Further, the fact of the prevalence of the double and single huf or hub of sixty and thirty acres over so large an area once Roman province, irresistibly suggests a connexion with the double and single yoke [p392] of oxen given as outfit to the Roman veteran, with such an allowance of seed as to make it probable, as we have seen, that the double yoke received normally fifty or sixty jugera, and the single yoke twenty-five or thirty jugera.

It is worth remembering, further, that in the Bavarian law before quoted, limiting the week-work of the servi on the ecclesiastical estates to three days a week, an exception is made allowing unlimited week-work to be demanded from servi who had been supplied with their outfit of oxen de novo by their lord. So that there is a chain of evidence as to the system of supplying the holders of 'yard-lands,' 'huben,' and 'yokes,' with an outfit of oxen, of which the Kelso 'stuht,' the Saxon 'setene,' the outfit of the servus under this Bavarian law, and that of the Roman veteran, are links.610

It is hardly needful to repeat that it does not follow from this that the system of allotting about thirty acres (varying in size with the locality) to the pair of oxen was a Roman invention. The clear fact is that it was a system followed in Roman provinces under the later empire, as well as in Germany and England afterwards; and, as the holding of thirty acres was found to be the allotment to each 'tate' or household under the Irish tribal system, it may possibly have had an earlier origin and a wider prevalence than the period or extent of Roman rule.

The scattering of the strips composing a yard-land or hub, over the open fields should also be once more mentioned in comparing the two. It was not [p393] confined to the 'yard-land' or 'hub.' It arose, as we have seen, in Wales, from the practice of joint ploughing, and was the result of the method of dividing the joint produce, probably elsewhere also, under the tribal system. It is the method of securing a fair division of common land in Scotland and Ireland and Palestine to this day, no less than under the English and German three-field system. And the remarkable passage from Siculus Flaccus has been quoted, which so clearly describes a similar scattered ownership, resulting probably from joint agriculture carried on by 'vicini,' as often to be met with in his time on Roman ground. This passage proves that the Roman holding (like the Saxon yard-land and the German hub) might be composed of a bundle of scattered pieces; but this scattering was too widely spread from India to Ireland for it to be, in any sense, distinctively Roman. It perhaps resulted, as we have seen, from the heaviness of the soil or the clumsiness of the plough, and the necessity of co-operation between free or semi-servile tenants, in order to produce a plough team of the requisite strength according to the custom of the country; and this necessity probably arose most often in the provinces north of the Alps.

Another point distinctive of the 'yard-land' and the 'hub' was the absence of division among heirs, the single succession, the indivisibility of the bundle of scattered strips in the holding. And this finds its nearest likeness perhaps, as we have seen, in the probably single succession of the semi-servile holder, or mere 'usufructuarius' under Roman law, and especially under the semi-military rule of the border provinces. [p394]

Lastly, before leaving the comparison between the yard-land and hub it may be asked why the serf who held it in England was called a Gebur.

The word villanus of the Domesday Survey is associated with other words, such as villicus, villata, villenage, all connected with serfdom, and all traceable through Romance dialects to the Roman 'villa.'

But the Anglo-Saxon word was 'Gebur.' It was the Geburs who were holders of yard-lands.

We trace this word Gebur in High German dialects. We find it in use in the High German translation of the laws of the Alamanni, called the 'Speculi Suevici,' where free men are divided into three classes:—

(1) The 'semperfrien' = lords with vassals under them.

(2) The 'mittlerfrien' = the men or vassals of the lords.

(3) The 'geburen' = liberi incolæ, or 'fri-lant-sæzzen' [i.e. not slaves].611

The word 'gebur' or 'gipur' occurs also in the High German of Otfried's 'Paraphrase of the Gospels,' 612 of the ninth century, and in the Alamannic dialect of Notger's Psalms for vicinus.613

Here, again, the South German connexion seems to be the nearest to the Anglo-Saxon. [p395]

VI. THE HIDE, THE HOF, AND THE CENTURIA.

From the yard-land, or hub, the holding of a serf, we may pass to the typical holding of the full free landholder, connected in England with the full team of eight oxen.

