|p ^llf»l l ,i|B| l ,IM l l ^:^^^^Sf^^s^ 1 1 L i )jii' 8 ag ' aMi>W"n>tu«iW'JJii"tt"" %'U4NJ>' jr T ■f r^*^,.iO T} .«r4 J '♦ 4^.Jf- ^C^^^^^i^^jg-^ cy5t»iTfir^Tr''l -.^-^TttsTs^t^s m- HD 117 SI Cornell XDlniversit^ OF THE 1RewJ?ork State CoUeoe of aariculture ig,5.i?.l. .\cmii. 3778 HD 117.S4°'"^""""''^««y Library '^SiS''V acres £ Cornell University Library The original of tiiis book is in tine Cornell University Library. There are no known copyright restrictions in the United States on the use of the text. http://www.archive.org/details/cu31924013859479 CUSTOMARY ACRES AND THEIR HISTORICAL IMPOETANCE BY THE SAME AUTHOR. THE ENGLISH TILLAGE COMMUNITY. With 13 Maps and Plates. 8vo, 4s. 6(2. net. THE OXFORD EEFOEMERS-JOHN COLET, ERASMUS, AND THOMAS MORE : a, History of their Fellow-Work. 8vo, 4s. 6d. net • THE ERA OF THE PROTESTANT REVOLU- TION. With 4 Maps and 12 Diagrams. Fop. 8vo. 2s. 6d. (Epochs of Modern Hidtory.) LONGMANS, GRBBN & 00,, 39 Paternoster How, London : New Tork, Bombay, and Oalcutta. CUSTOMARY ACEES AND THEIR HISTORICAL IMPORTANCE BEING A SBEIES OP UNFINISHED ESSAYS BY THE LATE FREDERIC SEEBOHM HON. LL.D. (EDIN.) LITT.D. (CAMB.) D.LITT. (OXFOED). LONGMANS, GEEEN AND CO. 39 PATERNOSTER ROW, LONDON NEW YORK, BOMBAY AND CALCUTTA 1914 [All rights reserved] PREFACE. In publishing these unfinished studies of my father's it is my duty as well as my desire to state clearly, to anyone into whose hands this book may come, that I am responsible for its publication, not he. His standard of workmanship was so high that had he hved to go on with this work, its final form would have been very difierent. Eealising that he would not be able to complete his inquiry, he devoted his time — working sometimes even in his bed' — to grouping the subject into sections. He also drafted an introductory chapter to indicate its scope and to suggest the directions in which he thought it might be followed up. In all his work my father looked forward to the probability that other scholars with other evidence would amplify and modify his results, and with this in his mind he liked to consider his books as ' essays ' or ' contributions ' towards the solution of the questions he dealt with. In this particular work, he was conscious that his material had led him away from the original purpose he had set out with, into a far wider field of prehis- torical inquiry. But he had not time to remodel his notes to accord with a new point of view. Starting with his old theme of the Economic History vi Preface. of the Village Community, his idea was to investigate further the relation of the ' shell ' (as he called it), i.e. the actual division of the land on the ground, to the institutions that governed it. He had traced the Open Field System in England in its relation to the EngUsh Village Community. He found the same open fields still existing in France, obstructing indivi- dual enterprise and trammelling the development of agriculture. In continuing his long chain of studies he set himself to inquire whether a closer examina- tion of this shell, and of its modifications, would not throw light upon the one thing that interested him before all, viz. the gradual economic growth of the communities that inhabited it. Hence the dry bones of Metrology, the length and breadth of the acres on the surface of the ground, became ahve for him. But working at the customary acres, he found that what was conspicuous about them was not so much their relation to local customs as their relations to each other and to other land-measures in an ever-widening circle throughout Europe. Follow- ing this evidence he could not help feehng that, if there was anything at all in the remarkable results he obtained, they might have a value for students in a different field and might shed some side light upon the early movements of peoples in Western Eiirope. If it is true that in the seventeenth century, nearly all over England, the milestones of England still denoted the distances in the Old British Mile ; if this Old British Mile is the same as the Gallic Leuga which the Romans found existing . in Gaul when they fijst conquered it ; Preface. Vll if the relation of the fuxrows of the customary acres to the mile or itinerary measure is as ancient as it appears to be universal ; then we seem to be taken back 2000 years at a stride into an intricate network of intertribal or international relations. The acre, or unit of land measurement whatever it was called, was not merely a means of stating the area of a piece of land : it was a unit of cultivation and had its shape determined for it by the actual convenience of the plough. In wheat-growing plains where the eight-oxen plough was in use, a long furrow was a necessity, the width of the strip proportionately lessening. Further South and East where the single yoke of oxen was usual, the shape was often square or two squares side by side. In the typical village of Britain or Northern France the land lay round the village, as described in Piers the Plowman and as shown in the maps before the Enclosures (see map facing p. 123), divided into as convenient strips as possible. Just as a ploughman now will know very nearly how the land he is ploughing in a field divides into acres, so in the old open fields it would, be well known how far each strip corresponded to the tjrpical customary acre or a part of it. When the villagers came out to play on the stubbles, it was easy for them to choose an acre- strip to throw their ball across from balk to balk. The exact width of the statute acre — twenty-two yards — is now the cricket pitch. For practice with the long- bow, Henry VIII decreed that the shortest butts should be a furlong — eleven score — or exactly the acre's length. The stade was a furrow before it was viii Preface. a footrace; and the length of the furrow and back again still survives with us as a standard race in the quarter mile. The serious work of agriculture was the Ufe of the people, only interrupted by war. Caesar had to sub- ordinate his campaigns to the growing of corn, and Alfred found that his levies melted away from his standard whenever their fields called them. If the shell in which, year by year, this vital process took place can take us back 2000 years, to whom does it belong and at what period did it become crystalUsed ? These are questions rather for the ethnologist ; there is no attempt to deal with them in these pages. But they are questions that will arise in the mind of the student in proportion to the light he may be able to bring from other lines of study to the material here tentatively set before him. Although my father shrank from the thought of these his unfinished studies being pubhshed, he left it entirely to my discretion to deal with them. I believe it to be quite fair to his memory as well as in accordance with his wish that they should be put, just as they are, within reach of any student of History whose purpose they may serve. I am sure they will not be misjudged. HUGH E. SEEBOHM. PoYNDEKs End, HiTOHIN. CONTENTS. INTRODUOTOEY CHAPTER TO LIMIT THE SCOPE OF THE INQUIRY PART I. BRITISH AND IRISH EVIDENCE. FAQE 1 CHAPTER I. THE WELSH UNITS OF TRIBUTE AND FOOD-RENT. SBOT. I. THE TEBP AS THE UNIT OF TRIBUTE ... 15 II. THE LAND MEASUEES OF THE WELSH CODES . . 20 III. THE MILLTTE OE PAEASANG OF 3 GALLIC LEUG^ . 25 IV. THE ERWSj TEEPSj AND MAENOLS OP THE VENEDOTIAN CODE ........ 27 Y. THE EKW AND TEEF OF THE GWENTIAN CODE . . 33 VI. THE EEW AND TEEP OP THE DIMETIAN CODE . . 34 VII. GENEEAL EESULTS ...... 35 CHAPTER II. THE IRISH UNITS OF TRIBUTE AND FOOD-RENT. I. THE LAND MEASUEES OP THE BEEHON TEACTS II. THE SMALLER TIE-CUMAIL III. THE LAEGER TIE-CUMAIL IV. THE TEICHACED . . . • • V. CONCLUSION ..... 39 41 43 51 51 Contents. CHAPTER III. PAGE THE UNITS OF TBIBVTB IN OAELIO SCOTLAND 54 CHAPTER IV. THE UNITS OF TBI BUTE AND ASSESSMENT OF THE DOME SD AT SUBVET. SECT. I. THE COBNISH GREAT ' ACtEE ' OR CAEITCATE . . 64 II. THE HIDE ........ 66 CHAPTER V. GENEBAL CONCLUSION 68 PAET II. THE OLD BBITISH MILE 79 PART III. ON THE CUSTOMAET AOKES OF BRITAm, IRELAND, AND ARMORICA. CHAPTER I. INTBODUOTOBT 95 CHAPTER II. THE BBITISH GUSTOMABT ACSES. I. THE CUSTOMAET ACEES ...... 100 II. THE ENGLISH STATUTE ACRE ..... 103 III. THE COENISH AND DOKSET ACEES . . . 104 IV. THE FOREST ACRES ...... 106 T, THE POWTS ACEE ...... 107 Contents. XI BBUX. Pi^QE VI. THE NORTH WALES CUSTOMARY ACKE . . 108 VII. THE SCOTTISH^ NOETHUMBEIAN AND CUMBRIAN ACRES 109 VIII. THE HALF-ACRES IN FORM OF 1 X 10 . . . 112 IX. THE CONNECTION OF BRITISH CUSTOMARY ACRES WITH IRISH MEASURES . . . . . .115 CHAPTER III. TBE BRETON OPEN FIELD SYSTEM 117 CHAPTER IV. THE ACRES OF THE GORN-OROWING DISTRICTS OF FRANCE. I. THE BRETON ARPENT II. THE NORMANDY CUSTOMARY ACRE . in. THE 329 M. ACRE IN TWO HALF-ACRES IV. THE DORSET OR STADE ACRE . V. THE GROUP OF ACRES . 124 132 184 135 136 CHAPTER v.*-' THE ACRES TRACED INTO THE CORN-GROWING REGIONS AT THE MOUTHS OF THE PO AND THE DANUBE. I. THE RHENISH 1 X 10 ACRES AND THBIE GERMAN NEIGHBOURS ....... 138 II. THE LINK BETWEEN AEMORICAN 1 X 10 ACRES AND THOSE OF THE PO VALLEY AND THE DELTA OF THE DANUBE ........ 143 III. THE VENETIAN ACRES ...... 153 • IV. THE PLOUGH AND THE PLOUGH-TEAM OF EIGHT OXEN IN THE PO VALLEY .,,... 158 xii Contents. PART IV. THE CUSTOMARY ACRES OF NORTHERN AND EASTERN EUROPE. CHAPTER I. PAOE THE STUNDE OR PABA8ANG OF NORTHERN AND EASTERN EUROPE 165 CHAPTER II. THE WERST OF EASTERN EUROPE 170 CHAPTER III. THE AGRICULTURAL UNITS OF NORTHERN AND EASTERN EUROPE 173 CHAPTER IV. THE BALTIC REGION 178 CHAPTER V. THE ACRES OF THE LOW COUNTRIES 182 CHAPTER VI. THE EASTERN ORIGIN OF THE EUROPEAN PARASANGS 186 PART V. THE LAND UNITS OF THE MEDITERRANEAN BASIN. CHAPTER I. ■/HOMERIC PLOUGHING 193 CHAPTER II. THE EGYPTIAN LAND UNITS 203 Contents. xiii CHAPTER III. TRACES OF GREEK COLONISATION IN MAGNA GR^CIA^^'^^ AND SICILY 207 CHAPTER lY. THE AGRICULTURAL UNITS OF ITALY 214 CHAPTER V. THE OVERFLOW OF THE V0RSU8 INTO THE LIGVRIAN DISTRICT TO THE WEST OF THE MARITIME ALPS 228 CHAPTER VI. THE SPANISH AGRICULTURAL UNITS 233 CHAPTER VII. THE LAND UNITS OF THE DISTRICT BETWEEN THE LOIRE AND THE GARONNE 246 PART VI. THE PROBLEM OE THE BEITISH AND AEMOEICAN ACRES 251 INDEX .... 269 INTRODUCTORY CHAPTER TO LIMIT THE SCOPE OF THE INQUIRY. I AM fully aware of the dangers involved in an attempt to follow a line of historical inquiry on a single branch of a complicated problem into a, stratum of facts under- lying that of ordinary history. But ever since I came to realise the importance in Economic History of tribal landholding and its develop- ment into the open field system of the village community, I have felt more and more the necessity of a deeper study of both and of the transition from the one to the other, if we are to understand the line of economic development in Western Europe. I have been driven to recognise how the open field system of the village community acted as the shell under cover of which the common village life of an agricultural people was kept alive through the period of anarchy which followed upon the break-up of the Roman Empire, the almost breathless succession of Teutonic invasions and the internecine wars between the invading tribes. In the first volume of my series of economic studies I tried to point out how the open field system of the English village community was intimately connected with British Economic History and the growth of the manorial system during the later period. But it enters also (far more than I think has yet been acknowledged) B 2 2 Introduction. into the wider question of the economic development and evolution of human society in Europe. The secret of the economic strength of the open field system in the organised form taken by it in Britain is not far to seek. The holding of the individual peasant was so inter- twined and intermixed with those of his neighbours that he was incapable of independent action. As his yardland or virgate consisted normally of 30 acre strips or sometimes twice as many half-acre strips scattered over the common fields he could not cultivate his strips as he liked. He was bound to follow the course of husbandry prescribed by the community whether on the two-field or the three-field course of rotation of crops and fallow. Even when the corn crop was removed, the strips of the peasant were not at his own disposal. Everyone's strips all aUke on a given day became common pasture for the herds and flocks of the whole community under the care of the common herdsman or shepherd. But it was not a specially British institution. Practically the same advanced tjrpe of open field system was common to the corn-growing districts of Britain and Gaul. Indeed, to get a full and vivid sense of the economic and social power and grip of the system, the Enghsh student must go to the great corn-growing regions of France and look down upon its vast open fields from the Cathedral towers of Chartres or Amiens. Nor may one reasonably be afraid to recognise how deeply it entered into the tissue of domestic and com- munal fife — ^how the rehgious and poetic feeling of the people for ages, probably from the days of the Introdvxiion. 3 Druids, and certainly through the Christian centuries, were influenced by it. The temple and the church became visibly its centre. One may safely recognise with Ruskin how the sculptures on the great doorways of Amiens Cathedral indicate that it was built for an agricultural community. One may learn from Millet's ' Angelus ' what a part the church bell plays in the ecanomy of the open field, and see, too, in the ' Shepherdess ' of the same peasant-painter how solemn a responsibihty is involved in the task of leading the communal flock in the exercise of the vaine pdture by night as by day so as to secure that every one's strips get their fair share of the droppings of the sheep by which the stubble of the common field is manured. I have walked over the common fields both of Amiens and Chartres. On the latter a peasant was found planting vines on one of his strips near the town. When reminded of the vaine pdture and the injury it would do to his vines, he shrugged his shoulders. They must run the risk of that ! He dare not enclose his strip. Even to this day the individual peasant holder is bound by the strong grip of custom. To defy the immemorial rules of his commune might be to bring down his house over his head. Nor could one gain anywhere in Britain so vivid a picture of open field husbandry as one gets from the height above the old Roman town of Andernach, looking down upon the flat plain bounded by the wide sweep of the Rhine and stretching away into the far distance like a great map. Or going farther into the heart of Central Europe the same may be said of the view from the tower of Uhn Cathedral over the rich 4 Introduction. Bavarian corn-fields. Under modern legislation and industrial advance the rigid rules of the system may be gradually relaxed, but over a large part of Continental Europe in French, German, and Slavonic countries it is still more or less a living system, whilst in England we have to look back to the tithe maps of a century ago and those made for the very purpose of destroying it by enclosure. It is fast fading away out of memory. Still, even in England, the system has in many places left indehble marks on the ground which will remain for centuries longer, and perhaps for ever, to tell of its existence in the past. From the railroad between Cambridge and Oxford, passing through tracts of country now all but entirely devoted to pasture, the marks of the once arable strips remain as clearly defined as on a map. The ancient habit in some districts such as these, of ploughing each strip into what was called a high-backed land gave a permanence to its boundaries which nothing but a re-ploughing of the land can ever efface. Especially remarkable instances of this may be seen in the case of many of the manors around Worcester, which for more than a century have been altogether pasture and which yet retain the marks of the old ploughing and the division into strips as clearly defined as in the maps made before the enclosures of the eighteenth century, and still preserved in the Cathedral Library along with the manor rolls, going back sometimes to the thirteenth and fourteenth centuries. In these rolls are recorded the changes of the holders of the yardlands generation after generation. And the fact is unexpected and striking that in the schedules of holders attached to the eighteenth century maps, made just before the enclosure, the holdings are Introduction. 