Before concluding this Essay it may be well in a final chapter to consider its results in their bearing upon the conditions of early Anglo-Saxon society, and especially with regard to the division of classes and the character of the holdings.

The object has been to approach these difficult questions from the point of view of tribal custom.

The main clue to an understanding of the division of classes has been the amount of the wergelds.

The trouble taken to arrive at a correct knowledge of the currencies in which the wergelds were paid, tedious as it may have seemed to the reader, will not have been thrown away if it has led to the recognition of the fact that there was a very general correspondence in the amount of the wergelds tenaciously adhered to by the tribes of Western Europe, whether remaining in their old homes or settled in newly conquered countries. The amount of the wergelds was not seemingly a matter of race. Cymric and German customs were singularly similar.

If the Irish eric fine formed an exception, Irish tribal custom nevertheless had many things in common with Cymric and German custom in other respects.

It was from a study of the wergelds and the rights and liability of relatives in their receipt and payment that some idea was gained of the solidarity of the kindred under tribal custom. And this solidarity of the kindred was found to be closely connected with the family character of tribal land-holdings, of which the Cymric gwely was a typical example. Where direct evidence of this family element was wanting the liability of the kindred for the wergeld remained as an indication that it once had existed.

In reviewing the evidence of these matters and attempting to bring the results to a focus, we begin with the fact that with comparatively few exceptions the normal wergeld of the full or typical freeman was everywhere so large—200 gold solidi, the heavy mina of gold, traditionally representing 100 head of cattle. This wergeld was too large by far for the individual slayer to pay, and possible only as a payment from one group of kindred to another.

We have seen reason to infer from the Kentish, Wessex, and Mercian wergelds that the Anglo-Saxon tribes shared in these traditions, and, so to speak, brought their wergelds with them into Britain. And we have found that Anglo-Saxon custom as regards the wergelds was substantially similar to that of the Continental tribes.

From Beowulf we learned that, as there could be no feud within the kindred, a homicide within the kindred could not be avenged or compounded for. There was no galanas or wergeld in such a case under either Cymric or German custom, and evidence was found in the so-called Laws of Henry I. that it had been so also under Anglo-Saxon custom. Up to the time of the Norman Conquest the punishment of parricide was practically left by the laws to the spiritual jurisdiction of the Church (supra, p. 335).

The principle which required both paternal and maternal relations to join in the payment and receipt of wergelds, and nearly always in the proportion of two thirds and one third, was also common to Cymric and German tribes. This principle depended upon a view of marriage likewise common to both. A blood relationship was established as regards children of a marriage, while husband and wife for many purposes remained in their own kindreds. There being no blood relationship between husband and wife, the husband’s kindred alone were liable for his crimes and the wife’s alone for her crimes, and neither the husband nor the wife received any portion of the other’s wergeld or was liable for his or her homicides. Such was the custom under the Cymric codes and the laws of the Bretts and Scots, and Anglo-Saxon custom as described in the so-called Laws of Henry I. was similar.

The tribal feeling which allowed tribesmen and strangers to live side by side under their own laws, and made the Salic and Ripuarian Franks award a full wergeld to tribesmen of allied German tribes, while it gave only a half wergeld to the Gallo-Roman possessor who was not of their blood, was, it would seem, brought with the invading tribes into Britain.

Danish and English tribesmen were allowed to live side by side under their own laws and acknowledged as ‘equally dear,’ with a similar wergeld, while, at all events in the cases which come under notice, complete strangers in blood were awarded only a half wergeld as in the Continental laws.

We have not attempted to settle the question how far there was a Romano-British population left in the towns, but we have found incidental traces and hints that in Northumbria, Wessex, and Mercia there were ‘wilisc’ men—Welsh or British—who had only a half wergeld, being treated as strangers both in this respect and also as regards the substitution of the ordeal for the oaths of kindred (p. 403).

The principle that a man who could not bring to his protection the oath of his kinsmen must be brought to the ordeal was one of widely extended tribal custom. And it was emphasised by the adoption of the ordeal as a Christian ceremony solemnly performed in the churches under both Frankish and Anglo-Saxon law.

There can be little doubt that in the solidarity of the kindred under tribal custom we have to do with the strongest instinct which everywhere moulded tribal society. So far as it had its way and was not confronted by more potent forces it must have almost necessarily ruled such matters as the division of classes, the occupation of land, and the modes of settlement.

When we inquire into the grades of society under tribal custom they seem everywhere to have their roots in the principles of blood relationship. A man who has no kindred to protect him needs and seeks the protection of a chieftain or lord. By the force of tribal gravitation he sinks into the dependent condition of living upon another’s land.

Whether he be a freedman who has risen from the rank of the theow or thrall, or a free tribesman of low position, or one of a conquered race, or a stranger immigrant, and whether he be cottier or the holder of the typical yardland, until in the course of generations a kindred has grown up around him, he remains in the dependent condition. He is indeed a freeman as compared with the theow or thrall, but when Alfred and Guthrum make their compact and agree that Dane and English shall be reckoned as equally dear at the normal wergeld of the full freeman it is not of the dependent class they are thinking. They give to this class and to the Danish leysing or newly made freedman a twy-hynde instead of a twelve-hynde wergeld.