The Saxon hide, or the familia of Bede, was Latinised in Saxon charters into 'casatum.' We have found in the St. Gall charters the word 'casa' used for the homestead. The present Romanish word for house is 'casa,' and for the verb 'to dwell,' 'casar.' And there is the Italian word 'casata,' still meaning a family. Thus the connexion between the 'familia' of Bede and the 'casatum' of the charters is natural. Bede wrote more classical Latin than the ecclesiastical scribes in the charters. The hide was the holding of a family.614 Hence it was sometimes, like the yard-land or holding of a servile family, called a 'hiwisc,' which was Anglo-Saxon, and also High German for family.615 But the Saxon hide, also, was translated into ploughland or carucate, corresponding with the full team of eight oxen.

Generally in Kent, and sometimes in Sussex, Berks, and Essex, we found in addition to or instead of the hide or carucate, or 'terra unius aratri,' solins, sullungs, or swullungs—the land pertaining to a 'suhl,' the Anglo-Saxon word for plough. This word is [p396] surely of Roman rather than of German origin. The Piedmontese 'sloira,' and the Lombardic 'sciloira,' and the Old French 'silleoire,' are surely allied to the Romanish 'suilg,' and the Latin 'sulcus.'

Again, in Kent the quarter of a 'sulung' (answering to the yard-land or virgate of other parts) is called in the early charters a 'gioc,' 'ioclet,' or 'iochlet,' 616 i.e. a yoke or small-yoke of land. We have seen in the St. Gall charters, also, mention of 'juchs' or 'jochs,' which, however, were apparently jugera. This word gioc is surely allied to the Italian 'giogo,' and the Latin jugum.

Here, then, we have the hide the typical holding of a free family, as the centuria was under Roman law. A free Saxon thane might hold many hides, and so might and did the lord of a Roman villa hold more than one 'centuria' within its bounds. Still Columella took as his type of a Roman farm the 'centuria' of 200 acres,617 and calculated how much seed, how many oxen, how many opera, or day-works of slaves, or 'coloni' were required to till it. The hide, double or single, was also a land measure, and contained eight or four yard-lands, and so also was the 'centuria' a land measure divisible into eight normal holdings allotted with single yokes. Both also became, as we have seen, units of assessment. But in England the hide was the unit. Under the Roman system of taxation the jugum was the unit. [p397]

This variation, however, confirms the connexion. The Roman jugum, or yoke of two oxen, made a complete plough. Nothing less than the hide was the complete holding in England, because a team of eight oxen was required for English ploughing. The yard-land was only a fractional holding, incomplete for purposes of ploughing without co-operation. Hence it would seem that the complete plough was really the unit in both cases.

How closely the English hidation followed the lines of the Roman 'jugatio' has already been seen. When to the many resemblances of the hide to the 'centuria,' and of the 'jugum' to the virgate, regarded as units of assessment, are now added the other connecting links found in this chapter, in things, in figures, and in words, between the Saxon open-field system, and that of the districts of Upper Germany, so long under Roman rule, the English hidation may well be suspected to go back to Roman times, and to be possibly a survival of the Roman jugation. When Henry of Huntingdon, in describing the Domesday Survey, instead of saying that inquiry was made how many hides and how many virgates there were, uses the words 'quot jugata et quot virgata terræ,' 618 he at any rate used the exact words which describe what in the Codex Theodosianus is spoken of as taxation 'per jugationem.' 619

Not, as already said, that the Romans introduced into Britain the division of land according to plough teams, and the number of oxen contributed [p398] to the plough team. It would grow, as we have seen, naturally out of tribal arrangements whenever the tribes settled and became agricultural, instead of wandering about with their herds of cattle. It was found in Wales and Ireland and Scotland, in Bohemia, apparently in Slavonic districts also and further east.620 It is much more likely that the Romans, according to their usual custom, adopted a barbarian usage and seized upon an existing and obvious unit as the basis of provincial taxation.