5 still called ' yardlands.' So that while these maps bear witness to the persistence of the holdings, in yard- lands, of scattered strips until the enclosure put a sudden end to the system, the marks of the strips them- selves will remain practically for ever to bear witness to future generations that the open field system once and for ages was the prevalent system of agriculture of the English village community under the ecclesiastical manorial lords of the Severn valley. But when we speak of the particular form of the open field system connected with the village community we must not forget that there were various other forms of it. In Britain, as already said, there have been all along two prominent forms of it. Side by side with the village system of parts of England mainly agricultural we have to recognise the tribal form of it fully in force in Wales under tenth century customary rules, described incident- ally rather than as newly enacted in the Venedotian Code. What they amount to is this. Though the tribesmen were mainly ^ engaged in pastoral life a certain amount of corn was a necessity even to them. So far as land was concerned their rights were family rights and sometimes joint and undivided shares with others in rights of pasture ; but the occupants of a pastoral community are represented as contributing oxen to a common plough-team and each year ploughing up by coramon consent enough for the coming year's crop. The common plough with its team of 8 oxen under this system of co-aration ploughed each year the required number of strips or erws, each strip being a day's work of the plough, a cyvar or co-ploughing, and the strips are described as taken by the individual contributors to the plough- team in the same order of rotation as their 6 Introduction. oxen might stand in the team. Finally, after the crop was gathered, the strips of the year's ploughing went back into the common pasture. Such is the description in the Venedotian Code of North Wales as prevalent in HowelFs time. Here obviously in this co-aration of the waste is an embryo form of the more advanced open field system of the settled agricultural village com- munity. It is only necessary to extend the corn crop over a wider area and to subject the strips to a perma- nent rotation of crops, and the result would be holdings with scattered and intermixed strips and the vaine fdture over the stubble. In Britain the 8-oxen plough- team, entering as it did through the carucate and the hide into the structures of the manorial system, forms a connecting link pointing strongly to the long- continued existence side by side of tribal and more or less settled agricultural life such as the testimony of Caesar, and of Pytheas, centuries ear her than Caesar, might well prepare us to expect. It is not difficult to conceive of the process from tribal and mainly pastoral life to that of the agricultural village as population increased and the single plough was no longer all that was required. But the transi- tion was not always so easily accomplished. The constitution of the tribal family unit of the gwely did not readily adapt itself to the system of regular holdings, in yardlands normally representing the contribution of the single yoke of 2 oxen to the plough-team of 8. The right of each tribesman to maintenance was not easily shifted from an allotment of cattle with common rights of grazing to a definite holding like the yardland. The evidence of both Irish and Welsh history show that some power from outside following upon conquest Introduction. 7 and followed by something like lordstiip was needful to consolidate tlie village community into the per- manent form in which we find it and which enabled it to hold its own apparently unchanged through countless generations. Now if the open field system with its network of scattered strips or acres under the grip of tribal or communal instinct was strong enough to preserve from destruction during ages of anarchy the village life of the agricultural population, let us say of Britain and Armorican Gaul, what shall be said of the customary acres of which, so to speak, the intermixed network of strips consisted ? In spite of their division only by unploughed turf balks, must they not have shared in the permanence resulting from the same tenacity of custom which kept the system itself alive ? Shall we not learn to trust their evidence also when we find them holding their own from a far back traditional origin lost in the past ? There is no doubt that the lie of the land and the character of the soil must have had something to do with their form and size — in any particular field. But if it can be shown that the British customary acres, which in spite of legislation have fin- gered in particular districts on this side of the Channel, form a normal group coinciding with a corresponding group on the other side of the Channel similar in area and in the form of 1x10, like those in England, these customary acres will take a place of new interest in economic inquiry. It may be possible with more or less success to follow them back to the central home from which the Celtic tribes or possibly earher im- migrants wandered into the western extremities of Europe bringing with them as a part of their racial 8 Introduction. possession whatever of civilisation they had already attained to, whether derived from a still earlier home, or gathered, since their settlement in Central Europe, from the agricultural methods of the great corn- growing regions of the nearer or farther East. Such an inquiry cautiously and sanely conducted and confined strictly to the facts bearing upon it ought not, I think, to be unproductive. It may possibly lead us along unexpected paths, but if pursued with caution it ought not to lead us astray. If this evidence of customary acres stood alone I confess that, remarkable as it always must be, it would not bear the full weight of the inference which I am seeking to draw from it. But it does not stand alone. Through its connection with the open field system of agriculture it becomes involved in the system of tribal tribute and food-rents which become at last the basis of medieval taxation. The methods of tribute to the chieftain \mder tribal custom in Wales becomes at once closely connected with the plough, even in the shape of co-aration of the waste. The nature and meaning of the tribal systems of tribute become thus important elements in economic history, and though to trace with any fulness the passage of the ' 'firma unius noctis ' of the tribal chieftain and the still more mysterious hide as elements in the natural evolution of land-management and the manorial system in Britain, would lead us beyond the limits of this particular inquiry, yet from the point of view of its object we shall find that these areas of tribal tribute play a part in Britain which brings them strictly within its limits. They play an important part in the understanding Introdwiion. 9 of the relation of the tribesmen to their chieftain, supplementing om: knowledge of the patriarchal relation of the tribesmen in their gwelys to their ancestral head. Even dm:ing the tribal period the tribute of the chieftain becomes detached from the tribesmen and charged upon the tref or land of the tribal settle- ment. It easily becomes so because of the shifting character of the occupants. But the important point is that, even during tribal occupation, the tribute as a charge upon the land becomes a kind of property, sometimes a family property ; for the chieftainship is itself a family inheritance, following the same tribal instinct which ruled in the gwely. The descendants of Cunedda have to be dealt with as members of a Royal House and have their rights to share in the tribute or to take it sometimes in the ' nine nights' entertainment with their retinue and dogs.' So a privileged host of subchieftains or mactigerns become a prominent feature in the actual conditions of later tribal society. And the strange thing is that if we try to foUow the course of Wessex and Mercian conquest, whether in the Severn vaUey or in West Wales, we find amongst the conquerors a similar class of petty kinglets with rights of property and claims resembhng those of the petty chieftains of the conquered district. If the early charters of the Severn valley, strictly genuine or not, may be trusted, the Mercian or Wessex petty chieftains could join, and had to join, in the donations to the monasteries because they had rights which had to be surrendered by the grant. If we turn to the Redon Cartulary and seek to learn the conditions of Breton tribal hfe we are at first disappointed. The learned Editor seems to get nearly 10 Introduction. all his information from the Welsh side of the Channel. We have to face the charters of the eighth century with httle previous knowledge. And what do we find 1 Practically, in a word, a settled agricultural population, the tref under some name or other being still the usual geographical unit and a whole host of mactigerns or petty chieftains in many cases owners of the trefs. Whilst the main result of this inquiry may be to lead to the recognition of a very ancient and widely extended substratum of common agricultural tradition and methods, there can be no reason why full recognition should not be given to the traces found even in modem customary measures of modifications of earher custom owing to the influence of colonisation and conquest. Colonisation is indeed a very different thing from conquest. When an organised body of colonists settle in a country they may naturally bring with them their own pecuhar methods of agriculture, and, becoming the dominant inhabitants of a limited district, displace the methods of the earher inhabitants or impose their own methods upon them. We shall find the customary agricultural measures of the Greek colonies in Africa, in Magna Grsecia, and in Sicily from this point of view especially instructive. We shall find, too, on the other hand that the evidence of the Roman Agrimensores becomes equally striking in their recognition of agricultural measures, not their own, and left undisturbed after Roman conquests even in Italy itself and also in the interesting instance of the Greek colony of Cyrene taken over by the Romans from the last of the Ptolemies. Nor can we rightly shut our eyes Introduction. 11 altogether to such influences as may have come along the routes of comutnerce, though to enter fully into so intricate a subject would be beyond the Umited scope of our purpose. The same may be said with regard to the mass of archaeological evidence pouring in from all sides, which could not possibly be dealt with even were the writer competent to deal with it, in a subordinate section of such a volume as this. This Essay, or series of Essays, is avowedly limited in its scope. It is but a tentative attempt to bring together and present for consideration of the economic student facts relating to one section of evidence bearing upon a wide and intricate subject which is obtaining more and more attention, viz. the gradual growth of Western civihsation regarded from an economic point of view. Even should the facts placed before the reader in this volume leave many problems unsolved, they will, I venture to think, have done something to emphasise the importance to the understanding of Economic History, of the recognition of the existence through the ages of a substratum of ancient agricultural tradition and custom embedded in the moral and mental habits of the peasantry, extending over a wide geographical area, having its roots it may be in East, but deeply rooted also in the social and economic hfe of Central and Western Europe. The object of the first section of this volume will be to show that within the Umits of Cymric, Gaehc, and Brehon tribal custom there was a consistent and common knowledge and use of natural measures both itinerary and agricultural ; that the areas of tribal 12 Introduction. occupation and tribute were closely connected with the ancient itinerary measures common to Gaul and Britain, the GalUc leuga of 1500 paces being prevalent in Britain and so deeply rooted in the minds of the British peasantry all over Britain that it held its own in common use till the time of the Tudors and Stuarts and the general introduction of the statute mile. After examina- tion of the various customary acres prevalent in Britain, in passing to the corresponding group of the great corn-growing district of France, it was necessary to examine, however imperfectly, some of the special features of the Breton open field system with which they were so closely connected. The result of the examination of the Armorican customary acres will be to show their identity in area with the British acres, accompanied generally with a remarkable change in form, i.e. from the British form of 1 X 10 to that of 1 X 5, the French acre thus embracing two half acres, each 1 x 10. The geographical spread of the several acres will be marked on the map, thus showing that whilst the group as a group covers only a hmited area, individual instances of them can be traced sporadically eastward in a somewhat slender and often broken line across to the Rhenish and Danubian valleys to the plains on the mouth of the Danube on the one hand and over the Alpine pass into the Po valley on the other hand. A further examination of the agricultural units of Italy and those of the Ligurian district of France, and lastly those of Spain, brings back the inquiry to the meaning of the group of associated customary acres of Armorica and Britain. Lastly, a chapter deals with some other questions Introduction. 