It might at first sight be supposed that this twy-hynde condition of the dependent class in England, so far as it may have included Anglo-Saxons, must have been the result of degradation in social status between the first settlements and the time of King Alfred, but we have sought in vain for evidence of an earlier higher position in the Laws of King Ine. And, on the whole, even when regarded solely from a tribal point of view, it does not seem unlikely that strangers in blood and freedmen and dependent followers of the conquering chieftains should find themselves after conquest and settlement in the economic condition of tenants and gafolgeldas on the lands of protecting lords. Nor would it be strange that, when in a new country and under other influences this uniform dependent economic condition had once become a general fact, the whole class, in spite of variety of origins, should find itself marked by a twy-hynde wergeld.

It does not follow, however, that because in the compact between Alfred and Guthrum the twy-hynde class were reckoned as equally dear with the Norse leysing that the Anglo-Saxon ‘ceorl who sits on gafol-land’ was generally in as low a social position as the Norse newly made freedman. It is enough that according to the evidence, he was a dependent tenant, let us say, under the lordship of a twelve-hynde man or if settled upon royal demesne of some gesith or official of the king.

Still it may be well to look once again at the position of the Norse leysing, because, after all, it is with the leysing that the Anglo-Saxon twy-hynde gafolgelda was equated in a compact made after King Alfred’s victory, and so when the two chieftains seemed to be treating on equal terms. Surely King Alfred was not intending to degrade the Anglo-Saxon dependent class. Presumably he was making a good bargain for them.

The early Norse laws were settled long after the date of this compact, upon the conversion of South Norway, and, as in other cases, they were framed with the express purpose of making room in the legal system for the Christian Church and so in some sense with its sanction. And yet so deep was the gulf between classes even then that a certain portion of the churchyard was set apart for leysings, and in no case were they to be buried in the portion reserved for classes above them. And if after giving his freedom-ale and so attaining the first step in freedom and independence the leysing should die leaving destitute children whose support ought not to be thrown back upon his lord, we have seen that the way out of the difficulty was to dig a grave in the churchyard into which the leysing’s children were to be placed and left to starve to death, the last survivor being the only one which the lord thenceforth had to maintain.[320] This was the position of the leysing at the bottom of the ladder of freedom.

But we found the leysing of the Norse laws rising by steps into greater freedom and better social position. And the process throughout was founded upon the gradual growth of kindred. It was the lack of kindred to swear for them and defend them which placed them low in the social scale, and it was the gradual growth of kindred generation after generation which marked the steps of their rise into better social position with higher wergelds.

When we turn to the Anglo-Saxon laws we seem to detect similar tribal principles originally at work but with differences which may very probably be referred to the circumstances attendant upon conquest and settlement in Britain.

The law of tribal gravitation here as elsewhere, aided, no doubt, by other potent forces, had been at work placing the man with an imperfect kindred in a dependent position at the bottom of the social ladder.

And it is important to note that at first the middle rungs of the ladder by which a man could climb out of the dependent position seem to have been present here as in Norway. The evidence is scanty, but sufficiently important.

From the Kentish laws the presence of stepping-stones into greater freedom may be inferred in the case of the three classes of lÆts with their rising wergelds. And in a precious fragment of ancient custom happily rescued from oblivion we found evidence that, originally at all events, there had been a way out of the ceorl’s twy-hynde condition at the fourth generation of landholding connected with payment of gafol to the king’s utware and direct service to the king. But we recognised that the collector of the fragment looked longingly back to ancient custom, speaking of it in the past tense, as if it was no longer in force.[321]

It would obviously not be wise to trust solely to the negative evidence of the silence of the laws, but in this case the silence seems to confirm the evidence of the fragment. For the pathetic tone of the fragment finds an echo in the fact that all traces of the middle rungs in the ladder seem to have vanished from the later laws. There is no mention in Ine’s laws or in Alfred’s of there being or having been several grades of freedmen or lÆts. Even the half wergeld of the six-hynde stranger who has risen to the possession of five hides silently disappears after King Alfred’s time. From whatever cause, so far as the evidence goes, the twy-hynde class seems to have become a homogeneous class in which, in spite of different origins, distinctions were merged in a common economic condition. Differences of origin were perhaps forgotten as the result of comradeship in the long struggle against the Danish foe.

We thus seem to be driven to recognise the width and to some extent the bridgelessness, already in King Alfred’s time if not in King Ine’s, of the gulf between the position of the twelve-hynde landed class and that of the twy-hynde dependent class of gafolgeldas and geburs who were tenants on their land.

It seems probable that, though technically and really free in the sense of not being thralls, the twy-hynde class, broadly speaking, may have found themselves very early, if not from the first, placed in an economic condition of service and servitude, including work as well as gafol, which by the ultimate disappearance of the middle rungs of the ladder might very easily slide into what is loosely called the ‘serfdom’ of later times.

In the meantime we realise that the abjectness of this semi-servile condition may be very easily exaggerated by modern associations with the terms ‘service’ and ‘serfdom.’

It is when we turn from the twy-hynde class to the position of the class above them, of gesithcund and twelve-hynde men, that we learn that a part at least of the risk of misunderstanding may lie in the difference between the tribal notion of service and freedom and the more modern one.

What, then, has tribal custom to teach us as to the position and services of the twelve-hynde class?