The Frisian tribute of hides was perhaps an example of this. The Frisians were a pastoral people, and a hide for every so many oxen was as ready a mode of assessing the tribute as counting the plough teams would be in an agricultural district. The word 'hide,' which still baffles all attempts to explain its origin, may possibly have had reference to a similar tribute. Even in England it does not follow that it was in its origin connected with the plough team. Its real equivalent was the familia, or casatum—the land of a family—and in pastoral districts of England and Wales the Roman tribute may possibly have been, if not a hide from each plough team, a hide from every family holding cattle; just as in A.D. 1175 Henry II. bound his Irish vassal, Roderic O'Connor, to pay annually 'de singulis animalibus decimum corium placabile mercatoribus'—perhaps a tenth of the hides he himself received as tribute from his own tribesmen.621 The supposition of such an origin of the connexion of the word 'hide' with the 'land of a family' [p399] or of a plough team is mere conjecture; but the fact of the connexion is clear. All these three things, the hide, the hiwisce, and the sullung, and their subdivision the yard-land, were the units of British 'hidation,' just as the centuria and the jugum were the units of the Roman 'jugatio.'

VII. THE GAFOL AND GAFOL-YRTH.

Passing now to the serfdom and the services under which the 'yard-lands' and the 'huben' were held, it may at least be said that their practical identity suggests a common origin.

We learned from the Rectitudines and from the Laws of Ine, to make a distinction between the two component parts of the obligations of the 'gebur' in respect of his yard-land.

There was (1) the gafol, and (2) the week-work.

The gafol was found to be a semi-servile incident to the yard-land. The week-work was the most servile one.

A man otherwise free and possessing a homestead already, could, under the laws of Ine, hire a yard-land of demesne land and pay gafol for it, without incurring liability to week-work. But if the lord found for him both the yard-land and the homestead, then he was a complete 'gebur' or 'villanus,' and must do week-work also.

Taking the gafol first, and descending to details, it was found to be complex—i.e. it included gafol and gafol-yrth. [p400]

The gafol of the 'gebur,' as stated in the Rectitudines, was this:—

• For gafol proper:—
  • 10d. at Michaelmas.
  • 23 sesters of beer, and 2 fowls, at Martinmas.
  •  1 lamb at Easter, or 2d.
• For gafolyrth:—the ploughing of 3 acres, and sowing of it from the 'gebur's' own barn.

Comparing the gafol proper with the census of the St. Gall charters, and the tribute of the 'servi' of the Church under the Alamannic laws of A.D. 622, the resemblance was found to be remarkably close.

The tribute of the 'servi' of the Church was thus stated in the latter:—

• 15 siclæ of beer.
• A sound spring pig.
• 2 modia of bread.
• 5 fowls.
• 20 eggs.

As regards this tribute in kind the likeness is obvious, and it further so closely resembles the food-rent of the Welsh free tribesmen as to suggest that it may have been a survival of ancient tribal dues—a suggestion which the word 'gafol' itself confirms. It seems to be connected with the Abgabe, or food gifts of the German tribesmen.622

We saw that the word gafol was the equivalent of tributum in the Saxon translation of the Gospels. 'Does your master pay tribute?' 'Gylt he gafol?'

Further, the French evidence seems to show [p401] that the later manorial payments in kind and services upon Frankish manors were, to some extent, a survival of the old Roman exactions in Gaul.623 And the tribute of the Alamannic and Bavarian laws, and of the St. Gall and other charters, was found to be equally clearly a survival of the Roman tributum in the German province of Rhætia and the 'Agri Decumates.'

But in addition to the 'gafol' in kind, there was the gafol-yrth; and of this also we found in the St. Gall charters numerous examples. In the many cases where the owner of homesteads and land surrendered them to the Abbey, and henceforth paid tribute to the Abbey, there was not only the tribute in kind, but also the ploughing of so many acres, sometimes of one, sometimes of two, and sometimes of one in each zelga or field—to be ploughed, and reaped, and carried by the tenant. The combination of the dues in kind and in ploughing, with sometimes other services, made up the tributum in servitium—i.e. the gafol of the tributarius, or 'gafol-gelder,' which he paid under the Alamannic laws to his lord, the latter thenceforth paying the public tributum for the land to the State.

