But how was it 150 years later, when Tacitus wrote his celebrated description of the Germans of his time?

The 'Germania' was obviously written from a distinctly Roman point of view.

The eye of the writer was struck with those points chiefly in which German and Roman manners differed. The Romans of the well-to-do classes lived in cities. City life was their usual life, and those of them who had villas in the country, whilst sometimes having residences for themselves upon them, as we have seen, cultivated them most often by means of slave-labour under a villicus, but sometimes by coloni.

What struck Tacitus in the economy of the Germans (and by Germans he obviously meant the free tribesmen, not their slaves) was that they did not live in cities like the Romans. 'They dwell' (he says) 'apart and scattered, as spring, or plain, or grove attracted their fancy.' 514 Of whom is he speaking? Obviously of free tribesmen or tribal households, not of villagers or village communities, for he [p339] immediately afterwards, in the very next sentence, speaks of the Germans as avoiding even in their villages (vici) what seemed to him to be obviously the best mode of building, viz. in streets with continuous roofs. 'Their villages' (he says) 'they build not in our manner with connected and attached buildings. There is an open space round every one's house.' And this he attributes not to their fancy for one situation or another, as in the first case, but 'either to fear of fire or ignorance of how to build.' 515

It is obvious, therefore, that the Germans who chose to live scattered about the country sides, as spring, plain or grove attracted them, were not the villagers who had spaces round their houses. We are left to conclude that the first class were the chiefs and free tribesmen, who, now having become settled for a time, were, in a very loose sense, the landowners, while the latter, the villagers, must chiefly have been their servile dependants. And this inference is confirmed when Tacitus comes to the second point and tells us that the servi of the Germans differed greatly from those of the Romans. There were some slaves bought and sold in the market, and free men sometimes sank into slavery as the result of war or gambling ventures; but in a general way (he says) their slaves were not included in the tribesmen's households or employed in household service, but each family of slaves had a separate [p340] homestead.516 They had also separate crops and cattle; for 'the lord (dominus) requires from the slave a certain quantity of corn, cattle, or material for clothing, as in the case of coloni. To this modified extent (Tacitus says) the German servus is a slave. The wife and children of the free tribesman do the household work of his house, not slaves as in the Roman households.'

Clearly, then, the vicus—the village—on the land of the tribesman who was their lord, was inhabited by these servi, who, like Roman coloni, had their own homesteads and cattle and crops, and rendered to their lord part of their produce by way of tribute or food-rent.

The lords—the tribesmen—themselves (as Tacitus elsewhere remarks) preferred fighting and hunting to agriculture, and left the management of the latter to the women and weaker members of the family.517

Now, if we could be sure that the tribal homestead was a permanent possession, and that the village of serfs around it had a single tribesman for its lord, the settlement would practically be to all intents and purposes a heim or manor with a village in serfdom upon it. It was evidently in a real sense the tribesman's separate possession, for, after speaking of blood relationships which bind the German tribesman's family and home most strongly together, Tacitus adds, 'Everyone's children are his heirs and successors [p341] without his making a will; and if there be no children, the grades of succession are brothers, paternal uncles, maternal uncles.' 518

But then this was also the case in Wales and Ireland. There was division among male heirs of the family land. And yet this family land was not a freehold permanent estate so long as a periodical redistribution of the tribe land might shift it over to someone else.

The embryo manor of the German tribesman, with its village of serfs upon it, might therefore, if the same practice prevailed, differ in three ways from the later manor. It might become the possession of a tribal household instead of a single lord; and also it possibly might, on a sudden redistribution of the tribal land, fall into the possession of another tribesman or tribal household, though perhaps this is not very likely often to have happened. Finally, it might become subdivided when the time came for the unity of the tribal household to be broken up as it was in Wales after the final redivision among second cousins.

It must be remembered that land in the tribal stages of economic progress was the least stable and the least regarded of possessions. A tribesman's property consisted of his cattle and his serfs. These were his permanent family wealth, and he was rich or poor as he had more or less of them. So long as the tribe land was plentiful, he as the head of a tribal household took his proper share according to tribal rank; and so long as periodical redistributions took place, even when the tribal household finally was [p342] broken up, room would be found for the new tribal households on the tribal land. But when at last the limits of the land became too narrow for the tribe, a portion of the tribesmen would swarm off to seek new homes in a new country. Frequent migrations were, therefore, at once the proofs of pressure of population and the safety-valve of the system.