13 relating to the passing of the tribal and pastoral con- ditions to the open field system of the village community, and finally an attempt is made to show how far further light has been shed by the facts examined upon the position of the open field system of husbandry as an important factor in Economic History. F. SEEBOHM. The Heemitage, Hitchtn : January 1912. PART I. BEITISH AND lEISH EVIDENCE. CHAPTER I. THE WELSH UNITS OF TBI BUTE AND FOOD-RENT. I. THE TREF AS THE UNIT OF TRIBUTE. The question to be considered in this section is the character of the economic or areal units paying the chieftain's food-rent in kind known as the gwestva, or the tunc pound in lieu of it, under early Cymric custom. What were the ' maenols ' and ' trefs/ responsible for the gwestva ? Following no doubt prevalent but more or less vague tradition, the Venedotian Code of North Wales points them back to the time of the mythical Dunwal, ' the great measurer,' who, ' before the crown of London and the supremacy of the island were seized by the Saxons, established good laws in the island which continued in force till Howell's time.' It goes on to say : ' The cause of his measuring the island was that he might know the tribute (mal) of the island, the number of " milltyrs " and its journeys in days.' And then follows the measurement of the erw or local acre and the number of erws in the trev and the maenol and the cantref. c 16 Welsh Units of Tribute. The cantref is said to consist of two cymwds each of 12 maenols. ' And of the twelve maenols four are assigned to ailUs to support dogs and horses and for progress and dovraeth nd one for canghellorship and one other for maer-ship, and the rest for free uchelwyrs.' ' And from those eight [maenols] of free uchelwyrs the brenhin [or head chieftain] is to have a gwestva every year, that is a found yearly from each of them.' ^ The gwestva or tunc pound was therefore the tribute paid to the chieftain by the uchelwyrs or free tribesmen in the free maenols and trevs, quite apart from that of the aillts and taeogs occupying the other maenols. In South Wales, according to the Gwentian and Dimetian Codes, the trev was the unit paying the gwestva or tunc pound, and whilst there is no state- ment how many tune-paying trevs there were in the cantref they are said to be grouped into maenols of 13 trevs. What then was the trev or tref ? According to the Codes it was a unit of occupation, traditionally reckoned as of a definite area and as normally containing so many tyddyns or homesteads. And I think we may go one step further, and fairly identify the group of its occu- pants with what in the Codes and law treatises is known as the trefgordd. In the phraseology of the Codes and treatises the gwelegordd was the family group of members of the gwely. The gosgordd was the group of persons in the retinue of the brenhin. So that the trefgordd would naturally be the group of occupants of the tref. And as the tref, or its multiple, was the areal gwestva- paying unit, its group of occupants — ^the trefgordd — would be the payers of it. 1 Ancient Laws and Inst, of Wales, 1841, vol. i., p. 189. The Tref. 17 What then was the trefgordd 1 The word ' trefgordd ' is in frequent use in the Welsh Codes and the legal tracts of later origin. I have tried, in my last volume, to reahse what it meant in terms of the actual tribal everyday hfe of the period — let us say of Howell's time. The typical trefgordd turned out, according to the hints to be gathered from the traditions recorded in the Welsh Codes and later legal treatises to be the normally complete or tjrpical self-contained group for the purpose of tribal husbandry or occupation. It appeared to be a group of occupants with its one herd of 24 cows and one bull, under one herdsman and his dog, with one common churn for the milk of the herd, and with one plough and plough-team of 8 oxen for such co-aration as was required by a mainly pastoral group of occupiers. It seems probable that the tref was theoretically the area occupied by such a trefgordd, i.e. in theory the normal extent of land considered as sufficient for a complete self-contained group under pastoral conditions of tribal occupation, with a complete plough-team and cattle enough to keep up, roughly speaking, a milking herd of 24 cows. The unit responsible for the food-rent or tribute was apparently regarded at the time of the Codes, not as the group itself but as the area occupied by it. The occupiers for the time being pay the tribute due from the tref occupied by them. According to the Codes the trefs of free tribesmen seem, as akeady said, to have been originally kept distinct from those of the non-tribesmen or taeogs. But the Denbigh Extent bears witness to the fact not c 2 18 Welsh Units of Tribute. only of the ease with which the occupants of a tref or ' villata ' with their cattle could be moved, but also that in later times at any rate the occupj^g group might be a mixed one composed of tribesmen from more than one gwely and even of non-tribesmen also. A gwely might be scattered with its cattle in several tune-paying villata and have sometimes only a small undivided share in any particular one. The cattle, we infer from the earlier evidence, though in one herd and under one herdsman, often had several owners. As in Swiss mountain pastures now, so in Wales at the time of the Codes, the results or value of the milk of the herd in the common chum was divided among the o^vaers according to a test-milking of their several cows on a certain fixed day. And further, as in Switzerland to this day the cattle do not remain in the same place all the year round, so in Wales there were summer and winter trefs in tribal times. It is not needful to go further into details except to emphasise the fact borne out by the Codes and also the Denbigh Survey that the group of occupiers consisted of several households with their separate tyddyns or homesteads, according to the description of Giraldus Cambrensis scattered about here and there, and that in the tyddyns and their cattle-yards there must have been provision for the oxen contributed to the plough- team as well as for the winter shelter of the other cattle. Thus the agriculture of the strictly pastoral tref differed widely from that of the settled village community with its permanent agriculture and its scattered holdings in yardlands on the open field system. It was a co- aration by the common plough of portions of the pasture The Tref. 19 or waste cKosen each year, the crops of the cyvars or erws ploughed being taken by the contributors to the full team of 8 oxen according to the rules laid down in the Codes, i.e. in the same order of rotation as the position of their oxen in the plough-team. So that regarded as a taxable or gwestva-paying area the tref would not consist of so many yardlands nor even of so many erws of pasture and so many of permanent arable, but each year what was required for arable would be taken out of the common pasture, to become pasture again after the removal of the crop, thus laying the foundation in the course of pastoral husbandry for what in more settled agricultural communities afterwards became so prominent an element in the open field system, as already pointed out. Whether the gwestva-paying area was a tref ox, as in Venedotia, a ' maenol ' of four trefs does not matter for the present purpose. It is enough that we can see how, even from a mainly pastoral point of view, it would be easy for a conquering chieftain to estimate roughly by its area how many such economic units or centres of food-rents a country side or a larger district theoretically should contain ; or to ascertain how many at any time it did contain, by counting the ploughs or the herds or the churns, or, in the case of a mountain district devoted to sheep, by counting the shepherds. It is easy to see also how the tune-pound paying area might by gradual transition from pastoral to more strictly agricultural times easily come to be regarded even as a plough-land, just as in England hides more or less tended to become carucates, under similar circumstances. 20 Welsh Units of Tribute. The Welsh word ' tref in its Irish equivalent ' treab ' with its wide meanings is significant enough to illustrate this point. Its dictionary meanings, according to Dr. Atkinson's glossary of the Brehon tracts, are the following : — Treb, dwelling house, a farmstead. Trebad, act of ploughing, husbandry. Trebaim, I plough, manage a farm. Trebaire, farmer, householder. Bo ireb(ith)tha, plough-ox. Lastly, it hardly need be said that in taking the tref to be the typical economic and areal unit of pastoral occupation and also of tribute, it is by no means to be regarded as probable or even possible that once chosen as the typical unit, let us say in Howell's time, it could remain in each instance ever after unchanged. There is nothing to show that in another generation the growth of population might not require an extra plough-team, or that the typical herd of cattle might not have so increased as to require another herdsman with his dog. And yet the tribute of the tref might remain the same tunc pound as at first fixed. Indeed, we so find it in the Denbigh Extent of the fourteenth century. II. THE LAND MEASURES OF THE WELSH CODES. However much the maenols and trefs of the Codes, regarded as the units for payment of the gwestva or tribute of the free tribesmen, may have varied from time to time in area or in contents, we are bound by the Codes to regard them as containing theoretically in Howell's time a normally definite area. Before, however, we can realise what was the meaning Land Measures of the Codes. 21 of the areal extent of the fiscal unit of each of the three districts we must first ascertain upon what system of measures their area was determined. Nor need this in itself be a fruitless or uninteresting object of research, for if the result should be to bring these Cymric normal units of tribute more or less into direct connection with known historical realities, the gain will be worth the trouble involved in the sense of reahty given at the outset to the inquiry on which we are engaged. The measurement of the trefs is described in the Venedotian Code as having been fixed by Dunwal ' the great measurer." But let it be clearly understood at the outset that in dealing with what for convenience we may speak of as DunwaFs land measurements, we are not deahng with the merely imaginary methods of a mythical personage. We are really dealing with the statement in the Venedotian Code of what were the measurements of the erw, the tref, and the maenol as declared to be actually in use in Howell's time or at the date of the Code. We are dealing with facts of every- day life in North Wales in the early tenth century, one of these facts being that the customary land divisions were not of new invention, their origin being already lost in ancient tradition. That they were attributed to a mythical pre-Saxon ruler of Britain shows only that they were already estabhshed by immemorial custom, and as such we must regard them, quite irrespective of whether the mythical ruler ever existed or not. The text of the Code proceeds to describe the milltyr which formed the basis of the measurements thus : — 22 Welsh Units of Tribute. Starting with the barleycorn — 3 lengths of the barleycorn in the thumb. 3 thumbs in the palm. 3 palms in the foot. 3 feet in the pace. 3 paces in the leap. 3 leaps in the land (tyr), (in modern Welsh called a grwn). 1000 lands in a ' mylltyr.' And the clause concludes with the statement ' And that measure we still use here.' The ' tir ' or ' land,' 1000 of which made the miUtyr, was thus 27 natural feet or 9 natural steps in width. It was in fact the small end of the strip of open field ploughing, frobably the day work of the plough in the practice of co-aration with the customary 8 oxen plough. Thus the system which hes at the base of the measure- ments in the Welsh Codes, hke those of ancient Egypt and Greece, was based upon natural measures, and in Wales the milltyr as the great itinerary measure was obviously intimately connected with the system of agriculture. The custom of reckoning distances in ' acres,' i.e. by so many small ends of the strip or ' land ' in open field ploughing, is not an otherwise unknown one. Distances up to our own times have been traditionally measured in several counties in acres, i.e. by so many short ends of the local customary acre, and the practice has found its way with emigrants into Canada. That the milltyr as the great itinerary measure was 1000 ' land-ends ' need not therefore surprise us. Although customary acres in different districts vary in size, it is clear that Dunwal's traditional ' land-end ' of 27 natural feet or 9 natural steps, was that generally adopted by custom. That it continued long in use in Land Measures of the Codes. 23 Wales is shown by the following passage from a later legal treatise ^ which graphically describes the method of settling a dispute as to land measurement at a much later date than the Codes : — If there be a suit as to land and soil between two persons and one is adjudged to meer (tirvynu) to the other ; he is to bring rehcs ; and the owner who swears is to swear at the end of every nine stops (ym pen pob nau cam) he shall walk, and that with his feet on his own meer (tervyn). And so from place to place he shall begin measur- ing unto the place where he shall conclude. And these are called the ' nine steps of meer (nau cam tervyn).' The nine steps were the 27 feet of Dunwal's ' land- end/ and thus the phrase is proof of the continuance of the habit of measuring by so many short ends of the ' tir or land ' of 9 single steps. Local custom may be of very ancient origin, but, even if following the tradition of the Venedotian Code, it were allowable to connect the arrangement of Welsh units of tribute with so late a period as that of the break up of Eoman rule and so (as suggested by Sir John Rhys) with the coming or migration from Ciimbria into Wales of ' Cunedda and his sons,' and if we might further take into account the suspected connection of Cunedda with the mihtary arrangements of the district of the Roman wall, so closely bordering upon the Northumbrian and Cumbrian region, it would still be most unlikely that Cunedda's system of measure- ment should be, even then, one sprung de novo out of the clouds. It surely would point to traditional methods already in use in the district from which they came and probably of wider prevalence. Upon what system of assessment, it might well be asked, was the tribute in corn levied which Agricola found to be so ' Ancient Laws, ii. p. 207. 