Perhaps we may go one step further.

From the remarkable resemblance of the English gafol-yrth and its South German equivalent the inference was drawn that this peculiar rent taken in the form of the ploughing of a definite number of acres, was probably a survival of the Roman tenths, [p402] or other proportion of produce claimed as rent from settlers on the ager publicus of the 'Agri Decumates,' and of Rhætia. Indications were found that the agrarium, or tenth of the arable produce, may have been taken in actual acres like the Saxon tithes—i.e. in the produce of so many 'andecenæ,' the ploughing, sowing, reaping, and garnering of which were done by the tenant.

But under Roman usage the proportion taken was not always a tenth. The State rent was nominally a tithe. But it was in fact so extortionately gathered as sometimes in Sicily to treble the tithe.624 Hyginus also says that the 'vectigal,' or tax, was taken in some provinces in a certain part of the crop, in some a fifth, in others a seventh.625 In Italy the dues from the Agri Medietates perhaps surviving in the later métayer system, amounted sometimes to one-half. At any rate, the proportion varied.

Now the Saxon 'gafol-yrth' of the yard-land of thirty acres seems, according to the 'Rectitudines,' as we have seen, to have been the produce of three acres in the wheat-field, ploughed by the 'gebur' and sown with seed from his own barn. For it will be remembered that the first season after the yard-land was given there was to be no gafol, and in the gebur's outfit only seven out of the ten acres in the wheat-field [p403] were to be handed over to him already sown, leaving three unsown, i.e. probably the three which otherwise he must have sown for the gafol-yrth due to his lord. As ten acres of the yard-land were probably always in fallow, three acres of wheat was a heavier gafol-yrth than a fairly gathered tithe would have been.

It would therefore seem probable that as the 'gafol' in kind may be traced back to the Roman tributum, itself perhaps a survival of the tribal food-rents of the conquered provinces, so the 'gafol-yrth' may be traced back to the Roman decumæ, or other proportion of the crop due by way of land-tax or rent to the State. And this survival of the complex tribute or gafol, made up of its two separate elements, from Roman to Saxon times, becomes all the more striking when it is considered also that it was due from a normal holding with an outfit of a pair of oxen, both in the case of the Saxon yard-land and of the Roman veteran's allotment.

VIII. THE BOON-WORK AND WEEK-WORK OF THE SERF.

Proceeding still further, besides the gafol and gafol-yrth, and yet distinct from the week-work, was the liability of the serfs on the Saxon manor to certain boon-work or services ad preces; sometimes in ploughing or reaping a certain number of acres of the lord's demesne land in return for grass land or other advantages, or without any special equivalent; sometimes in going errands or carrying goods to market or otherwise, generally known as averagium. 'He shall land-gafol pay, and shall ridan and averian [p404] and lade lædan' for his lord. So this boon-work in addition to 'gafol' is described in the 'Rectitudines.'

The various kinds of manorial 'averagium' were, as we have seen, often called in mediæval Latin angariæ, a going on errands or postal service; paraveredi, or packhorse services; and carroperæ, or waggon services.

We have seen how these services resembled the angariæ and the parangariæ and paraveredi, which were included among the 'sordida munera' or 'obsequiæ' of the Theodosian Code in force in Rhætia in the fourth century, found still surviving, though transformed into manorial services, in the same districts in the seventh century and afterwards, under the Bavarian laws and in the monastic charters. The carrying services and other boon-work on Saxon manors closely resembled those of the Frankish charters and the Bavarian laws, and probably therefore shared their Roman origin.

There remains to complete the serfdom its most servile incident, the week-work—that survival of the originally unrestricted claim of the lord of the Roman villa to his slave's labour which, limited, as we have seen, according to the evidence of the Alamannic laws, under the influence of Christian humanity by the monks or clergy, in respect of the servi on their estates, to three days a week, became the mediæval triduanum servitium. The words of the Alamannic law are worth re-quoting.