The emigrating tribesmen in their new home would form themselves into a new sept or tribe, take possession of fresh tracts of unoccupied land, and perhaps, if land were plentiful, wander about for a time from place to place as pasture for their cattle might tempt them. Then at last they would settle: each tribesman would select his site by plain, wood or stream, as it pleased him. He would erect his stake and wattle tribal house, and daub it over with clay519 to keep out the weather. He would put up his rough outbuildings and fence in his corn and cattle yard. Round this tribal homestead the still rougher homesteads of his serfs, each with its yard around it, would soon form a straggling village, and the likeness to the embryo manor would once more appear.

Indeed, when we turn to the famous passage in which the German settlements and their internal economy are described, the words used by Tacitus seem in themselves to indicate that he had in his eye precisely this process which the example of the Welsh and Irish tribal systems has helped to make intelligible to us. Tracts of country (agri), he says, are 'taken possession of' (occupantur) by a body of tribesmen (ab universis) who are apparently seeking new [p343] homes; and then the agri are presently divided among them.

This passage, so often and so variously construed and interpreted, is as follows:—

It is unfortunate that the first few lines of this passage are made ambiguous by an error in the texts. If the true reading be, as many modern German critics now hold, 'ab universis vicis'—by all the vici together, or by the whole community in vici—there still must remain the doubt whether the word vicus should not be considered rather as the equivalent of the Welsh trev than of the modern village. The Welsh 'trev' was, as we have seen, a subordinate cluster of scattered households. Tacitus himself probably uses the word in this sense in the passage where he describes the choice of the chiefs, or head men (principes) 'qui jura per pagos vicosque reddunt.' 522 The vicus is here evidently a smaller tribal subdivision of the pagus, just as the Welsh trev was of the 'cymwd,' and not necessarily a village in the modern sense.523 [p344]

If, on the other hand, the true reading be 'ab universis in,' or 'per, vices' or 'invicem,' the meaning probably is that fresh tracts of land (agri) are one after another taken possession of by the tribal community when it moves to a new district or requires more room as its numbers increase.

The new agri, the passage goes on to say, are soon divided among the tribesmen or the trevs, 'secundum dignationem,' according to the tribal rules, the great extent of the open country and absence of limits making the division easy, just as it was in the instance of Abraham and Lot.

In any case it is impossible to suppose that Tacitus meant by the words in vices or invicem, if he used them, that there was any annual shifting of the tribe from one locality to another, for it is obvious that the very next words absolutely exclude the possibility of an annual movement such as that described by Cæsar. 'Arva per annos mutant et superest ager.' They change their arva or ploughed land yearly, i.e., they plough up fresh portions of the ager or grass land every year, and there is always plenty left over which has never been ploughed.524 Nothing could describe more clearly what is mentioned in the Welsh triads as 'co-aration of the waste.' The tribesmen have their scattered homesteads surrounded by the lesser homesteads of their 'servi.' And the latter join in the co-tillage of such part of the grass land as year by year is chosen for the corn crops, while the cattle wander over the rest. [p345]

This seems to have been the simple form of the open field husbandry of the Germans of Tacitus.

And this is sufficient for the present purpose; for whichever way this passage be read, it does not modify the force of the previous passages, which show how manorial were the lines upon which the German tribal system was moving even in this early and still tribal stage of its economic development, owing chiefly to the possession of serfs by the tribesmen. It gives us further a clear landmark as regards the use by the Germans of the open-field system of ploughing. Tacitus describes a husbandry in the stage of 'co-aration of the waste.' It has not yet developed into a fixed three-course rotation of crops, pursued over and over again permanently on the same arable area, as in 'the three-field system' afterwards so prevalent in Germany and England.

These are important points to have gained, but the most important one is that, notwithstanding the strong resemblances between the Welsh and German tribal arrangements, there was this distinct difference between them. The two tribal systems were not working themselves out, so to speak, on the same lines. The Welsh system, in its economic development, was not directly approaching the manorial arrangement except perhaps on the mensal land of the chiefs. The Welsh tribesmen had as a rule no servile tenants under them. The taeogs were mostly the taeogs of the chiefs, not of the tribesmen. Thus, as we have seen, when the conquest of Wales was completed, the tribesmen of the till then unconquered districts became freeholders under the Prince of Wales, and with no mesne lord over them. The taeogs [p346] became taeogs of the Prince of Wales and not of local landowners. So that the manor did not arise. But even in the time of Tacitus the German tribesmen seem to have already become practically manorial lords over their own servi, who were already so nearly in the position of serfs on their estates that Tacitus described them as 'like coloni.'

The manor—in embryo—was, in fact, already in course of development. The German economic system was, to say the very least, working itself out on lines so nearly parallel to those of the Roman manorial system that we cannot wonder at the silent ease with which before and after the conquest of Roman provinces, German chieftains became lords of villas and manors. The two systems, Roman and German, may well have easily combined in producing the later manorial system which grew up in the Roman provinces of Gaul and the two Germanies.