24 Welsh Units of Tribute. unjust and whicli he (according to Tacitus) so justly reformed ? It must surely have been collected upon some system akeady in use, for tribal custom and tradition do not easily change in such matters. That the primitive method of measuring land in Cumbria and North Britaki under early British and Gaelic custom was by putting one foot closely before the other is illustrated by one of the poems in Mr. Skene's ' Book of the Dean of Lismore ' : ' The piteous tale of Diamed's slaying the great boar.' ' Measure, Diamed, the boar down from the snout,' cried Finn, hoping to secure his death. ' Along the back he measures now the boar.' Diamed is still unwounded and Finn orders him to measure it the other way ' against the hair.' ' He measured once again the boar. The envenomed pointed bristle sharply pierced The sole of him. The bravest in the field Then fell and lay upon the grassy plain.' ' The next point to be noted is that in the Gaelic districts of Scotland in the twelfth century (i.e. in King David's time) this primitive natural system of measure- ment was still in use. Even in legal enactments at that date no other system or standard seems to have been available. The length of the legal rods by which land was measured could not, it seems, have been otherwise described or understood than in thumbs and feet and steps of a middle-sized man. And it happens that though some variation in the length of the standard of the middle-sized man's foot must be allowed for, there is one instance in which the recognised length of the natural foot in GaeHc Scotland can be accurately 1 Diamed was a ' yellow haired and blue eyed ' Celt in these poems. In Mr. Skene's view they embodied tradition common to both sides of the Solway Firth. In another poem the tusk of a bear is measured by the ' thumb of Finn ' — thirty thumbs of Finn. To face p. 25. k rhumb H/" ■ O i OO i Pa/m K - - Thumb - - ^ ft*^ O ' O i O - Pa/m - .0 ,^ .^ .<^ «" :v # SOf* Fia. 7. The tref, or gwestva-paying unit of 1248 erws, would contain 10,223,616 square natural feet, i.e. nearly exactly ^\ of the square milltyr, i.e. 10,125,000 square feet. Tojace p. 35. & ,(<' vtf" *; b^ A\^ ,& 1.1'\6' 444.3S8 ares DunwaVs Venedotian Maenol paying the tunc pound according to the description in erws in tiie Code. 22,391,880 square natural feet. 22,781,250 = Jj square milltyr. Area on the standard of -iil foot according to Code 13,834 arcs. As division of the square milltyr 13,886 ares. Gicentian tref paying the tunc pound as described in eiws in the Code. 14,556,672 square natural feet. 14,580,000 = j'j square milltyr. 8881 arcs. 8887 ares. ^5 Dimeiian tref paying the tunc pound as described in erws in the Code. 10,223,616 square natural feet. 10,125,000 = j\r square milltyr. 6217 ares. 6172 ares. Results of Welsh Evidence. 35 It would take the form of two squares and its area would be 6217 ares, the 7-3 of the square milltyr being 6172 ares. And it would equal one-sixth of Howell's maenol, and | the square leuga. VII. GENERAL RESULTS. To sum up the result of this examination of the areas of the Welsh tune-paying units of tribute in their relation to the square milltyr, we have found them to work out successively as under : — Howell's maenol . . ■ j^ oi tlie square milltyr. Dunwal's „ . . • sV Gwentian tref Dimetian „ 1 TO The diagrams will show to the eye how Howell's maenol falls as a division of the square milltyr into three squares while the others fall into two squares. However mysterious may be the traditional origin or author of these measurements of the units of tribute adopted in the Welsh Codes, the correspondence of them as divisions of the square of the great itinerary ' milltyr,' described as in use for the purposes of the ' mal,' goes very far, I think, to show that they may be regarded with some confidence as in a true sense realities. They must be regarded no doubt as statements of normal or typical areas, i.e. of the amount of land traditionally reckoned, if we are right, as sufficient for the occupation of the typical trefgordd with its one plough, its one herd of cows, and its one churn. As such, and so far, they bear the stamp of reahty. Further it happens that there is another khid of evidence, altogether independent of the Codes, which, so far as it goes, incidentally confirms the 36 Welsh Units of Tribute. genuineness of these typical areas. There is the evidence of the Report of the Agricultural Commission of 1820, already alluded to, as to the customary local rods by which agricultural areas had for generations been measured in Wales, and which were still in cuistomary use. They are described by the Commissioners in English feet, not therefore with absolute exactness, but sufficiently for our purpose. (1) The paladyr of North Wales, the llath of Merioneth- shire, and the pedair llathen of Montgomeryshire, all of which are alike described as 4^ yards or 13| English feet i.e. = 4-12 m. (2) The bat or eglyshaw described as 11 English feet, i.e. = 3"35 m. (3) The Cornish gad or goad, known in North Wales as cyvelin, 9 English feet .... i.e. = 2-75 m. Now these customary local rods are not mentioned in the Codes. We know of them onty by the fact that they have been from time immemorial the customary agricultural measuring rods in local use. The fact is important that these customary local rods are distinctly connected, not only with the local customary acres of Wales, but also with the milltyr of the Venedotian Code. (1) 1600 paladyrs = the milltyr (6592 m.) (2) 2000 bats = „ (6700 m.) (3) 2400 gads = „ (6600 m.) And what is still more remarkable is the relation of these customary rods to the double squares of the maenols and trefs described in the Codes. (1) 200 X 400, i.e. 80,000 square paladyrs, would make the two squares of Dunwal's Venedotian maenol (824 x 1648 m.) (2) 200 X 400, i.e. 80,000 square bats, would make the two squares of the Owentian tref . . (670 X 1340 m.) (3) 200 X 400, i.e. 80,000 square gads, would make the two squares of the Dimetian tref . . (550 X 1100 m.) Results of Welsh Evidence. 37 The vuiiform correspondence of 200 x 400 of the Welsh customary measuring rods in actual use, with the length of the sides of the two squares into which the maenols and trefs fall as divisions of the square milltyr, afiords strong and independent incidental evidence of the reality of the description of them given in the Codes. Finally, before we pass from the consideration of these Welsh units of tribute, there yet remains the further important question whether the actual area of Wales will admit of the possibiUty of maenols and trefs with so large an area as these figures suggest. Before we can fully rely upon Howell's version of what he regards as Dunwal's measurements for the purpose of his ' mal,' we must ascertain how far they are consistent with the area with which his system had to deal. It is notorious how often traditional records of figures contain within themselves the proof of un- certainty and exaggeration. It is not so in this case. According to the statement in the opening clause of the Gwentian Code there were in Cymru in the time of Howell the Good : — In Venedotia . . . .18 cantrefs In Deheubarth or South Wales 64 ,, In all . . .82 cantrefs (i.e. 8200 trefs) There were also 3 score trefs beyond the Cyrchell 60 „ and 3 score trefs of Buallt 60 „ Making a total of . . 8320 imfs in all. There are now 4,721,823 English acres in Wales and 368,399 in Monmouth, making a total of 5,090,222 acres. This divided by 8320, the number of trefs, gives 609 acres for the tref. Sir John Price in his ' Description 38 Welsh Units of Tribute. of Wales ' (1559 a.d.) estimates the number of cantrefs differently, thus : — In Gwynedd or North Wales . . .16 cantrefs In Powys ....... 13 ,, In Mathrafael between Wye and Severn . 4 „ In Dynefawr 26 „ 59 cantrefs in aU, making the number of trefs 5900. This estimate, which is considered by recent writers to be still excessive, would allow 800-900 acres to each tref. Ares. statute statute acres acres. in tret. Howell's Venedotian maenol of 4 trefs ..... 37,300 930 233 Dunwal's Venedotian maenol of 4 trefs ..... 13,886 347 87 Gwentian tref .... 8,887 — 222 Dimetian tref .... 6,172 — 154 Obviously there would be room for all the trefs with ample surroundings of moor and mountain even on the highest estimate of the number of cantrefs. On the whole, therefore, we may, I think, fairly conclude that figures which touch reality in so many directions cannot be put aside as mere random guesses. They seem to be founded upon local knowledge of the facts and customs which formed the basis of the system of tribute described in the Welsh Codes, and they are all the more to be trusted from their close connection with the agriculture and actual methods of ploughing under the pastoral conditions of tribal life. CHAPTEE II. THE IRISH UNITS OF TRIBUTE AND FOOD-RENT. I. THE LAND MEASURES OF THE BREHON TRACTS. We are now in a position to extend our inquiry to the Brehon or Irish land measures and units of food-rent or tribute. Like the Welsh, they are based upon natural measures starting from the ' ordlach ' or inch, described as of three barleycorns put end to end. But, hke the Greeks, the Brehon writers make their calculations in an arti- ficial foot of 12 inches or 4 palms, instead of the 9 inches and 3 palms of the natural foot. It will be easy, however, to translate the results into natural feet of 3 pahns for purposes of comparison. Without going too much into detail it may be said generally that from the thirteenth century records down to the Carew MS. of the time of James I. the prominent units of land measurement are described as plough-lands. The Brehon tracts, whilst disclosing that co-aration by tribesmen was a necessary and important element in tribal husbandry, as in Wales, describe the tribal arrangements in terms and in the language of pastoral hfe, with remarkable survivals from a wilder hfe still, in which the chief export of the island was the female slave. In the Brehon tracts values are generally stated 40 Irish Units of Tribute. in cows, and cumals of cows. The cumal or female slave came to be reckoned as equal to three cows, and was the chief unit of value and payment. As I have shown in my last volume ^ the payment in cumals or ancilloB forms a striking connecting hnk between the Ireland of the Brehon tracts and the Armorican district. For not only did the cumal of the Brehon tracts find its equivalent in the ancilla of the Breton ' Penitentials,' but even the equation with silver was the same in both cases. The cow was reckoned as equal to the ounce of silver in both cases, and the female slave of 3 cows to 3 ounces of silver, the ounces on both sides of the Channel being apparently the Eoman ounce of 576 wheat grains in weight. As com- pared with the Welsh Codes the silver value of the ' cumal ' was roughly equal to that of the Welsh cow, equated with 3 ounces of somewhat differing weight. Leaving aside then for the present the plough-lands of the later documents and breathing freely the air of the Brehon tracts with their payments in cumals and cows, the inquiry must be confined to the question what units of areal measurement are described in these tracts as in use for the purpose of food-rent to the chieftain, or in payment for rights of grazing under tribal customary arrangements. The answer is that throughout the Brehon tracts the amounts of land held by the various grades of tribesmen and chieftains are described in so many tir-cumails ; and that two chief normal units of area for the purpose of grazing and food-rent or tribute are independently described in two separate tracts 1 Tribal Custom in Anglo-Saxon Law, 1902, chap. iv. The Smaller Tir-cumail 41 (in 'Ancient Laws of Ireland,' III. 325 and IV. 276), both, these normal areas being called by the ancient term ' tir-cumail,' and each of them being in the form with which we are now so familiar, of two squares. II. THE SMALLER TIR-CUMAIL. The smaller unit (III. 339) easily justifies its name as a ' tir-cumail ' — ' the land of the cumal ' or female slave. It may not at first sight seem to be directly connected with food-rent or tribute. Its direct con- nection seems to be rather with the loan of stock by the chieftain to his tenant, the cumal of three cows being the return for the grazing of the ' tir-cumail ' of good pasture-land for a year. But as every chieftain made loans of cattle to his tenants, the practice had its due place in tribal economy. That this was the case is shown by a quite inde- pendent passage (IV. 304) which states that the ' tyr-m ho ' i.e. ' land of one cow ' — is that on which seven cows are grazed and for which one cow is given at the end of the year. (See also IV. 133 and III. 129). The smaller ' tir-cumail ' thus represented the normal area for 21 cows' grazing of good land and a return or rent at the end of the year of a cumal of 3 cows. The 3 cows rent added to the other 21 cows made a herd of 24, which with their bull would make what we shall find was in Ireland the normal herd connected with the plough-land. It would thus corre- spond very closely with what we have seen to be the normal herd of Welsh tribal custom associated with the single plough. 42 Irish Units of Tribute. The dimensions of this smaller ' tir-cumail ' are thus described and the figures of actual length are easily added on the basis of the variation in the natural foot : — 3 grains (elsewhere de- scribed as barleycorns, set end to end) 4 orlachs 3 bais 12 troigeds 12 fertaigs 6 forraigs 12 „ orlaoh (thumb) bais (hand) troiged (foot) fertalg (yoke) forraig •328--335 m. 3-926-4.02 m. 47 -232^8 -24 m. breadth of tir-cumail 283-4r-289-44 m. length of „ 566-8-579 m. (making an area of 1606 to 1676 ares in two squares) the side of each square being equal to ^ of the leuga. ■\ <'*'^ oi Fm. 8. The small tir-cumail and the square leuga (of higher standard). The tir-cumail taken on the lower standard = ^^ °^ t^^ square leuga taken at the higher standard, and it would contain 400 British statute acres in its two squares. That the orlach was equal to the Welsh thumb The Larger Tir-cumail. 43 or inch is shown by its being equal to 3 barleycorns, set end to end. But instead of proceeding to describe the palm of 3 orlachs or inches, the writer makes a ' hais ' or Jiand, i.e. palm with thumb, of 4 orlachs, and an artificial foot of 3 hands or 12 orlachs, which woiild equal 4 Welsh palms. Converting for purposes of comparison his fertaig of 12 feet of 4 palms into 16 natural feet of 3 palms, the Brehon fertaig becomes equivalent to the Welsh ' long yoke ' of the Venedotian and Dimetian Codes. The forraig of 12 fertaigs becomes 192 natural feet or 12 long yokes. The breadth of the tur-cumail becomes 1152 natural feet or 72 long yokes, and the long side of the double square becomes 2304 natural feet or 144 long-yokes. So that intimate con- nection is at once indicated both with the leuga and with Welsh agricultural measures. III. THE LARGER TIR-CUMAIL. The larger tir-cumail is described as follows (in IV. 276) in the tract ' Divisions of Land ' : — 3 grains 6 inches 2 hands 6 feet 6 paces 6 intritts 6 laits = proper inch (ordlach innraic) = hand (dom) (with thumb extended) foot (traiged) . . . ■328--335 m. pace (deisceim) . . . r968-201 m. intritt .... 11 •808-12-06 m. lait 70-85-72-36 m. forrach 425-10-434-16 m. 6 forrachs = airceand (breadth of tir-cumail) . 2551-2605 m. 12 „ = fot (length of „ ) . 