the elder sons going out of the tribal household to found tribal households of their own, thus leaving the youngest to inherit the original homestead. But among savage peoples where the youngest son inherits the homestead, he does not do so because of a tribal custom such as that to be found in the European evidence. It is because of the conditions of the marriage rites. Thus among the Kafir peoples of South Africa

Dr. Nicholson, from whom I quote this evidence, goes on to say that the

Quite independently of Dr. Nicholson I had come to the same conclusion;[226] and Dr. Nicholson, after handsomely acknowledging my priority in the "discovery," very properly alludes to the not unimportant fact of two workers in the same field coming to like conclusions. It is remarkable that the same distinction between the succession of the youngest son and of the son of the youngest wife appears in folk-tales.[227] Now clearly it would be quite wrong to suggest a parallel between the heirship of the youngest among the Kafir peoples of Africa and heirship of the youngest among the tribal people of early Europe. They are not comparable at all points, and it is just where the point of comparison fails that it becomes so important to science.[228]

I will take one other example, and this is the important practice of human sacrifice which looms so largely in anthropological research, and which is considered by so good an authority as Schrader to have taken a prominent place among the Aryans,[229] though he takes his examples, not from language, but from the unexamined customs of the Greeks, Romans, northerns, Indians, and Persians. We know more about the development of sacrifice now that Professor Robertson Smith has dealt with the Semitic part of the evidence. Without resting on the fact that the occurrence of human sacrifice in a country occupied by Aryan-speaking people does not, of itself alone, imply that the rite was Aryan, it is far more important to point out that among the higher races "the feeling that the slaying involves a grave responsibility and must be justified by divine permission" appears, and "care was taken to slay the victim without bloodshed, or to make believe that it had killed itself."[230] This feeling marks distinctly the Greek sacrifice as at Thargelia and in the Leukadian ceremony, the Roman sacrifice at the Tarpeian Rock, the sacrifice at the Valhalla rock of the northerns, while among the Hindus there is much to show that the idea of human sacrifice in some of the early writings is a literary borrowing from the Hebrews; and that if it ever prevailed among the Aryas of India it was very early superseded by the sacrifice of animals.[231] Colonel Dalton has given good reasons for his views "that the Hindus derived from the aboriginal races the practice of human sacrifices."[232] Although, then, Greek ritual and Greek myth are full of legends which tell of sacrifices once human, but afterwards commuted into sacrifices where some other victim is slain or the dummy of a man is destroyed;[233] although the significant Hindu ceremonial of so throwing the limbs of an animal slaughtered to be burnt with the dead that every limb lies upon a corresponding part of the corpse;[234] although Teuton, Celt, and Norse[235] are credited with the practice by authorities not to be questioned, it appears by the evidence that the European form of human sacrifice has little in common with the savage form except in the nature of the victim. It occurred, as Grimm states, when some great disaster, some heinous crime, had to be retrieved or purged, a kind of sacrifice, says Mr. Lang, not necessarily savage except in its cruelty; and the victims were not tribesmen, but captive enemies, purchased slaves, or great criminals.

These two examples will serve as warning against the too general acceptance of the custom and belief of savage and barbaric races, as identical with the custom and belief of early or primitive man. Such identification is in the main correct; but it is correct not because it has been proved by the best methods to be so, but because, of all possible explanations, this is the only one that meets the general position in a satisfactory manner. In many cases, however, it is monstrously incorrect, and it is the incorrect conclusion which weighs far more against the acceptance of the results of folklore than do the correct conclusions in its favour.

The work which has to be accomplished by the comparative method of research is of such magnitude that it needs to be considered. The labour and research might in point of volume be out of proportion to the results, and it may be questioned, as it has already been questioned by inference, whether it is worth the while. The first answer to this objection is that all historical investigation is justified, however much the labour, however extensive the research. Secondly, considering the very few results which the study of folklore has hitherto produced upon the investigations into prehistoric Europe, it must be worth while for the student of custom and belief to conduct his experiments upon a recognised plan in order to get at the secret of man's place in the struggle for existence, which is determined more by psychological than by physical phenomena. Thirdly, if the psychical anthropology of prehistoric times is to be sought for in the customs and beliefs of modern savages, it is of vital importance to anthropological science that this should be established by methods exactly defined. Whatever of traditional custom and belief is capable of bearing the test and of being definitely labelled as belonging to prehistoric man, becomes thereafter the data for the psychical anthropology of civilised man. Edmund Spenser understood this when his official duties took him among the "wild" Irish. "All the customs of the Irish," he says, "which I have often noted and compared with that I have read, would minister occasion of a most ample discourse of the original of them, and the antiquity of that people, which in truth I think to be more ancient than most that I know in this end of the world; so as if it were in the handling of some man of sound judgment and plentiful reading, it would be most pleasant and profitable."[236]

Comparative folklore, then, to be of value must be based upon scientific principles. The unmeaning custom or belief of the peasantry of the Western world of civilisation must not be taken into the domains of savagery or barbarism for an explanation without any thought as to what this action really signifies to the history of the custom or belief in question. No doubt the explanation thus afforded is correct in most cases, and perhaps it was necessary to begin with the comparative method in order to understand the importance and scope of the study of apparently worthless material. A new stage in comparative folklore must now be entered upon. It must be understood what the effective comparison of a traditional peasant custom or belief with a savage custom or belief really amounts to. The process includes the comparison of an isolated custom or belief belonging, perhaps secretly, to a particular place, a particular class of persons, or perhaps a particular family or person, with a custom or belief which is part of a whole system belonging to a savage race or tribe; of a custom or belief whose only sanction is tradition, the conservative instinct to do what has been done by one's ancestors, with a custom or belief whose sanction is the professed and established polity or religion of a people; of a custom or belief which is embedded in a civilisation, of which it is not a part and to which it is antagonistic, with a custom or belief which helps to make up the civilisation of which it is part. In carrying out such a comparison, therefore, a very long journey back into the past of the civilised race has been performed. For unless it be admitted that civilised people consciously borrow from savages and barbaric peoples or constantly revert to a savage original type of mental and social condition, the effect of such a comparison is to take back the custom or belief of the modern peasant to a date when a people of savage or barbaric culture occupied the country now occupied by their descendants, the peasants in question, and to equate the custom or belief of this ancient savage or barbaric culture with the custom or belief of modern savage or barbaric culture. The line of comparison is not therefore simply drawn level from civilisation to savagery; but it consists, first, of two vertical lines from civilisation and savagery respectively, drawn to a height scaled to represent the antiquity of savage culture in modern Europe, and then the level horizontal line drawn to join the two vertical lines. Thus the line of comparison is

We thus arrive at some conception of the work to be accomplished by and involved in comparative folklore. The results are worth the work. They relate to stages of culture in the countries of civilisation which are recoverable by no other means. The stages of culture are practically lost to history. In ancient Greek and Roman history, and in ancient Scandinavian history, there are priceless fragments of information which tell us much. But these fragments are not the complete story, and they belong to relatively small areas of European history. Every nation has the right to go back as far in its history as it is possible to reach. It can only do this by the help of comparative folklore. In our own country we have seen how history breaks down, and yet historical records in Britain are perhaps the richest in Europe. The traditional materials known to us as folklore are the only means left to us, and we can only properly avail ourselves of these when we have mastered the methods of science which it is necessary to use in their investigation.