Reverting once more to the compact between Alfred and Guthrum, Dane and English are to be equally dear at eight half-marks of gold. The Englishman, without any limiting adjective, is the twelve-hynde man. And he is put on a level with the Danish typical free landholder, the hauld or odalman of the Norse laws, whose wergeld under Norse law was that of the typical freeman everywhere—equivalent to the normal wergeld of 200 gold solidi, the mina of gold, the traditional wergeld of 100 head of cattle. It was six times that of the Norse leysing, just as the twelve-hyndeman’s wergeld in England was six times that of the ‘ceorl who sits on gafol land.’

The English twelve-hynde man is therefore put on a level with the Norse odaller or typical landholder. And so, as we have seen, the ceorl who rose by the middle rungs of the ladder into the twelve-hynde position had inter alia to become a landholder of 5 hides, and his family became gesithcund only after the landholding had continued to the fourth generation. His great-grandchildren then became gesithcund with a twelve-hynde wergeld.

The twelve-hyndemen were therefore landholders, surrounded, in principle at least if not always in practice, by a kindred. But what kind of a landholding was it?

Approaching the question strictly from a tribal point of view, the solidarity of the kindred involved in the payment and receipt of wergelds would certainly suggest that those who had a right to receive and the obligation to pay held a position in their kindred quite different from that of the modern individual owner of land.

The analogy of Welsh and Irish and Salic and Norse and Scanian tribal custom would lead us to infer that the Anglo-Saxon settlers in England must have brought with them traditions of tribal or family ownership more or less of the type of the Cymric gwely, though doubtless modified by emigration and settlement in a new country.

After all that has been said, traditions and perhaps actual examples of the individual ownership of the ‘Romanus possessor,’ and, still more likely, actual experience of the Roman type of landed estates, may have survived in Britain from the period of the Roman occupation, and the Anglo-Saxon settlers may easily have been influenced in the matter of landholding by what as conquerors they came to supplant. But they can hardly have wholly cast off their own tribal traditions and instincts. The continued payment and receipt of wergelds show that they did not. Even, to take an extreme case, if they came to Britain as single settlers having left their kinsmen behind them, still kindreds would gradually grow up around their descendants in the new country. And tribal custom left to itself would give to them landed rights, quite different from those of the individual owner. But the interval, apart from other outside influences, may well have subjected tribal custom to a strain.

From the point of view of this interval it may not be unreasonable to revert to the clauses of King Alfred’s laws on ‘kinless men’ and the Norman precedent, that the king was to take the place of the missing maternal kindred and of advocate for a Norman if he had no other.[322]

Unless, therefore, the twelve-hynde settler was surrounded by a full kindred in the new country, he must, according to his own tribal custom, have found himself much more of an individual than he was used to be, and therefore more dependent upon the protection of his chieftain or king.

We must not, on the one hand, conceive of the twelve-hynde settler as having all at once adopted the independent position of the Roman ‘possessor,’ though circumstances may have sometimes severed him as completely from his ‘parentilla’ as the ceremony of the Salic law. Nor can we, on the other hand, conceive of him always as a tribesman surrounded by his kindred. He may evidently, on the one hand, be released from many of the trammels involved in membership of a kindred, but, on the other hand, he is thrown more than ever under conditions of service to the king.

Let us for a moment revert to the tribal conception of these trammels and services. They did not always involve degradation of social condition. They often, as we have seen, were the mark of the attainment of a higher position.

The kindred of the aillts or strangers who settled upon a chieftain’s land under Cymric custom was acknowledged in the fourth generation of continued occupation, but at the moment a kindred was acknowledged its members became adscripti glebÆ. When the Irish fuidhir did the same his descendants of the fourth generation found themselves not only bound to the land, but also bound together by something like the rules of the Cymric gwely, so that one of them could not sell or charge his share without the consent of the others. We found the same thing in Norway, where the rules for payment of the wergelds by relations were more elaborate than anywhere else, and where the growth of kindred seems so completely to have ruled the rise from one social grade to another, till at last a man whose great-grandfather’s great-grandfather was a freeborn landholder became an odaller. If at first sight we were to picture the odalman to ourselves as an individual freeholder of Roman or modern type we should soon find out our mistake when we learned that if he wanted to sell his odal he must first consult his odal-sharers. When examined closely the fact became evident that it was the group of kindred that by long settlement on the land had become odal, and that the shares of individuals in the odal were subject—with, of course, many differences—to some such tribal customs as those of the Cymric gwely. The odalman was thus not a single isolated landowner. He was surrounded by kindred odal like himself, reciprocally bound to fight for one another and swear for one another, and to share in the payment or receipt of one another’s wergeld. The odalman was protected by his kindred, but his freedom of individual action was restricted by it.

So also under Salic law the joint inheritors of the alod on terra Salica, with right of redivision between great-grandchildren per capita, were in the same way trammelled, and when by a solemn public form they released themselves from their obligations to their kindred they relinquished also all rights of inheritance and protection (p. 134).

Are we to consider these Continental analogies to be without relevance to Anglo-Saxon landholding?