II. THE TRIBAL HOUSEHOLDS OF GERMAN SETTLERS.

Now, if we were to rely upon this evidence of Tacitus alone, the conclusion would be inevitable that the German and Roman land-systems were so nearly alike in their tendencies that they naturally and simply joined in producing the manorial system of later times. And there can be little doubt that, speaking broadly, this would be a substantially correct statement of the case.

But before we can fairly and finally accept it as such, it is necessary to consider another branch of evidence which has sometimes been understood to point to a kind of settlement not manorial. [p347]

The evidence alluded to is that of local names ending in the remarkable suffix ing or ingas. It is needful to examine this evidence, notwithstanding its difficult and doubtful nature. It raises a question upon which the last word has by no means yet been spoken, and out of which interesting and important results may eventually spring. The impossibility of arriving, in the present state of the evidence, at a positive conclusion, is no reason why its apparent bearing should not be stated, provided that suggestion and hypothesis be not confounded with verified fact. At all events, the inquiry pursued in this essay would be open to the charge of being one-sided if it were not alluded to.

The reader of recent literature bearing upon the history of the English conquest of Britain will have been struck by the confidence and skill with which, in the absence of historical, or even, in some cases, traditional evidence, the story of the invasion and occupation of England has been sometimes created out of little more than the combination of physical geography with local names, on the hypothesis that local names ending in 'ing,' or its plural form 'ingas,' represent the original clan settlements of the German conquerors. Writers who rely upon G. L. Von Maurer's theory of the German mark-system have also naturally called attention to local names with this suffix as evidence of settlements on the basis of the free village community as opposed to those of a manorial type.

Local names with this suffix, it is hardly needful to say, are found on the Continent as well as in England. [p348]

How, it may well be asked, does the evidence they afford of clan settlements or free village communities comport with the thoroughly manorial character of the German settlements on the lines described by Tacitus?

Now, in order to answer this question, it must first be considered how far the description of Tacitus covers the whole field—whether it refers to the Germans as a whole, or whether only to those tribes who had come within Roman influences, and so had sooner, perhaps, than the rest, relinquished their earlier tribal habits to follow manorial lines.

So far as his description is geographical it is very methodical.

(1) There are the Germans within the Roman limes.525 These included the tribes who, following up the conquests of Ariovistus, had settled on the left bank of the Rhine in what was then called the province of Upper Germany, including the present Elsass and the country round the confluence of the Rhine with the Maine and Moselle. These tribes were the Tribocci, Nemetes and Vangiones.526 Further, there were the tribes or emigrants, many of them German, gradually settling within the limits of the 'Agri Decumates.' Lastly, there were the Batavi and other tribes settled in the province of Lower Germany at the mouths of the Rhine, shading off into Belgic Gaul.

(2) There were the Northern tribes outside the Roman province,527 some of them tributary to the [p349] Romans and some of them hostile, the Frisii, the Chatti (or Hessians), and other tribes, reaching from the German Ocean to the mountains, and occupying the country embracing the upper valleys of the Weser and the Elbe, some of which tribes afterwards joined the Franks and Saxons.

(3) There were the Suevic tribes528 so familiar to Cæsar, and amongst whom were the Angli and Varini, the Marcomanni and Hermunduri, always hovering over the limes of the provinces from the Rhine and Maine to the Danube: some of them hostile and some of them friendly; some of whom afterwards mingled with the Franks and Saxons, but most of whom were absorbed in the Alamannic and the Bavarian tribes who finally, following the course of the previous emigration, passed over the limes and settled within the 'Agri Decumates' in Rhætia, and in the Roman province of Upper Germany.

(4) Behind all these tribes with whom the Romans came in contact were others vaguely described as lying far away to the north and east.

The habits of which of these widely different classes of German tribes did Tacitus describe?

Probably it would not be safe to go further than to say that the Germans whose manners he was most likely to describe were those chiefly Suevic tribes hovering round the limes of the provinces, especially of the 'Agri Decumates,' with whom the Romans had most to do. It is at least possible that he left out of his picture, on the one hand, those distant northern or eastern tribes who may still have retained their early nomadic habits, and on the other hand those [p350] Germans who had silently and peaceably settled within the limes of the Roman provinces, and so had become half Roman.529

But to what class are we to refer the settlements represented by the local names with the supposed patronymic suffix?

The previous study of the Welsh and Irish tribal system ought to help us to judge what they were.