6102-5210 m. Its area is 130,152 to 135,721 ares. It will be observed that the dimensions of this greater tir-cumail are 9 X 9 times those of the lesser 44 Irish Units of Tribute. tir-cumail, and so the area is 81 times that of the lesser tir-cumail, ^nz. 214,000,838 square natural feet. Fig. 9. Great tir-oumail of higher standard, 5000 medimni, 400 square stades. Postponing for the moment the question of the relation of this greater tir-cumail to itinerary measures, the more immediate point from a Brehon point of view is whether it corresponded with any widely prevalent traditional or customary land division mentioned in Irish Annals in the Brehon tracts or in the ancient The Larger Tir-cumail. 45 Irish poem ^ which most directly deals with the ancient land divisions of the coimtiy, and which Mr. Skene regarded ' as giving probably the oldest views of the land divisions all over Ireland.' ^ In the ' Annals of the Four Masters ' there is an entry which tells us that the mythic King Ollamh Fodla appointed a Taoisech over every triocha ced and a Brughaidh over every haile. The trichaced and the baile were therefore recognised divisions in early times. According to the ancient poem Ireland contained 184 ' trichaceds.' Each trichaced contained 30 ' baile biataigs ' — i.e. ' places of food-rents '■ — each sustaining 300 cows, and, according to the poem, there were 5520 of them in the four provinces. The trichaced goes back far into the past of Irish poetry and legend. And the term itself, meaning ' thirty hundreds,' implies that the baile must have been in some sense properly a ' hundred.' It was seemingly a ' hundred ' 1 This poem, the text of which is given in the Celtic Society's edition of the Battle of Magh Leana (p. 107) and Mr. O'Donovan's translation of which is also found in Skene's Celtic Scotland, iii. p. 154, may not be itseK of very early date, seeing that it uses words and phrases which are not found in the Brehon tracts, e.g. ' seisrig ' for plough-land containing 120 ' acra ' (neither of these two words being found in Dr. Atkinson's Glossary). And it is even open to suggestion that the plough-land of 120 acres may have been intro- duced by early English emigration. But it no doubt represents, in general, very early tradition. The seisrig is translated probably wrongly, as a ' six-horse plough.' In the Crith Gabhlach (iv. p. 307) we read of ' a fourth part of ploughing apparatus, i.e. an ox and a ploughshare and a goad and a bridle,' showing that the Bre- hon plough-team was of 4 oxen, and implying co-operation in the ploughing. I am indebted to Mr. Norman Moore for the information that the modem word seisreachi, plural seisrighe, is now used in Ireland for a, ' plough-team yoked to a plough.' ^ Celtic Scotland, ui. 154. 46 Irish Units of Tribute. because according to the poem it found grazing for 300 cows, i.e. 100 cumals of cows. At any rate, it is difficult and perhaps needless to make any more Hkely suggestion. Now Ireland contains 19,759,600 English statute acres, and this area divided by the number of bailes would give about 3525 acres to the baile. The great tir-cumail would contain roughly 3200 English acres. So that if there be any correspondence between this greater tir-cumail and the ancient tribal divisions of the poem it must be with the baile or typical place of food-rent. The figures are near enough at least to suggest the identity of the greater tir-cumail with the baile. The poem describes the latter as a complex unit of agriculture and pastoral husbandry somewhat of the type of the Welsh trefgordd. It informs us that its area contained both arable and pasture, being normally reckoned as including room for 12 ploughs with 120 acres to each, besides pastvire for 300 cows, i.e. for ' 4 full herds ' of 75 cows each. It will be seen at once that the combination of ploughs and cattle practically corresponds with that of the Welsh trefgordd. Each of the 'four full herds ' of 75 cows is associated with 3 ploughs, so that each single plough is associated with a herd of 25 cows as in the Welsh trefgordd. Assuming, as we probably may fairly do, that the greater tir-cumail was identical with the baile or greater unit of food-rent, the next point is that the plough-land or area allotted to each single plough and the grazing of its 25 cows was one-twelfth of its whole area. 47 / ,/ 6 \«' e Bude.^ fe rude off lande in baroyis sal conten VI elne fat is to say xvii 1 With regard to the date of these documents. In the Begistmm of the Abbey of Aberbrothoc (p. 163) is a record of a perambulation in 1220 expressly recognising the legal Assize of King David as still in force and (at p. 89) there is a charter which describes two bovates of land as containing 26 acres, i.e. 13 a. to the bovate. ' Acts of Parliament of Scot- land, p. 751 red. in Gaelic Scotland. 59 fut off a mydljTi mane, fe rude off pe land in fe burghe mesurit ofE a midlyng mane sal be xx fut. The mesuring of landis. In fe first tyme pat fe law wes maid and ordanit fai began at }>e fredome of halikirk and syne at fe mesuring of landis fe plew land fai ordanit to contene VIII oxingang, J>e oxgajig sail contene xiii akeris. The aker sail contene four rude, }>e rude xl faUis. The faU saU hald vi ellis.' According to the first of these fragments the acre in Baronia of 4 X 40 of the 18-foot rod (allowing for the recognised variation in the natural foot •246-"251) would have a furrow of from 177 to 181 m. and an area of from 31 "33 to 32 "76 ares. Taking the higher standard of King David we have no difl&culty in recognising it as a half-acre put into the form of 1 x 10 of the well- known Irish statute acre, also the customary acre of the Northumbrian and Cumbrian district with a furrow of 256 m. and an area of 65 '62 ares. The importance of this Scottish half-acre with a furrow of 181 m. and its extension over a large Gaelic area is shown by the prevalence of what the books call the old Scotch mile of 1810 m., i.e. 10 of its furrows in length mentioned in the Keport of the Board of Agriculture as prevalent as far north as Ross and Cromarty. The plough-land according to the other fragment — 1 Nor was this arrangement of 13 acres to the oxgang and 104 to the plough-land confined to the old Northumbrian district. It was prevalent in Teviotdale and the Merse (Innes, p. 241 and in 16th century note on Moynet MS. Acts of Pari. 196 red pages to vol. 1). Mr. Innes (p. 270) thinks that it appKed also to Aberdeen, Banff, and Moray. And this may perhaps be confirmed by the statement in Coimty Reports of the Old Board of Agriculture, 1795-1814, that the ' mile ' of Ross and Cromarty was that of 1984 English yards, i.e. 1815 m., whilst in the Report of 1820 (Weights and Measures) the mUe of Nairn and Moray ' in the cross roads ' was ' the old mile oi 2640 English yards,' i.e. 2413 m. — IJ Enghsh miles. 60 Units of Tribute that on ' The Measuring of Land ' — contained 8 oxgangs each of 13 of the acres (or half-acres) in Baronia. It therefore consisted of 104 of them and would contain an area of from 3258 to 3407 ares. It would be difficult from a Gaelic point of view not to recognise in this Scottish plough-land so described the double of the smaller of the Brehon tir-cumails, which on the same standards varied in area, as we have seen, between 1606 and 1676 ares. A nearer approach to correspondence could not have been made in these acres without making the munber in the oxgang 12*8 instead of 13. The closeness of correspondence is most obvious when shown in natural feet. The tir- cumail contained 2,654,208 square natural feet, while the Scotch plough-land contained twice 2,695,680 square natural feet. It would be still nearer the double of the tir-cumail if this were regarded as -^th of the square leuga or 2,700,000 natural feet put into the form of two squares as suggested in the previous chapter. Nor could it be passed unnoticed that if Mr. Skene and Mr. Innes are right in identifying the Scottish plough-land of these fragments with the areal unit of 40s. of the • Old Extent ' of the Early Scottish Kings the Brehon tir-cumail would become the one found land. In other words, consciously or unconsciously, the tir- cumail would become the areal 40s. unit of assessment of the ' old extent ' of Scotland. Nor have we yet exhausted the traces of connection between Scottish and Brehon areas of tribute. The greater Scottish agricultural area— the Davach — according to the same authorities ^ seems to have been ' Innes, Scotch Legal Antiquities, 1872, p. 270. in Gaelic Scotland. 61 composed of four of the plough-lands, and if so, it con- tained in area 13,032 to 13,628 ares, i.e. almost exactly Y^th of the area of the greater Brehon tir-cumail, which contained, as we have seen, 130,152 to 135,721 ares : the same variation of the natural foot being adhered to in both cases. The closeness of correspondence becomes still more obvious if we substitute for two of the Scottish half- acres the parent acre itself of which they are halves, with its furrow of 256 m., known to-day as i^e Irish acre, for then 2000 of these acres without change of their form of 1 X 10 would exactly fill the two squares of the greater tir-cumail. Further hght seems to be thrown on these remark- able correspondences between Brehon and Scottish areal units of tribute when the other acre mentioned in the first of the fragments as the acre in Burgo is examined. It turns out to be identical with the Enghsh statute acre. It is described as 4 x 40 of a rod of 20 feet of a middhng man. This rod on King David's standard (•251) would equal 5"02 m. and 40 such rods would equal 201 m. which is the length of the furrow of the English statute acre. Now it is curious that whilst an odd number of 104 of the acre in Baronia was required to make the Scottish plough-land, the even number of 80 of the Enghsh statute acre already met the requirement, though the second fragment describing the plough-land is silent about it. Forty Enghsh statute acres, consciously or not, equalled in area the smaller tir-cumail. But the close- ness of correspondence is best shown by the fact that if the statute acre were turned (as in the other case) into two half-acres of the form of 1 X 10 with a furrow of 62 Units of Tribute 142 m., then 80 such Enghsh half -acres, without change of form, would exactly fill the two squares of the smaller tir-cumail. English half acre Oxgang of the fragment Smaller tir cumail Pig. 13. Scottish plough-land. What then is the meaning of this Scottish plough- land ? The language of the second fragment would point to its introduction from a district which had long been in possession of it under customary laws which had ordained the 'freedom of holy kirk.' Its wording has a manorial ring about it, and taken in in Gaelic Scotland. 63 connection with the 8-oxen plough-team it seems to suggest the importation into a hitherto pastoral country of agricultural methods for long in customary use in the neighbouring Northumbrian district where both of the acres mentioned in the fragments were known. Finally, as an agricultural unit, I think we may identify the Scottish plough-land with that of the two field systems described in Fleta thus : — There should be 160 acres to the carucate, half for fallow, half for -winter and Lent sowing, i.e. 80 acres in each of the two fields. Such a carucate would correspond with the Scottish plough-land and consciously or not each of its two halves would correspond with the lesser tir-cumail of Brehon custom. I do not wish to press too strongly the re- markable correspondence between Brehon and Scottish areas of tribute as proving or pointing to direct con- nection. They may have been independently arrived at from itinerary measures common to both. But both itinerary and areal measures seem to keep throughout within the range of local Celtic custom and bear the primitive mark of being founded, not upon any imported fixed standard, but rather as part of an inherited tradi- tion of common methods of measurement based upon the natural length of the thumb and foot and step of the ' middling man.' CHAPTER IV. THE UNITS OF TRIBUTE AND ASSESSMENT OF THE DOMESDAY SURVEY. I. THE CORNISH ' GREAT AGER ' OR CARUCATE. We have now to consider the units of tribute and assessment of the Domesday Survey — ^the Cornish ' great ager ' or carucate and the hide. In the Domesday Survey there is direct reference to a Cornish ' great acre ' or carucate as a fiscal unit, but it is not easy and perhaps it may not be possible to find out precisely what it was. In the Exon Survey there are several instances in which the phrase is used, ' ager terre, hanc potest 1 caruca arare.' It therefore was a ' plough-land/ but cases also occur in which the acres and the ntmaber of teams which can plough it do not correspond. So that we are left in doubt as to its area and with something more than a hint that its area was not always the same. Mr. Round long ago pointed out passages in the ' Testa de Nevill ' from which at first sight one might obtain information as to its area. At p. 204 of the ' Testa de Nevill ' is an extent of the Pomeray manor of Tregony in Cornwall. And in the description of this demesne land is the following statement ; — The Cornish Carucate. 65 In dominico ii aoras cornubienses continentes ii caruo' comu- bienses, que valent per annum xl'. And in the list of Libere tenentes holding at that date by mihtary service, the phrase so many acrcB cornu- bienses containing as many carucatod cornubienses occurs again and again. There are Cornish half-acres containing Cornish half-carucates and also ferlings containing a fourth part of a Cornish carucate. There is also in the ' Testa de Nevill ' (p. 201) a list of the 146 Libere tenentes of the Bishop of Exeter in Cornwall holding ' acres ' which from comparison with the holdings in the Domesday Survey (120 b) and the amoxmt of their services and rents must be ' great acres.' They are also divided into ferlings, but we are nowhere told how many ordinary acres they may have contained. In the absence of any information given for Cornwall of the number of acres in the great acre or carucate or in its ferhng we might perhaps at first sight be tempted to infer that as in the Extent of Beri, a Devonshire manor of the same Pomeray estate, it is stated that there were 16 acres to the ferhng, so there might be 64 acres in the Cornish carucate. But this inference is doubtful, for in the Extent of another Pomeray manor in Devonshire (p. 187) — Stockelegh Pomeray — the ferhng seems to have contained 32 acres ^ which would give 128 acres to the carucate. Nor is there any mention in the Extents of these Devonshire manors of the existence of the Cornish great acre or carucate. 