Dr. Konrad von Maurer, in those masterly papers contributed in 1855 to the ‘Kritische Ueberschau’ which are still so valuable, rightly lays stress upon the power of the kindred as the great rival of the power of the state in the development of Anglo-Saxon polity. We find but little direct allusion to the kindred in the laws, it is true. But incidentally and as it were by accident we have learned from passages mentioned in their proper place that so late as the time of Athelstan there were kindreds both twelve-hynde and twy-hynde powerful enough to defy the King’s peace.[323]

This is in itself a significant reminder that more or less of tribal custom remained in force behind the screen of the laws from which most of our evidence has been taken. And yet we seem to be almost forced to the conclusion that if we try to realise the position of the twelve-hynde settler we must regard him, at all events for the first few generations, as in a very different position from that of the Norse odalman in the old country. Even though as head of his family he may have brought descendants and dependents with him, he could not in the new country be at once surrounded by kinsmen and odal-sharers who with himself had hereditary rights in the land.

We thus come round again to the point that so far as he may have been separated from his kindred the first Anglo-Saxon settler must have found himself thrown upon the protection of his chief and into a position of individual service. He becomes, as we have seen according to the scanty evidence of the Laws of Ine, a king’s gesith, with military and judicial and administrative duties to discharge, put into a post of service which he cannot relinquish at pleasure. Service to the king has to some extent taken the place of the restraints of kindred, and so in a sense, like the twy-hynde man, he has become a gafolgelda, but paying his gafol and services direct to the king, and adscriptus glebÆ, but tied to an estate and an official position instead of to a yardland.

This view of the position of the gesithcund and twelve-hynde class rests very much upon the incidental evidence of the Dooms of Ine, but the truth of it is confirmed by the independent evidence of the precious fragment already referred to. For its interesting evidence shows that, in addition to his holding of five hides of land, it was precisely into this position of gafol-paying and service direct to the king that the ceorl of ancient custom had to climb in order to earn the gesithcund status and the twelve-hynde wergeld.

Thus we arrive at a definite and practical mark distinguishing ultimately, and perhaps more or less from the first, the twelve-hynde and twy-hynde classes.

The twelve-hynde or landed class paid gafol and did service direct to the king. The twy-hynde or dependent class paid gafol and did service to the landed class, who from this point of view were middlemen between the twy-hynde gafolgelda and the king.

We seem, therefore, thus early to arrive at something analogous to Professor Maitland’s technical definition of the Manor as the fiscal unit from which gafol is paid direct to the king, while its lord is the receiver of the payments and services of its tenants. The single landholder who is not under a manorial lord in the Domesday survey is said to hold ‘as for a manor’—though he may have no tenants.

It may be worth while in this connection to allude to another general feature of the manorial estate on both sides of the Channel which if not directly of tribal origin must at least have worked in close sympathy with tribal custom.

The gesithcund man officially charged with the control of a district or estate easily became in a manorial sense lord of the dependent tenants upon it. And the judicial and magisterial adjunct to the lordship became a prevalent feature of the typical manor.

We have seen that the ‘sac and soc’ of later times may have grown from the root of the tribal principle involved in the sacredness of the precinct or area of protection of the chieftain and, in degree, of every grade of tribesman who possessed a homestead. How large a place this principle occupied is shown by the prominence of the fredus in Frankish law and of the mundbyrd in the early Anglo-Saxon laws. The manor was a complex product of many factors, and tribal custom was certainly one of them.

Once again, what kind of a holding was that of the twelve-hyndeman? Was it a family holding, and what were the rules of succession?

Unfortunately, we do not know how far the immigrants came in kindreds and families or as followers and ‘gesiths’ of military chieftains. But, in any case, if we may take the evidence of place-names the great number of patronymic names of places would lead to the supposition that the holdings were family holdings. The ham may at first have been the estate of a gesithcund man held direct of the king with gafolgeldas and geburs under him holding yardlands and doing work on his demesne. But when it becomes an ingham the patronymic termination points to the lordship of the manor having been held, as time went on, jointly, in somewhat the same way as the Cymric chieftainship in the gwely. His sons and grandsons and great-grandsons may really have had their rights of maintenance all along, and ultimately, if they were allowed to do so, they may have sometimes divided the inheritance instead of continuing to hold it jointly. Tribal instincts working alone would probably follow some such line as this.

But it is easy to see that the nearer the holding of the twelve-hyndeman approached to a benefice or office the stronger would be the tendency towards single succession instead of divisions among heirs.

During the century or two after the first settlement there was time, no doubt, for the growth of kindreds, and the thane in the king’s service would soon become the head of a family group; but, on the other hand, many influences were at work undermining the solidarity of the kindred and strengthening the manorial element. Tribal instincts die hard. But probably there never was full opportunity for the growth upon English soil of anything like the solidarity in landholding of the Norse kindreds of odal sharers tracing back their family possession for four or five generations.

There is but little evidence on the rights or rules of succession to be found in the laws. And the silence is suggestive of the continuance of custom. Even the diplomatic evidence of wills and charters is so much restricted to boc-land that it perhaps throws a shadow rather than direct light upon the ordinary devolution of land which had not become the subject of the Romanised rules of ownership, conveyance and testamentary disposition.

But if Professor Vinogradoff is right in his view that folk-land was that land which was still held under ancient custom, then for anything we know, in spite of documentary silence, folk-land may still have been held more or less as family rather than individual property even in later times.

If the suggested analogy between the terra Salica of the Salic laws and the folk-land of Anglo-Saxon documents could be proved, the family character of the holdings in both cases would receive confirmation. At the same time the frequent concurrence of relatives in Anglo-Saxon dispositions of land and the common form of deprecation of future interference on their part would at least be consistent with the supposition.