In the first place we have clearly learned that in tracing the connexion of the tribal system with local names, the fixing of a particular personal name to a locality implies settlement. It implies not only a departure from the old nomadic habits on the part of the whole tribe, but also the absence within the territory of the tribe of those redistributions of the tribesmen among the homesteads—the shifting of families from one homestead to another—which prevailed apparently in Wales and certainly in Ireland to so late a date.

Following the parallel experience of the Irish and Welsh tribal system we may certainly conclude that in the early semi-nomadic and shifting tribal stage described by Cæsar the names of places, like those of the Irish townlands, would follow local peculiarities of wood or stream or plain, and that not until there was a permanent settlement of particular families in fixed abodes could personal names attach themselves to places, or suffixes be used which in themselves involve the idea of a fixed abode.

Then with regard to the nature of the tribal settlements which these local names with a patronymic [p351] suffix may represent, surely the actual evidence of the Welsh laws and the 'Record of Carnarvon,' as to what a tribal household was, must be far more likely to guide us to the truth than any theoretical view of the 'village community' under the German mark-system, or even actual examples of village communities existing under complex and totally different circumstances at the present time, valuable as such examples may be as evidence of how the descendants of tribesmen comport themselves after perhaps centuries of settlement on the same ground.

Now we have seen that the tribal household in Wales was the joint holding of the heirs of a common ancestor from the great-grandfather downwards, with redistributions within it to make equality, first between brothers, then between cousins, and finally between second cousins; the youngest son always retaining the original homestead in these divisions. The Weles, Gwelys, and Gavells of the 'Record of Carnarvon' were late examples of such holdings. They were named after the common ancestor and occupied by his heirs. Such holdings, so soon as there was fixed settlement in the homesteads, were obviously in the economic stage in which, according to German usage, the name of the original holders with the patronymic suffix might well become permanently attached to them.530

We may then, following the Welsh example, fairly expect the distinctive marks of the tribal household to be joint holding for two or three generations, and then the ultimate division of the holding among male heirs, the youngest retaining the original ancestral homestead. [p352]

We know how persistently the division among male heirs was adhered to in Wales and in Ireland under the custom of Gavelkind,531 though of the peculiar right of the youngest son to the original homestead we have no clear trace in Ireland. Possibly St. Patrick was strong enough to reverse in this instance a strong tribal custom. But in Wales the succession of the youngest was, as we have seen, so deeply ingrained in the habits of the people that it was observed even among the taeogs. The elder sons received tyddyns of their own in the taeog trev in their father's lifetime, whilst the youngest son remained in his father's tyddyn, and on his death succeeded to it.

The persistence in division among heirs and the right of the youngest were very likely therefore to linger as survivals of the tribal household.

Now it is well known that in the south-east of England, and especially in Kent, the custom of Gavelkind has continued to the present day, retaining the division among male heirs and historical traces of the right of the youngest son to the original homestead. In other districts of England and in many parts of Europe and Asia the division among heirs has passed away, but the right of the youngest—Jüngsten-Recht—has survived.

Mr. Elton, in his 'Origins of English History,' has carefully described the geographical distribution in Western Europe of the practice, not so much of division among heirs, as of the right of the youngest to [p353] inherit the original homestead, the latter having survived in many districts where the other has not.

In England he finds the right of the youngest most prevalent in the south-east counties—in Kent, Sussex, and Surrey, in a ring of manors round London, and to a less extent in Essex and the East Anglian kingdom,—i.e. as Mr. Elton describes it, in a district about co-extensive with what in Roman times was known as the Saxon shore. A few examples occur in Hampshire, and there is a wide district where the right of the youngest survives in Somersetshire, which formed for so long a part of what the Saxons called 'Wealcyn.' 532

Further, as the custom is found to apply to copyhold or semi-servile holdings, it would not be an impossible conjecture that previously existing original tribal households were, at some period, upon conquest, reduced into serfs, the division of the holdings among heirs being at the same time stopped, so as to keep the holdings in equal 'yokes,' or 'yard-lands,' thus leaving the right of the youngest as the only point of the pre-existing tribal custom permitted to survive.

A similar process, perhaps in connexion with the Frankish conquest of parts of Germany, possibly had been gone through in many continental districts. Mr. Elton traces the right of the youngest in the north-east corner of France and in Brabant, in Friesland, in Westphalia, in Silesia, in Wirtemberg, in the Odenwald and district north of Lake Constance, in Suabia, in Elsass, in the Grisons. It is found also in [p354] the island of Borneholm, though it seems to be absent in Denmark and on the Scandinavian mainland.533

Attention has been called to this curious survival of the right of the youngest because it forms a possible link between the Welsh, English, and continental systems of settlements in tribal households.