1 The list of free tenants consists of 7 holding each a ferling and each paying 5s. per annum and 12d. pro servicio ; 1 holding a J ferling paying 2s. 6d. per annum and F 6d. pro servicio ; 2 holding 8 acres each paying I5d. per annum and 3d. pro servicio ; 1 holding 4 acres paying l^d. 66 The Cornish Camcate. We remain therefore witli no certainty as to the number of acres in the Cornish great acre or carucate, and even if the number of acres were known we should still have to ask what acres they might be, whether Enghsh statute acres or Cornish or Devonshire cus- tomary acres. Lastly, there is reason to believe that the area of the Cornish acre or carucate varied. Even in the Exon Survey, as already mentioned, it was not always described as to be ploughed by the same number of carucw. And when we come to the statements in the ' Caption of Seisin ' 11 Ed. III. describing the manors of the Duchy of Cornwall, we find the number of Enghsh acres in the Cornish great acre varying considerably and justifying the statement of Carew in his ' History of Cornwall ' (p. 36) that whilst ' commonly 30 acres make a farthing land and nine farthing lands a Cornish acre, the rule is overruled to a greater or lesser quantity according to the fruitfulness or barrenness of the soil.' On the whole therefore it seems better to regard the Cornish great acre or carucate as of varying and uncertain area, and thus, for our purpose, out of court. n. THE HIDE. The hide, however various it may have been or may have become, as a unit of assessment ' ad geldum,' when considered as an actual agricultural area ' ad arandum ' came to be generally reckoned as containing four virgates or yardlands. In a period and in districts of settled agriculture, the hide had thus become ahnost identical with the carucate, connected with the ' caruca ' or team of eight oxen, The Hide. 67 yoked four abreast. The virgate in the same way had normally become the subordinate holding which con- tributed a pair of oxen to the common plough-team of the hide. Further, there was at least a dominant tendency to regard the virgate as a holding of normally 30 acres, making the normal hide an area of 120 acres of arable land. Now 120 statute acres (equal in area to 4855 ares) are so closely equivalent in area to 100 Cornish custom- ary acres (4820 ares), that we may reckon them ahke. And as the length of the furrow of the Cornish acre is Y^th of the leuga, the hide of 100 Cornish acres would be i^th the area of the square leuga, and hke the Cornish acre retain the shape of 1 x 10. Hence, when the clause in King Ine's laws which fixes the ' feorm ' of the unit of ' 10 hides to fostre ' or food-rent, is considered and connected with the numerous gifts of 10 hides or multiples of this unit by King Ine to the monasteries, the correspondence in area of the 10 hide unit of food-rent or tribute with that of the square leuga becomes remarkably significant. CHAPTER V. GENERAL CONCLUSION. The origin of the hide rests still in obscurity. But besides this remarkable relation to the leuga the question remains what, apart from its normal area, may have been its possible economic relation to the Welsh and Irish units of tribute examined above. It seems worth while before dismissing the subject to try to sum up the results of the foregoing inquiry as to these units, by way of realising the lesson to be learned from them and their relation to the hide viewed hke them as an economic unit and from a tribal point of view. Regarding the Celtic units or areas of tribute in their relation to each other, perhaps the most remarkable fact may be the association both in Ireland and Wales of the single plough-team with the maintenance of a herd of 25 cows and the connection of these combined elements with the tribute-paying group and area. This was not merely a chance result of Welsh custom. In Wales it was traditional in both Cymric and Goidehc districts. In Ireland it cropped up again in connection with both the tir-cumails. The small tir-cumail was the traditional area of good pasture allotted practically to the same typical herd, and the cumal of 3 cows was the annual return for the grazing of it. Admitting the variation within limits Celtic Units of Tribute. 69 in natural measures, we have found that Dunwal's maenol of North Wales contained nine of these smaller tir-cumails, the Dimetian tref four of them, and the Scottish plough-land of ' auld extent ' two of them. Here then keeping within the Unes of Celtic custom itself we find a direct connection between the unit of gwestva or tribute payment and the economic group with its typical plough-team and herd of 25 cows, and also we find that at any rate in Wales the normal areas occupied by this economic group, with its plough-team and herd, and charged with the payment of the tunc pound, turn out when examined to be divisions of the square milltyr of ' the great measurer ' who is stated in the Venedotian Code to have measured the ' island ' for the purposes of the ' mal ' or tribute. It is further worthy of remark that in nearly all cases the divisions are such as when geometrically reahsed take the prevalent form of two squares, as do also both the Irish tir-cumails. At any rate (whatever their origin) the independent evidence with regard to the tribute-paying areal imits of Ireland and Wales leaves us with the impression that, however many details remain in doubt, we have been dealing substantially with reahties and with widespread results of early Celtic custom common to both sides of St. George's Channel. With these Celtic examples in view we may recur once more to the Hide. Whether of foreign or of Celtic origin and whatever its subsequent Saxon and manorial relations may have been, we surely shall not be far wrong in regarding it as having grown out of an economic unit for payment of food-rent and tribute more or less like those of the Celtic group though probably belonging 70 Celtic Units of Tribute. to a more advanced stage of agricultuxal and economic life. The possibility must be admitted that it may have been to begin with, like the Irish plough-land and the Welsh trefgordd, the normal unit associated with the single plough in an agricultural rather than pastoral district ■under economic conditions widely different from those of a mainly agricultural population settled in village communities with several ploughs and plough- teams engaged in permanent open field husbandry with scattered holdings in yardlands. We seem even to get ghmpses of the stages by which the economic change from pastoral to agricultural conditions may have come about in the gradations of area in different districts in Wales. We trace in Wales what seem to be survivals of gwestva payments which go back to tribal conditions before the Norman or Saxon Conquest. The areas differ, but the commuta- tion into the tunc pound of the gwestva payments is the same throughout. Here again coincidences may easily deceive, but Mr. Maitland himself, in the concluding portion of his ' Domesday Book and Beyond,' would not let us rest in regarding the hide merely as a plough-land of 120 acres connected with a team of 8 oxen without further suggesting that it was also a ' fOund-faying unit.' After giving the figures of the valets of twenty counties in the Survey he adds : ' No one can look along these fines of figures without fancying that some force, conscious or unconscious, has made for one. pound one hide ' (p. 464). Whether this be so or not (and it must be allowed that it is by no means a imiversal rule) in passing from the Welsh and Irish to English units of taxation it will General Conclusion. 71 be hard, I think, when we set them in a row, not to recognise that the hide and the Scottish plough-land might well find a natural place at the end of the list. We cannot but recognise, I think, that the dwindling figures of area very instructively mark not only the passage from a mountainous to a more level district, but also a growing preponderance of the arable over the pastoral element even in Wales tiU the tribal ' co-aration of the waste ' had at last to give place to the settled agriculture of a village community gradually formed by the aggregation of homesteads and plough-teams for protection and convenience in the advanced economic stage vmder which it had become the typical form of rural occupation. The following table can hardly from this point of view fail to be instructive : — Eeserre Axel English Acrea for pasture after 120 acres arable Howell's Venedotian maenol 37,300 930 810 Dunwal's „ „ 13,886 347 227 Irish single plough-land 11,370 271 151 Gwentian trev .... 8,887 222 102 Dimetian trev .... 6,172 154 34 Domesday hide and Kentish sullung ..... 4,855 120 — Scottish plough-land . 3,258 80 — Bearing in mind that under both Irish and Welsh custom the herd of 25 cows was deemed to be the proper pastoral associate of the single plough, and further that not only in the case of the English hide and carucate, but also even in the Irish case stated in the ancient poem, the area required for arable seems by custom to be reckoned as 120 acres to the plough — it is clear that there must have come a time when these traditional figures would be found to be conflicting. Let us deduct 72 Celtic Units of Tribute. from the area of the £1 paying xmits of food-rent or tribute 120 acres for the arable and in a third column state the number of acres of pasture left for the herd of cows over and above the stubbles and fallows of the area actually ploughed. It surely is instructive to trace the extent of the change which the character of different districts or stages of tribal growth may have made in the complex economic unit associated with the single plough and responsible for the customary food-rents, commuted into the tunc pound. One is disposed to think that in the case of Howell's maenol of four trefs paying the tunc pound we ought possibly to regard it as an innovation of exceptional character. If so, to avoid exaggeration, we may reckon that in North Wales, outside the 120 acres of arable, something over 200 acres would be left besides the stubble and fallows for the pasture of the herd of 25 cows and the oxen of the plough-team. In the Irish plough-land of the Brehon poem about 150 acres were apparently left over in the same way for the pasture of a similar herd and the oxen. In the Gwentian district, bordering on, if not including, the rich plain of the Severn Valley, something over 100 acres were con- sidered enough to be left over for the past\ire of a similar herd ; whilst in Dimetia something like 40 acres were considered enough. And then, at last, we come to the contrast afforded by the hide. Let us examine more closely what it was from the same point of view. To each of its four yard- lands the typical outfit was something more than the two oxen for its share of contribution to the plough-team of eight. In the ninth century will of Abba (Birch, General Conclusion. 73 'Cart. Sax./ 412) the gift of a half-swulung (two yard- lands) carried with it 4 oxen, 2 cows, and 50 sheep, i.e. 8 oxen, 4 cows, and 100 sheep to the hide. According to the tenth century ' Rectitudines ' 2 oxen, 1 cow, and 6 sheep were allotted to the yardland. According to the twelfth century Grlastonbury Inquisition, 2 oxen, 1 heifer, and 1 cow and 6 sheep were so reckoned. ^ So that, marking the preponderance of agriculture over pasture, the typical herd associated with the hide or plough-land apparently all over England as compared with Ireland and Wales has dwindled down from 25 cows to 4 cows besides the 8 oxen of the plough-team. Surely this means that whether the hide as an assessable unit be regarded as of foreign importation or as a Celtic survival, its dwindled herd and restricted area mark very clearly the line of distinction between the pastoral stage of tribal hfe of the West and the settled agriculture of the village system of the East of Britain, already noticed even by Caesar. Returning to the evidence of the Domesday Survey, having translated the areas of the series of plough-lands into EngHsh acres, it is easy to compare them with the maximum number of acres possible to each actual Domesday plough-team as given for the various Enghsh counties in Mr. Maitland's table. I have ventured to transfer his figures to a map to facihtate comparison. If we compare the Gwentian tref of 222 acres with the figures for the adjoining EngHsh counties and the larger Venedotian areas with those of the forest districts of Shropshire, Staffordshire, and Derbyshire, the compari- son, I think, is instructive.^ 1 Roxburghe Club ed. 1882. " Maitland, Domesday Book and Beyond, p. 402. Fig. 14. Number of acres per plough-team. Sloping figures represent acreage divided by actual plough-teams in Domesday Survey. Upright figures represent acreage divided by estimated teamlands. — Maitland, Domesday Bock and Beyond,'^. 402. Celtic Units of Tribute. 75 These figures show clearly that already at the time of the Domesday Survey the number of plough-teams in the counties of Central England could not have been doubled without curtailing the 120 acres allotted to each, whilst naturally the two districts where the plough-teams absorbed the smallest proportion of the total acreage were the two great forest districts of Sherwood and the Weald. Thus in passing from the Irish and Welsh units of tribute to the hide we seem to have silently turned over an important page in economic history. The Celtic plough-team of 8 oxen keeps its place. It survives the great economic change and is the con- necting link between the old and the new conditions. If the hide was ever itself a Celtic unit of tribute, as- sociated like the others with the plough-team of 8 oxen, then it may be that the hide carried the plough-team with it as the new page in economic history was turned over. But if the hide came into Britain with conquerors from abroad, Saxon or other, then whatever it may have been originally as an economic unit or as a unit of assessment, it must have met in Britain with strangely tenacious traditions of earher Celtic custom. It must have had to succumb so far to their influence as to have been forced to reshape itself and its divisions upon the basis of the 8-oxen plough-team. In either case this plough-team accompanied and survived the turn over of the economic page. However this may have been, we have to recognise that the Celtic team of 8 oxen originally used in the Welsh tribal ' co-aration of the waste ' has become the great manorial plough-team with which the tenants of a manorial estate do service in ploughing on the lord's 76 Celtic Units of Tribute. demesne. If we are to look upon this from a tribal point of view, it seems to suggest a new method of paying tribute to a chieftain who has either him- self become a manorial lord or been supplanted by one. When and how this change was effected from tribal to manorial conditions in the South-Western parts of Britain we do not know. Whether the result of con- quest or of gradual agricultural evolution we cannot tell. But if in the days of Ine weekwork on the lord's demesne had already come to be regarded as a semi- servile method of paying gwestva or food-rent we may judge that the passage from tribal to manorial conditions may have been the result of conquest. And yet as in the structure of the hide the 8-oxen plough survived as the connecting hnk between tribal and manorial con- ditions, so possibly also the translation by Bede of the Saxon word hide or hiwisc into the Latin familia may serve a somewhat similar purpose. May there not be found in the word familia a reminiscence of something like the Welsh trefgordd — ^the tribal group of occupants of the pound-pajdng tref with their single 8-oxen plough and their normal herd of cattle— rather than the allotment or family holding of the single Saxon ceorl ? However this may be, we seem to stand upon firm ground when, going back to King Ine, we find in his unit of ' 10 hides to fostre ' reminiscence of the tribal system of food-rents, whilst in his mention of yardlands we recognise an indication that the hide had already for some purposes become connected with the 8-oxen plough- team. And thus we are led back once more to the fact most relevant for the moment to our present purpose, General Conclusion. 