That as time went on the growing force of feudal principles would demand single succession to landed estates whenever they could be regarded as benefices is what might be expected. And it is worth noting that under later feudal custom, by a kind of compromise, what was really a family holding was often artificially moulded for practical purposes into a single holding with apparent single succession.

In the Domesday survey are many instances in which thanes or soldiers here and there hold manors or fractions of manors ‘pariter’ or ‘in paragio.’ And when the feudal tenancy ‘in parage’ is examined in its fully developed form on the Continent, it is found to present many resemblances to what under Cymric custom the family holding of a tribal chief of kindred might be if the chief alone were regarded as the landed person doing homage to the superior lord for all his kindred and if, in the next stage, when the gwely was internally divided between brothers, one of them only did homage for the rest. There were indeed in tribal custom as to the chieftainship and the constitution of the gwely traits which easily allowed themselves to be developed on feudal lines. For the present purpose, however, the point seems to be that within what looks from the outside like a single individual landholding there may have been internal family divisions which are not apparent.

Passing now from what may be regarded as the holdings of the twelve-hynde class, more or less tending to resemble manorial estates, to the yardlands of the twy-hynde class, room may perhaps be found even in their case for the exceptional continuance of the family element in spite of the apparent single succession.

The Kentish holdings in sulungs and yokes instead of in hides and yardlands seem to go back to the earliest Kentish records. The fact that, in spite of the difference in date between the evidence of the earliest charters and that of the Domesday survey and the surveys in the Battle Abbey records and the ‘Black Book of St. Augustine,’ the holdings seem to have been throughout in sulungs and yokes points to continuity. And when these sulungs and yokes in the surveys are found to be very often held by ‘the heredes of so and so,’ or ‘so and so and his pares,’ it seems fair to suggest that in these Kentish holdings there may have been a survival of family ownership.

Whether it were so or not, this later Kentish evidence shows at least that the continuance of family holdings was not necessarily inconsistent with external uniformity in the sulungs and yokes of the open-field system in Kent. And if so, why may not the same thing be true in exceptional cases of the hides and yardlands of Wessex and Mercia?

Contrary principles have a strange way in practice of finding a modus vivendi till one of them at last overrides the other.

It will be remembered that one of the complaints of the existence of kindreds powerful enough to defy the king’s peace in King Athelstan’s time came from Kent. And if these facts may be taken as evidence that the solidarity of kindreds had been better preserved in Kent than elsewhere some tribal light might perhaps be thrown upon the survival of the custom of gavelkind in Kent.

It is not a matter upon which we must dwell, but evidently the gavelkind tenure must have been something different from the prevalent tenures of other districts. The difference cannot have been the equal division of the sulungs and yokes between sons as contrasted with the single succession to the yardlands of other districts, because the sulungs and yokes were apparently not interfered with by the gavelkind division among heirs. And when the right of the youngest son under the custom of gavelkind to succeed to the parental hearth is compared with the similar right of the youngest son in the case of the Cymric gwely the inference becomes very strong that the gavelkind holdings were family holdings and the gavelkind divisions internal divisions within the family, like those of the Cymric gwely, not necessarily interfering with the permanence of the sulungs and yokes of the open-field system of which the family holdings were composed or in which the family had rights.

The surveys of Kentish manors in the records of Battle Abbey and the ‘Black Book of St. Augustine’ present instances sometimes of sulungs and yokes held by the heredes of a deceased person and sometimes of others which maintain their unity for purposes of payments and services although in the possession of several holders. The sulung in these cases seems to have continued to be the unit liable for the fixed ploughing and other services irrespective of the question who were its occupants.[324]

Once more perhaps some light may be gained from Cymric tribal custom.

We have learned from the Cymric evidence that a district might be divided for purposes of revenue and food rents into sub-districts, irrespective of who might be the occupants. And we have seen also how the Cymric trefgordd or unit of pastoral occupation, with its one plough and one churn and one herd of cattle under a single herdsman, could remain a permanent taxable unit paying the tunc pound in lieu of food rents, whoever might at the time be its occupants and have cattle in the herd. Within the lines of tribal custom itself the members of a Cymric gwely might be spread over a district and their cattle distributed among many trefgordds, while from the chieftain’s point of view the local units of taxation were uniform and regular.

But this must not blind our eyes to the fact that the yardlands on Anglo-Saxon estates were, so far as we can see, for the most part really individual holdings with actual single succession. However hard tribal custom may have fought for the family element, the manorial element in the end seems to have prevailed on most manors so as to secure, for the purposes of the lordship and the convenience of manorial management, single succession to the yardlands. The fact that as early as King Ine’s time we see new individual holdings of geburs being made by the allotment of yardlands and homesteads to individual tenants in return for gafol and work, when taken in connection with the ‘Rectitudines’ brings us back to the likeness of these holdings to the holdings of the villani of later times. We see in the allotment of stock to the gebur, of which we trace scattered evidence, the fact on which the principle of the later villenage was based. Only when both homestead and yardland came from the lord was there to be work as well as gafol under King Ine’s laws. The stock of the holding according to the ‘Rectitudines’ belonged in theory to the lord and upon the tenant’s death returned to the lord. The continuance to another tenant on the payment of a relief involved the admission that the holding and its outfit were a loan from the lord.