This remarkable passage in the Alamannic code of A.D. 622 seems to be the earliest version extant of the Magna Charta of the agricultural servus, who thus early upon ecclesiastical estates was transformed from a slave into a serf.

IX. THE CREATION OF SERFS AND THE GROWTH OF SERFDOM.

There is yet another point in which the correspondence between British and Continental usages is worth remarking.

The community in serfdom on a lord's estate was both by Saxon and Continental usage recruited from above and from below.

Free men from above, by voluntary arrangement with a lord, could and did descend into serfdom. The Saxon free tenant could, by free contract, arrange to take a yard-land, and if he were already provided with a homestead and oxen, he became a 'gafol-gelder,' or tributarius of his lord, without incurring the liability to the more servile 'week-work,' just as was the case when, under the Alamannic laws, free men made surrender of their holdings to the Abbey of St. Gall. In both cases, as we saw, week-work was added if the lord found the homestead and the outfit.

On the other hand, whenever a lord provided his slave with an outfit of oxen, and gave him a part in the ploughing, he rose out of slavery into serfdom. To speak more correctly, he rose into that middle class of tenants who, by whatever name they were [p406] known at first, afterwards became confounded together in the ranks of mediæval serfdom.

There were, in fact, grades in the community in serfdom not only like those of the Saxon geburs and cottiers, but also corresponding to the historical origin of the serfs. Thus, as we have seen in the 'Polyptique d'Irminon' and in many other cartularies and surveys of monastic estates, there are coloni and liti among the serfs, names bearing witness to the historical origin of the serfs, though the difference between them had all but vanished.

There is a passage in the Ripuarian laws, 'If any one shall make his slave into a "tributarius," or a "litus," &c.' 626 The 'lidus' of the 'Lex Salica' was under a lordship, and classed with 'servi,' and by a legal process he could be set free.627 We have noticed the passage in the Theodosian Code which speaks of 'coloni' and 'tributarii' on British estates, and also the mention by Ammianus Marcellinus of 'tributarii' in Britain. We have noticed also the three grades of 'læts,' the only class of tenants mentioned in the laws of Ethelbert.

Now, whatever doubt there might be as to what were the 'læts' on Kentish 'hams' and 'tuns' in the sixth century, if they stood alone as isolated phenomena; taken together with the 'tributarii' and 'coloni' and 'liti' on Continental manors, there can be hardly any doubt that they belonged to the same middle [p407] class of semi-servile tenants to which allusion has been made. Their presence on the manorial 'hams' and 'tuns' of England revealed in the earliest historical record after the Saxon Conquest, taken in connexion with the many other points brought together in this chapter, makes the inference very strong indeed that they, like the 'coloni,' 'tributarii,' and 'liti' on Continental manors, were a survival from that period of transition from Roman to German rule, during which the names of the various classes of semi-servile tenants, afterwards merged in the common status of mediæval serfdom, still preserved traces of their origin.

X. THE CONFUSION IN THE STATUS OF THE TENANTS ON ENGLISH AND GERMAN MANORS.

In one sense both in England and Germany the holders of the 'yard-lands' and 'huben,' though serfs, were free. As regards their lords they were serfs. As regards the slaves they were free. In this respect they resembled very closely the Roman 'coloni' on a private villa.

On the Frankish manors there were two classes of these semi-servile tenants—'mansi ingenuiles,' who were free from the 'week-work;' and 'mansi serviles,' from whom 'week-work' was due. Probably owing to the nature of the Saxon conquest the first of these classes seems to have practically become absorbed in the other. The laws of Ine, indeed, mention the gafol-gelder who, providing his own homestead, did not become liable to 'week-work' like the 'gebur.' But [p408] in the statements of the services on the manors of Hisseburne and Tidenham no such class appears. In the 'Rectitudines' there is no class mentioned between the thane, who is lord of the manor, and the 'geneats'—i.e. the 'gebur' and the 'cotsetl.' In the Domesday Survey there are no tenants above the villani, as a general rule, except in the Danish districts, where the 'Sochmanni' and the 'liberi homines' appear.