11 viz. that the normal area of the hide and ' carucate ' reckoned as 120 statute acres seems to have been yo of the square leuga and the 10 hide unit of Ine's laws therefore equal in area to the square of that great Gallic itinerary measure. To face p. 79 PART II. THE OLD BEITISH MILE. Note. — In the annexed map I have tried to show the length in metres of the ' Old British mile ' as calculated from the ' vulgarly computed ' and ' measured ' distances in statute miles in Ogilby's Itinerary or Road Book. The Gallic leuga of 1500 Roman paces was 2220 metres. The English statute mUe 1609 metres. The examination of the Welsh areas of tribute led to the conclusion that they were closely connected with the system of itinerary measures described in the Venedotian Code, and traditionally attributed to the period preceding the Saxon Conquest. The great itinerary measure was the milltyr of 1000 ' lands ' or strips for ploughing. The breadth of the land or erw was 9 single steps of 3 natural feet, making the milltyr 9000 single steps or 4500 double steps or paces. Reckoning the natural foot of an average middle- sized man at "251 m., according to the incidental evidence of the laws of King David of Scotland, the length of the double step or pace of six such feet would equal 1-506 m. and the milltyr of 4500 paces would measure 6677 metres. The GalUc leuga was reckoned by the Greeks and Romans as 12 Greek stades of 185 m., or 1500 paces G 80 The Old British Mile. of 1'48 m. according to Roman standard, and thus equalled 2220 m. The Welsh milltyr of 4500 paces would divide into three such leugse of 1500 paces or 2226 m., the variation in actual length expressed in metres being obviously within the limits of the inevitable variation in natural measures. If we had nothing but this evidence of the Welsh Codes to guide us, it would be difl&cult not to raise the question whether after all the leuga of 1500 natural paces may not have been the ancient customary itinerary measure of Britain as it was that of Gaul, during or even hefore the Roman occupation. But there is other evidence of an altogether inde- pendent kind from which I think the same inference may be drawn. It has long been known that when in quite recent times distances on the roads came to be systematically ascertained from actual measurement by the wheel in statute miles, the figures were found to be very much larger than those of common com- putation as recorded in the old itineraries and on the old milestones.^ It may be worth while to consider the meaning of these discrepancies. It will be convenient to begin with the fact that there still exist in Yorkshire (and probably elsewhere) old milestones the distances marked on which are obviously based on a mile much larger than the statute mile. From information kindly furnished to me by a local 1 If we may trust the Enoyolo- pjedias no one seems to know when the statute mile came into obliga- tory or common use on the roads. But the reader of the article on ' Mile ' in the Penny Cycloprndia will find a very fair statement of the discrepancy already noticed between the figures of modern measurement and those of the old itineraries. The Old British Mile. 81 antiquary^ I have been able in 19 cases to compare the distances recorded on old milestones with the actual distances in statute miles carefully obtained from the Ordnance map. The several distances marked on the milestones being in even miles, necessarily are not exact, but a com- parison of 117 milestone miles covered by them with the 162 miles of actual statute mileage ought to yield a fair average result. 162 statute miles of 1609 m. = 260,658 metres, and divided by the 117 milestone miles the average milestone-mile becomes one of 2221 m., i.e. the Gallic leuga of 1500 paces.^ Nor does the value of these figures rest here. If we divide the hst into smaller divisions, we may get some practical guidance as to the extent of the variations which may be expected in the average length of a customary mile derived from distances handed down by tradition or locally computed by the number of paces of an average walker, 1500 paces being reckoned to the mile. Divided into five groups as they happen to come in the list, the average miles vary from 2104 to 2308 m. We may therefore expect that the old customary mile wiU be found to easily vary more or less than 100 metres above or below the normal 2220 m. of the Galhc leuga of 1500 paces reckoned upon the Roman standard. It is true that the unit of itinerary distance is called a ' mile,' but the evidence of the milestones certainlv 1 Mr. J. J. Brigg, of Kildwick Hall, near Keighley, Yorkshire. Since published ; see ' Some Old West Riding Milestones ' in York- shire ArchcBological Journal, Vol. 3 A further hst, since kindly- added, covering 83 old milestone miles and 121 J statute miles, brings an average of 2356 for the old milestone mile. 0,2 82 TU Old British Mile. would lead us to regard what is called ' a mile ' as a measure of 1500 paces rather than a ' mille passuum.' In attempting to trace back the mile of the old milestones little help can be got from the Statute Book. The first mention of what seems to be the statute mile is in reference to building restrictions in a statute of 35 Ehzabeth in which the ' mile ' for the special purpose is described as 8 furlongs of 40 perches of 16 J feet, as in the case of what is now known as the ' statute mile.' Nor are we informed when the present statute foot as a fixed standard measure came iato use. On the other hand the ancient use of the leuga, rather than the mile, rests upon substantial evidence. Whatever doubts may arise as to the exact actual length in early times of the perch or the pace or foot we have historical evidence of the use of the leuga on this side of the Channel maintaining its ancient structure as 12 stades or 1500 paces — as in Gaul. There is first the evidence of the Domesday Survey in which the extents of woods, &c., are described not in miles but in ' leucce ' of 12 ' quararttence.' There is also the thirteenth century evidence of Walter of Henley, who, in order to show that an acre of land could be ploughed in a day, calculated the distance to be travelled by the plough-team in doing it, not in miles but in leagues of 12 furlongs of 40 perches of the King's perch ' of 1^\ feel. (' Husbandry,' ed. Lamond, 1890, p. 9.) What the length of the King's perch may have been we do not certainly know, for still there may be doubt as to the length of the foot. But the fact is important that supposing the foot to have been the modern statute foot of '305 m. the league of Walter The Old British Mile. 83 of Henley's statement would be 2400 metres in length, instead of tke normal reckoning of 2220 m. on the Roman standard. It would appear, however, from ' Harrison's Descrip- tion of Britain ' prefixed to the 1577 edition of ' HoHns- hed's Chronicle,' that the standard of the foot was not yet fixed even at that date.i The author gives in a diagram the exact length both of the English and of the French foot. And as his French foot correctly measures -325 m., i.e. the French standard foot, it is remarkable that his Enghsh foot should measure only •298 m. {i.e. the length of the Roman foot) instead of •305, the Enghsh standard foot. Walter of Henley's league of 12 furlongs of 40 perches of the King's perch of 16| feet (and presumably also the Domesday Survey league of 12 quarantense, if founded on the same King's perch) may therefore have measured 2344 m. instead of the 2220 of the Gallic and Roman standard. But however hazy Harrison may have been with regard to statute miles and feet, what is more to our purpose is that he was quite clear in his knowledge of the existence of an ancient well-known customary measure which he called ' the old British mile ' and that it was a mile of 1500 paces, though he was probably unconscious of the fact that if the standard of the paces were the same the mile would be identical with the GalKc leuga. His words are these : — As for the old British mile that includeth 1500 paces Enghsh, it shall not greatly need to make any discourse of it, and so much the less sith it is yet in use and not forgotten among the Welsh men, as Leland hath noted in his Commentaries of Britain. Moreover, he added to his treatise a copy of a 1 The later editions omit the information here giyen. 84 The Old British Mile. valuable Itinerary in which the distances (apparently unconsciously to him) when carefully examined seem to be given in the old British mile of 1500 paces and not in the statute mile. This valuable Itinerary is headed ' How a man may journey from any notable town in England to the City of London or from London to any notable town in the realm.' It is not known who was its author. Commencing as it does with ' The way from Walsingham to London,' it may well go back to the days when pilgrimage was the most typical form of travel, and the pilgrim the typical traveller. But what is most to our purpose is the fact that when Ogilby, nearly 100 years later, at the command of Charles II. in 1675 made ' an actual survey and men- suration by the wheel ' he found that the actual distances in statute miles exceeded very largely those recorded in Harrison's Itinerary. (Ogilby's ' Traveller's Guide,' London.) Where the routes coincided he adopted Harrison's figures, calling them the distances by ' vulgar computa- tion.' In his celebrated work he records for each route (1) the direct ' horizontal distance ' (as the crow flies) ; (2) ' the vulgar computation ' ; (3) the actual ' dimen- suration in statute miles.' His total for the direct routes out of London ex- tending in all directions as far as Berwick and Carhsle, St. David's, Holyhead, and Land's End, and also a good many direct shorter routes, were : — As the crow flies .... 3781 statute miles. By ' vulgar computation ' . . . 3507 miles. By actual measurement . . . 4536 statute miles. These figures would make the average customary The Old British Mile. 85 mile 2069 m. But Ogilby Mmself became conscious that as these direct routes out of London passed beyond the London district the difference between the figures of vulgar computation in Harrison's lists and the measured mileage in statute miles became greater than was the case within that district. In order to test the result of this fact I have roughly separated the London ends of the longer routes and the shorter direct routes out of London from those beyond the London influence. The separated figures work out thus : — By vulgar compuca- tion. By measure- ment. Making the average computed mile. For 24 cases in London district For the remainder of the routes 1130 2377 3507 1355 3181 4536 1928 m. 2152 m. Thus the average customary mile on the country ends of Ogilby's long direct routes becomes raised by this somewhat rough division from 2069 to 2152 m. Passing from the direct routes out of London, Ogilby gives in two further divisions particulars of what he calls cross roads. He has no longer Harrison's figures to assist him, so that presumably the figures of ' vulgar computation ' on these cross roads were the result of local repUes to his own local inquiry. His figures are as follows : — r As the crow flies . .2019' Group \ By vulgar computation . 1831 No. 1. 1 By measurement in statute (, miles .... 2505 . Making the cus- ^ ternary mile >■ 2201 m. of computation ) Group No. 2. ( As the crow flies . . 505 | ^^^ ^^^ ^^^ By vulgar computation . 475 V ^ ^.^^ ( Measured in statute miles . 636 J ■;} 2152 m. 86 The Old British Mile. From Ogilby's and Harrison's figures taken together the general result seems therefore to point to if not to prove the prevalence throughout the country districts of England and Wales of a traditional customary mile very closely resembling the Gallic leuga of 1500 paces or normally 2220 m. in length. But the problem can hardly be stated quite so simply as this. For besides the shortening of the computed mile in what may be spoken of as the London district there are exceptional variations in some other special districts which may mean disturbance .from without, and thus have historical interest. An examination of the figures on the map will show that in Norfolk and Sufiolk and down the Icknield Way as far as the rich Aylesbury vaUey, under the chalk down of the Chiltern range, the average computed mile rises to something like 2400 m., thus roughly corre- sponding with the league of Walter of Henley's ' King's perch of 16| feet,' i.e. IJ Enghsh statute miles. Then again in the figures both of Harrison and Ogilby on the northward course of the route from London to Cockermouth and CarUsle there is a distinct rise in the length of the computed mile as Newcastle- under-Lyme and Warrington are approached, and then a sudden drop in the figures to 1768 m. all through Lancashire, the higher figures being resumed again from Lancaster to Kendal and on to Carlisle. At this moment I am not attempting to explain these variations, I am only pointing them out. They may have historical importance or not. But as a matter of caution it may at least be well to consider how far (except in the case of the London district and that of Lancashire) these variations exceed The Old British Mile. 87 what might fairly be expected in distances measured by the natural methods on which the computed figures must have been based. It seems most probable that the distances by ' vulgar computation ' must have been arrived at either by the time commonly taken to travel the distance between the two places or at best by counting the number of steps or paces of an ordinary walker. Then, again, the steps and paces even of the same walker would vary with the uphill and downhill of the route and between the straightforward travel on a well-made road and the irregular stepping on a packhorse track. If the normal mile of 1500 paces, i.e. 1\ miles of Roman standard, was 2220 m. the single pace of the walker would be 1'48 m. The mile of 2400 m. would be covered by 1500 paces of TG m. The difference in the pace in the two cases would be httle more than 4 inches, i.e. 2 inches in the step. Again if in some cases the computed mile sank to 2100 m., the difierence would be covered by a drop of little more than 2 inches in the pace or one inch in the step. Recognising that