The fact that in exceptional cases family holdings were able to maintain their own under manorial management must not be allowed to lead us to underrate the power of the manorial element. There were in tribal custom itself as described by Tacitus elements of what we have elsewhere spoken of as the embryo manor, but this must not blind our eyes to the fact that something more was required to produce the general uniformity of holdings and single succession upon manorial estates than tribal custom working alone.

If from a tribal point of view we try to understand the growth of manorial serfdom and see how on the Continent it was seemingly the result of the combination of two leading factors, tribal custom and Roman methods of land management, it becomes hardly possible to ignore the presence of something like the same combination of two interacting factors on British or English ground.

With the manorial side of serfdom in its connection with the widely prevalent open field system we have already attempted to deal in a former volume. That there may have been some continuity and continuance of estates managed on the Roman system can hardly be denied. However far the policy of extermination of the old inhabitants was carried, it never extended over the whole area. And the whole of Britain was not conquered in the same century. Even if the continuity of estates in Britain should be considered to have been entirely broken by the Anglo-Saxon invasions (which is hardly conceivable), it must be admitted that continuity and likeness between England and the Continent as to land management was very soon restored on monastic and other ecclesiastical estates, and perhaps also upon what was Royal domain. Nor can it be doubted that herein was a force greatly strengthening the manorial element.

If we limit our view to the tribal side only of the problem, we recognise that in Scandinavia and in the Cymric districts of our own island and in Ireland tribal principles working alone tended powerfully, without help from the Roman side, to produce a class of tenants becoming adscripti glebÆ after four generations of occupation, but it did not produce either in Norway or in Wales or Ireland or in Celtic Scotland that general and typical form of occupation in uniform yardlands or ‘huben’ so prevalent in England and Germany on manorial estates with ostensibly single succession and services in so many points resembling those of the Roman colonate.

Whether the manor was the indirect or direct successor of the Roman Villa—i.e. whether the continuity was broken or not—the manorial use of the open-field system of agriculture seems to be required to produce the uniformity of holdings in yardlands and the single succession which marked what is roughly called the serfdom of the manorial estate.

It is hardly necessary to repeat that the open-field system itself was not of manorial origin. It was essentially an economic result and differed very greatly in its forms. Its main object seems to have been fairness and equality of occupation. Under tribal custom, in Wales, it arose out of coaration of portions of the waste or pasture by the common plough-team to which the tribesmen or the taeogs, as the case might be, contributed oxen. The strips were day-works of the plough taken in rotation by the contributors according to the place of their oxen in the plough-team for the season, and they returned into common pasture when the crop had been removed. The tribesman in the pastoral stage was the owner of oxen but not of the strips ploughed by them. They were merged again in the common pasture of the district in which he had rights of grazing for his cattle. And the cattle, and not the corn crops, were the main thing upon which the system turned.

Whatever method of distribution may have been followed, as arable farming increased and the strips became more and more permanently arable, mostly on the two-field or the three-field system, the area of unploughed land was more and more restricted and the pasture over the stubbles and fallows obviously became more and more essential. The cattle, on the one hand, required the pasture on the stubbles and fallows, and the land, before being ploughed again, required the manure arising from the pasturing of the flocks and herds upon it.

Where open-field husbandry still subsists in Western Europe, whether on this or the other side of the Channel, the owner of the strips has still no right of grazing upon his own strips till upon the appointed day when the common right begins of all the holders to graze their cattle in a common herd or flock over the whole area. This right is known in France as the ‘vaine pÂture,’ and it is still the most important and indestructible element in the open-field husbandry. In the great open fields around Chartres a man may plant his strips with vines if he likes, but to this day, if he does so, he must let the sheep of the commune graze over them after a certain date, in exercise of the immemorial right of the vaine pÂture.

In all this no manorial element need be present, and when the manorial element is absent there is not necessarily any uniformity or single succession in the holdings. But when manorial management comes upon the top of this widely extended and all but universal system of agriculture, whether in Roman times or later, the bundle of scattered strips which under tribal custom could be ploughed by a pair of oxen whether alone or in joint ploughing is very naturally taken as the typical holding. And thus when we find in the Laws of Ine and later records gafolgeldas and geburs settled upon yardlands and doing service by week-work on the lord’s demesne the natural inference must be that it is the result of manorial land management and that there has come into existence already something like a manor with something like a community in serfdom upon it, using the prevalent open-field system as the shell in which it will henceforth live so far as its agriculture is concerned.

And so it seems natural to attribute to the manorial management and the manorial requirement of fixed services and dues the uniformity of the holdings and the single succession by which the uniformity was preserved. The power which seeks and makes uniformity seems to come from above. Agricultural communities of free tribesmen who had become individual freeholders (if such could be conceived of as prevalent in King Ine’s and King Alfred’s time) would probably have used the open-field system in a quite different way. And we see no trace of it in the evidence.

When, however, we have said this we have no disposition to ignore or make light of the later evidence upon which great stress has quite rightly been laid by Professor Maitland in his remarkable work on ‘The Domesday Survey and beyond,’ showing that there were in some districts villages, in which the manorial element was apparently absent in the time of Edward the Confessor, though appearing as manors after the Conquest. He has suggested that in these villages not only the manor in name but also the manor as a thing was apparently non-existent. There was in these cases apparently, in King Edward’s time, no demesne land upon which the services of a tenantry in villenage could be rendered, and the tenants were often sokemen who had individually put themselves under the protection of this lord or that, instead of there being one lordship over the group, as in a manor.