Comparing the status of English and German holders of 'yard-lands' and 'huben,' the resemblances are remarkable, and they confirm the suggestion of a common origin. Both are 'adscripti glebæ.' In both cases there is the absence of division among heirs. In both the succession is single, and in theory at the will of the lord. In both there are the gafol and customary services.

In both cases there is the distinction in grade of serfdom between the man who freely becomes the holder of a yard-land or hub by his own surrender, or by voluntary submission to the semi-servile tenure, and the man who is a nativus or born serf.

In both cases there is a regular contribution towards military service or the equipment of a soldier, and apparently no bar in status from actual service, though doubtless in a semi-menial position.

In all these points we have noticed strong analogies between the semi-free and semi-servile conditions of the various classes of tenants on Roman villas, and on the Roman public lands, which we have spoken of as the great provincial manor of the Roman Empire. And the natural inference seems to be, that even the curious confusion of the free and servile status may [p409] be, in part, a survival of the like confusion in the Roman provinces. It naturally grew up under the semi-military rule of the German provinces, and possibly in Britain also; whilst the Saxon conquest of the latter, no doubt, as we have said, tended to reduce the confusion into something like simplicity by fusing together classes of semi-servile tenants of various historical origins, in the one common class of the later 'geneats' or 'villani,' in whose status the old confusion, however, survived.

XI. RESULT OF THE COMPARISON.

To sum up the result of the comparison made in this chapter between the English and the Continental open-field system and serfdom. The English and South-German systems at the time of the earliest records in the seventh century were to all intents and purposes apparently identical.

The mediæval serf, judging from the evidence of his gafol and services, seems to have been the compound product of survivals from three separate ancient conditions, gradually, during Roman provincial rule and under the influence of barbarian conquest, confused and blended into one, viz. those of the slave on the Roman villa, of the colonus or other semi-servile and mostly barbarian tenants on the Roman villa or public lands, and of the slave of the German tribesman, who to the eyes of Tacitus was so very much like a Roman colonus.

That peculiar form of the open-field system, which was the shell of serfdom both in England and on the Continent, also connects itself in Germany [p410] distinctly with the Romano-German provinces, whilst at the same time conspicuously absent from the less Romanised districts of Northern Germany.

It seems therefore inconceivable that the three-field system and the serfdom of early Anglo-Saxon records can have been an altogether new importation from North Germany, where it did not exist, into Britain, where it probably had long existed under Roman rule.

We have already quoted the strong conclusion of Hanssen that the Anglo-Saxon invaders and their Frisian Low-German and Jutish companions could not introduce into England a system to which they were not accustomed at home. It must be admitted that the conspicuous absence of the three-field system from the North of Germany does not, however, absolutely dispose of the possibility that the system was imported into England from those districts of Middle Germany reaching from Westphalia to Thuringia, where the system undoubtedly existed. It is at least possible that the invaders of England may have proceeded from thence rather than as commonly supposed, from the regions on the northern coast. But if it be possible that a system of agriculture implying long-continued settlement, and containing within it numerous survivals of Roman elements, could be imported by pirates and the emigrants following in their wake, the possibility itself implies that the immigrants had themselves previously submitted to long-continued Roman influences.

On the whole we may adopt as a more likely theory the further suggestion of Hanssen, that if the three-field system was imported at all into England, [p411] the most likely time for its importation was that same period of Roman occupation during which he considers that it came into use in the Roman provinces of Germany.628

Nor is there anything inconsistent with this suggestion in the irregular lines of the English open fields and their divisions, so different from those produced by the rectangular centuriation of Roman 'Agrimensores.' We must not forget that the open field system in its simpler forms was almost certainly pre-Roman in Britain as elsewhere; so that what the Romans added to transform it into the manorial three-field system probably was rather the three-course rotation of crops, the strengthening of the manorial element on British estates, and the methods of taxation by 'jugation,' than any radical alteration in the land-divisions or in the system of co-operative ploughing.629