These lordless villages on the eve of the Conquest as shown by the entries T. R. E. in the Domesday survey and especially in the ‘Inquisitio Eliensis,’ merit more careful study than has yet been given to them, and so far as they can be shown to prove the existence of free villages of liberi homines or socmanni, after the Conquest merged sometimes in the class of villani, I am ready to welcome the evidence. But unless they can be traced back to earlier times, their occurrence mostly in the Danish districts interspersed with other villages which were manors and had demesne land, together with the singular fact that the holders in these villages were commended to several lords, suggests that their peculiar position may date from the time of the Danish invasions, and be the result of the devastations as to the effects of which the ‘Liber Eliensis’ contains so much evidence. Many a manor may have lost both lord and tenants, and have been filled up again by the great lords of the district with new tenants—soldiers and servants who had served in the wars, it may be. Thus these cases, in which many features of the ordinary manor were apparently missing in the time of Edward the Confessor, may be of recent date and so, while important when viewed in relation to the Domesday survey and the changes made by the Conquest, not specially instructive as regards earlier Anglo-Saxon conditions.

Unfortunately, as we have seen, the laws of the Danish period, while recording existing and modified Anglo-Saxon customs on various points, leave us in the dark as to Danish custom, whether of old standing in the Danelaga or newly imported in King Cnut’s time. It was, no doubt, known to the invaders, and it was enough for them to say ‘as the law stands,’ though we do not know what it was. The whole question of the Danelaga was purposely omitted from the scope of my former volume, and now, after twenty years, still remains a subject requiring careful examination by future inquirers.

But this cannot be done completely until the minute work which Professor Maitland and Mr. Round and Mr. Corbett are gradually doing upon the Domesday survey itself in its local details has been further pursued, and it lies, with so many other branches of a difficult subject, beyond the limits of the inquiry made in this volume.

Reference may, however, be incidentally made to the numerous cases in which, in order to describe the nature of the tenure of socmanni and others under what were perhaps new conditions, the fact was recorded in varying phrases whether this person or that could or could not leave or sell his land. Of some it is stated ‘possunt recedere,’ of others ‘non possunt recedere’—of some ‘possunt vendere,’ of others ‘non possunt vendere.’ Though these tenures may have been comparatively modern and may belong to a period of advanced feudal conditions, still it may be possible that some trait of tribal custom may lurk at the root of the distinction. From the manorial point of view, it was necessary to record of the socmanni whether they had only limited rights in the land subject to the performance of services and ‘consuetudines’ (which, by the way, seem to have been very much like those of the villani) or whether they were permanent freeholders who could sell their holdings and leave the land when they liked.[325] The position of the tenants in this respect was probably dependent upon the tenure under which they held, i.e. upon whether they were tenants with only life interests, or for successive lives, or, as we should say, tenants in fee. After the devastations of war many new tenants must have been put upon desolated manors, and Professor Maitland has very rightly laid stress in another connection on the traditional habit of granting leases for three lives only, so that a holding might ultimately return to the lord. He has pointed out that when Bishop Oswald (A.D. 962-992), exercising manorial rights over the great domain of the Church of Worcester made these leases to thanes on certain services for three successive lives (i.e. for the lives of father, son, and grandson) he did it expressly for the purpose of securing to his successor full power to renew them or not.[326] And from a tribal point of view it may be a pertinent question whether the restriction to the three generations had not some indirect connection with the tribal custom or instinct, so often alluded to, which gave to the fourth generation of uninterrupted occupation fixity of tenure and status.

Recurring to the scattered cases of thanes holding ‘in paragio’ and by no means confined to the Danish districts,[327] it was necessary to state in the Domesday records, as in the case of the socmanni, whether they had or had not power to leave or to sell, and it may be useful that we should be reminded by these cases, in which feudal custom had possibly arisen out of tribal custom, that tribal custom was not unknown to the Danish and Norman conquerors of England. The Danish immigrants came from a district in which tribal custom was still fresh and vigorous. The Normans too, as is shown by the so-called Laws of Henry I., found Anglo-Saxon custom by no means altogether alien to their own instincts.

Before concluding this essay perhaps a further observation should be made.

We have learned in the course of this inquiry that it does not do to take too insular a view of Anglo-Saxon conditions. The similarity of wergelds, and indeed of tribal custom generally, has throughout become very apparent. But perhaps it is hardly more striking than the similarity in the modifications of tribal custom found in the laws on both sides of the Channel.

In their migrations and conquests the conquering tribes found themselves everywhere breathing a moral atmosphere in which it was difficult for the old tribal instincts to live. In such matters as the responsibility of a master for his slave’s homicides and of relatives for their kinsman’s crimes we have watched as it were modifications of tribal custom in the course of being made, here and there, on almost identical lines. May it not have been so also in regard to the important matter of the division of classes?

If we have recognised rightly the tribal principles originally at the root of the distinction between the twelve-hynde and twy-hynde classes there is no reason why we should not recognise also that besides the potent force of manorial management there may have been other influences at work widening the gulf between the two classes, and, so to speak, reducing to a level the members of each class by breaking away the rungs of the ladder between them.

It must not be overlooked that in the earliest Continental laws most nearly contemporary with those of Kent—Alamannic, Bavarian, Burgundian, and Wisigothic—the divisions of society have a very artificial look, as though largely based upon wealth rather than the tribal principles of kindred.

Now, German writers are not agreed upon the point whether these artificial divisions found in these earliest of the laws ought to be regarded as belonging to ancient German custom or whether they may not rather be traced to Roman influences.[328]

We have already seen how necessary it is in connection with these early laws to discriminate between ancient custom and the new influences which were working in them in the direction of individualism and the disintegration of the kindred. The earliest laws are, as we have seen, just those in which tribal custom had fared the worst.

In the Alamannic Pactus of the sixth century (Fragment ii. 36) the grades for wergelds were as under:—

And these were subdivisions of the ingenuus class, for there were below them the lidus and the servus. In another clause (iii. s. 25) a similar division is applied to animals. The penalties are given for killing ordinary, ‘mediana,’ and ‘meliorissima jumenta.’

In the Burgundian law the division of society into three grades—optimates, mediocres, and inferiores—is found in the Lex Romana and is applied to Romans and Burgundians alike. These divisions seem to supplant those of kindred, and to have no tribal principle at their root.[329]

In the Wisigothic laws the disintegration of tribal society is so far advanced that the wergelds of the ingenuus class are regulated, not by kindred or social position, but, as we have seen, according to the age of the individual.

It is difficult not to connect the substitution of artificial grades for those dependent on kindred with the Roman tendency to divide society into ‘patrician’ and ‘plebs,’ and the ‘plebs’ according to position and wealth into honestiores and humiliores.

Already in CÆsar’s time we see how difficult it was from a Roman point of view to understand the relation under tribal custom of the dependent tribesmen to their chieftain. CÆsar does not seem to have recognised the link of blood-relationship between them. To his view the chieftains were equites and the tribesmen almost their servi. It was difficult otherwise to bring the two classes within some recognised category of Roman law.

So it was no doubt, in degree, at the later period in the case of the conquering German tribes, when the Romanising forces were mainly in clerical hands.

The influence of the Church also told in favour of the artificial and anti-tribal division of the people into great men and small men. Its tenets of individual responsibility favoured individualism.

Canon XVI. of the Council of Orleans (A.D. 549) shows that the ecclesiastical mind in Gaul was familiar with the division into classes ‘majorum et mediocrium personarum.’

A canon of an earlier Council (A.D. 511) shows how by taking refuge in a church a homicide received protection till composition was arranged, and how thus the question of wergelds was brought within clerical recognition. Once brought within its power the Church was not likely to let it slip from its grasp. And the collections of FormulÆ of the Merovingian period show how the clergy joined with the other authorities in arranging the payment of wergelds and the prevention of private vengeance. From these formulÆ it would seem that the payment and perhaps the amount of the wergeld had become to some extent a matter of mediation and arrangement through the intervention of ‘boni homines’ who were sometimes ‘sacerdotes.’[330] And when the award was given and the payment made, it was natural that a formal charter of acknowledgment in stay of vengeance on the part of the relations of the slain should be insisted upon. Each set of formulÆ contains a form for this purpose. The matter of wergelds had become a subject of Franco-Roman conveyancing.

Romanising and clerical influences thus working together in connection with wergelds would naturally tend to exclude from consideration the question of kindred, and to make the payment of the wergeld a matter for the homicide alone.

Long before the time of King Ine these Romanising influences must have been at work in England, as elsewhere, introducing new considerations of justice and the position of classes founded on Roman law and Christian feeling, and not upon tribal custom.

We have recognised some such action as this in the nearly contemporary Canons and in the Kentish laws, as well as in the later Anglo-Saxon laws, and indeed again and again throughout this inquiry, so that while we have had to notice again and again the extent to which the Church succumbed to tribal custom when it suited its purpose to do so, it must not be forgotten how much of the modification of custom found in the laws was due to the influence of the Romanised Church.

It is not, therefore, enough to recognise only Romanised forms of land management under clerical influence. We must recognise also something of the same persistent antagonism of the Church to tribal custom which on the Continent had already in the sixth and seventh centuries sometimes succeeded in extruding considerations of kindred from the matter of wergelds, and to a great extent also from the question of the division of classes.

With this further recognition of outside influences, this contribution towards the understanding of a difficult question must come to an end. All that can be claimed on its behalf is that a few further steps in advance may have been made good. It may seem to have resulted rather in the restatement of some of the problems than in their solution. But this is what might be expected from the attempt to approach a subject which has many sides especially with light from the tribal side only. Following the true method of working from the known to the unknown, it is not until such a problem has been approached separately from its different sides that a final solution can be reached; and this involves the fellow work of many historical students.

In the meantime, without ignoring or seeking to minimise the force of other important influences, it may, I think, safely be said that we have found the influence of tribal custom upon Anglo-Saxon polity and economic conditions as apparent, all things considered, as there could be reason to expect.

It was a factor in economic development which, among others and in due proportion, has to be reckoned with, and its study has the special value that it helps to bring the student of the Anglo-Saxon laws to regard them from the point of view of the Anglo-Saxon settlers themselves.