in the computed distance we are deahng with natural measures and methods I think it may be fair provisionally to conclude that if no other explanation can be found for them there is nothing in the variation of the miles of ' vulgar computation ' in different parts of England to lead us necessarily to doubt Harrison's statement that the ' Old British mile ' was of 1500 paces or that allowing for local differences in the pace this old British mile and the Galhc leuga were identical. Even in the case of the shorter computed miles of 88 The Old British Mile. the London district the difierence in the pace would only be 7J inches and in the single step only 3f , Or the diminution in computed distance might be the result of the formation of better and straighter roads required by the increasing traffic as the metropolis was approached, and resulting naturally in a lengthened step of the walker. So that we may well take warning that too much must not be made of differences in the figures which, after all, in no part of England sink at all nearly to the statute mile. Without claiming absolute accuracy in the attempt I have made to place on the map the length of the customary mile of computation from a dissection of Ogilby's figures, I think the general diffusion all over England of a computed mile hovering round the Roman estimate of the length of the leuga of 1500 paces at 2220 m. will be accepted as pointing very strongly to the prevalence of this great historical itinerary measure in Britain as well as in Gaul. Before passing from Harrison's and Ogilby's Itineraries it may be worth while to recur to the York- shire milestones. Harrison's route from London to Richmond passes through the same district. His route from Hahfax through Keighley and Skipton to Carleton, included 34 computed miles against 47f statute miles, making the customary mile of the district 2240 m., thus so far incidentally confirming the evidence of the local milestones. There is also another route the figures of which may perhaps have a special interest. The Itinerary of Antonine gives the distances on the few British routes included in it, seemingly in Roman miles. There is no appearance of reference to The Old British Mile. 89 any other mile than the Roman * mille passuum.' Nor do the totals of each route which can be identified in detail suggest any other. They never, I think, fall short of the distance as the crow flies, as those of the ' vulgar computation ' constantly do. They seem very fairly to represent distances in Roman miles of 1000 paces where the position of the stations can be identified. Now if there could be any district where one might expect the Roman mile to have become rooted in the local mind by common acceptation during the centuries of Roman rule it surely must have been that in the neighbovirhood of the Roman Wall. Quite independently of Harrison, Ogilby gives the details of the route along the Roman Wall from ' Tinmouth to Carlisle ' as follows : In total — As the crow flies 59 statute miles. The vulgar computation Measurement in statute miles . 50 miles. . 69 miles 5 furlon The details are as follows : — Computed. Measured. Miles. Furlongs. From Tinmouth to Newcastle 7 9 2 on to Ovingham Hexham 8 8 11 10 7 4 „ Haltwesel 10 14 7 „ Chapelbourn „ Corby . „ CarHsle 6 7 4 50 8 9 5 2 6 1 69 miles 5 furlongs. From these figures the resulting customary mile would be 2236 m. Could there be a more striking instance or illustration pointing to the fact that down to the sixteenth and seventeenth centuries the old British mile of popular customary reckoning was not 90 The Old British Mile. the mile of 1000 paces, but a leuga of 1500 paces like that prevalent in ancient Gaul ? There remains the question how far other evidence exists filhng up the gap between the Itinerary included in Harrison's ' Description ' of 1577 and that of Walter of Henley and the Domesday Survey. In the first place the Itinerary pubUshed in 1577 in Harrison's ' Description ' is only a copy of one already known at the time ; for in 1579 a French guide- book was pubhshed in Paris for the use of French travellers in England, and it contained a copy of most of the chief routes taken from the same source as Harrison's with precisely the same figures, except in a few cases of evidently accidental errors in copying. In some of these the French copy is clearly the correct one, showing that the French copyist had access to the original source and was not copjdng Harrison. There can be little doubt that Leland had something to do with the common source. Henry VIII. had instructed him to travel and to collect materials, and wherever he went, mostly on horseback, he recorded the distances in his note-book with other information. When he arrived at an important place he apparently inquired what were the distances thence to the various towns all round, and the answers he received were evidently in old customary miles as locally known. When, for instance, he arrived at Worcester he made the following notes {inter alia) in his note-book, of what he had learned as to the distances from Worcester of places within 20 miles round. Leland's miles. Now reckoned locally in statute miles. Hereford . 20 26-28 Ludlow . . 20 30 Gloucester . 19 26 The Old British Mile. 91 Leland's miles. Now reckoned locally in statute miles Evesham . 10-12 16 Alcester . 12 17 Bewdley Tewkesbury . Bromsgrove . Winchcombe . • 12 (long) . 12 . 12 . 18 15 16 13 21 9 distances . 137 182 — ^making the customary mile 2138 m. The old British mile of 1500 paces was evidently the ordinary itinerary unit of local usage. Through the kindness of the Rev. J. M. Wilson, Canon of Worcester, I am furnished with independent confirmation of this. I am able to give the following distances from the monastery of the churches whose livings belonged to the Priors of Worcester, as stated in the MS. A. xii. in the Cathedral Library. MiUiaria as reckoned in 151G-1532 A.D. Now reckoned locally in statute miles Wolverley 13 17 Tibberton 3 5 Himbleton 5 n Stoke Prior . 8 10 Cropthome . 9 12 Overbury 10 17 Sedgebarrow . 12 18 Lenchwick 12 15 72 101 — ^making the customary mile 2257 m. It seems to be clear then that the mile of local knowledge and reckoning was the mile of 1500 paces and not the Roman or other mile of 1000 paces. The reader need not be troubled with further details. It is enough to say that the results of Leland's figures throughout his travels, where I have been able to verify them, correspond very closely with those of the Itinerary of 1577— thus carrying the ' old 92 The Old British Mile. Britisli mile ' of Harrison back another half-century, to 1535-1543 A.D. Still further back than Harrison and Leland the Itinerary of William of Worcester (a.d. 1478) affords independent evidence chiefly with reference to West Wales. Travelling in Devon, Somerset, and Cornwall, as far as I have been able to judge somewhat roughly from the map, he reckons stages covering about 355 statute miles, as 250 miUiaria, so making the customary mile of West Wales 2285 m. Here again, hidden within the average. He considerable variations in the details of the several journeys included in it. W. oJ W. Stat, miles. Somerset . 41 54 making the average mile 2118 m. of computation Devon . 38 59 » 2490 m „ . 55 72 „ 2106 m ComwaU . 77 . 23 . 16 111 34 25 2319 m 2380 m 2514 m I must, however, admit the difficulty of following the distances in Cornwall so as to translate them with exactness into statute miles from the map. The above figures must therefore be taken with this reservation. Of the Domesday Survey, with its 12 quarantense of 40 poles, we have already spoken. With regard to the Saxon period we seem to be thrown back upon the evi- dence of the pre-Domesday Venedotian Code of the tenth century, which, whilst describing the areas of tribute as closely connected with the mUltyr and the milltyr itself in natural measures, throws back its traditional use to a period before the Saxon invasions. Lastly, when we find in the Brehon tir-cumails, also described in terms of natural measures, evidence of a community of The Old British Mile. 93 agricultural measures possessed alike by Irish, British, and Scottish peasantries, the evidence throws back the common use of the milltyr and the leuga historically to a very early date. And thus brought back to the point from which we started we may let the question of the identity of itinerary measures in ancient cus- tomary use in Gaul and Britain rest for the moment, to be taken up again when the customary acres on both sides of the Channel have been examined. Note : — Mr. Petrie in his article on Weights and Measures in the Encyclopcedia Britannica, considers that the Belgic foot of the Tungri (•333 m.) passed over as a building foot into Britain and entered into the old English itinerary measures, which he considers were based upon the furlong of the statute acre of 600 Belgic feet. Thus :— foot -333 m. yard -999 fathom 1-998 chain 19-98 = 20-1 m. furlong 199-8 = 201 m. mile 1998 = 2010 m. = old English mUe. For the 10-furlong mile he refers to his article in Proc. Roy. Soc. Edin. (1883-4). In this way he considers that ' we can restore the old English system of long measure from the buildings, the statute prohibition, the surviving chain and furlong, and the old Enghsh mile shown by maps and itineraries,' and adds that such ' a regular and extensive system could not have been put into use throughout the whole country suddenly in 1250.' Hence he favours the view of its Belgic ' origin. ' But why the perch should be divided into 5J statute yards the Belgic origin does not explain. The length of the Belgic foot of the Tungri is not in dispute, it is described by Hyginus as 18 digits of the Roman foot and is in fact the Greek foot of 4 natural palms. And if the evidence for the 10 fur- long mile were complete, this ' old EngUsh ' mile of 1998 to 2010 m. would perhaps afford a Belgic explanation of the itinerary com- puted miles of the London district. But Mr. Petrie goes on to say that the ' Oallic leuga is a different unit, being 1-59 British miles by the very concordant itinerary of the Bordeaux pilgrim. This ap- pears to be the great Celtic measure as opposed to the Old English or Germanic [? Belgic] mile. In the N.W. of England and in Wales this mile lasted as 1-56 British miles till 1500, and the perch of those parts was correspondingly longer till this centm-y. The " Old London mile " was 5000 feet, and probably this was the mile which was modified to 5280 feet and so became the British statute mile.' PART III. ON THE CUSTOMAEY ACRES OF BRITAIN, IRELAND, AND ARMORICA. CHAPTEE I. INTEODVGTORl. If the facts considered in the previous chapter may be rehed on as good evidence that Gaul and Britain shared traditionally in the common possession of their cus- tomary mile or leuga of 1500 paces we are brought at once to the main problem of this inquiry. It is the common possession of Britain, Armorica, and Ireland, of a group of similar customary acres on both sides of the Channel. It is of course a matter of common knowledge that in various districts in Britain there have been prevalent customary acres remaining in local use in spite of all the efforts made to drive them out and to substitute for them the statute acre. It is also well known that the customary arpents or acres of the great corn- growing Armorican districts on the other side of the Channel have held their own with equal tenacity. But the fact that gives importance to the problem is that the group of customary acres of the Armorican corn district are practically the same as those of Britain. The problem is not quite so simple as this. H 96 Introductory. I will therefore explain at once why I have used the expression ' f radically the same/ The form of the acres in nearly all cases is 1 x 10. But on both sides of the Channel a curious practice is observable of changing the form of the acre, without altering its area. The general form being 1 x 10, the acre is turned into two half -acres each of the form of 1 X 10. But the remarkable point is that this change of form occurs most generally, so to speak, in the act of crossing the Channel. Thus the Aimorican arpents or acres, corresponding in area to British 1 x 10 customary acres, become on the other side of the Channel two half-acres, 1 x 10, combined into an arpent 1x5. We are thus brought to consider as a prelimiaary question how and why in practice the change from the acre in the form of 1 x 10 to the two half -acres in the same form was effected. It is all the more needful to clear up this question at the outset inasmuch as it involves the curious further point that the change of form must inevitably dislocate the relation of the length of the furrow of the original acre to the itinerary measure. If the furrow of the full acre (1 x 10) be a given division, say, of the old British mile or leuga of 1500 paces, then the furrow of the half-acre m the same form must needs be a similar division not of the leuga, hut, tnathematically, of half the diagonal of its square. All this seems at first sight to complicate the problem. But these facts are not confined to British and Armorican agricultural measures. Historically they bring us at once into contact with a wider world. Let us go back to the fact that in our previous examination of Celtic areas of tribute we have been Customary Acres of Britain, Ireland, and Armorica. 97 resting exclusively on what we may call the natural system of measures— the thumb, the palm, the foot, the step of the ' middhng-sized man.' We have followed no other standards of length beyond the suggestion that the 10 Koman feet of the Roman decempeda may be taken to equal 12 natural feet of minimum standard, the natural foot varying within limits between about 246 and 252 miUimetres. The variation in the standard of the foot and pace does not and cannot alter the fact that the leuga of 1500 paces = 9000 natural feet and that of the Welsh milltyr = 27,000 natural feet. All these measures will share in the variations of the natural paces and feet. There were, no doubt, in various parts of the world many artificial and derivative feet and cubits which may seem to compUcate matters ; but whatever doubts and difierences there may be between metrologists with regard to their origin I think it will not be disputed that the prominent Greek and Grseco-Egyptian and Babylonian agricultural areal units were based upon this natural system of measures ; and that the hmit of variation in the natural foot was very similar through- out, whilst ultimately the common recognition and use of particular itinerary measures, whether we take that of the Roman ' mille passuum ' of 8 stades or 1000 paces, or whether we take the Galhc leuga of 12 stades or 1500 paces, formed a prominent Unk of connection and stabihty of standard between East and West all along the route of their prevalence. With regard to the similarity in the range of variation in the natural foot I wish I could transfer to these pages the beautiful example, given by Dr. Dorpfeld many years ago, involved in the ground plan of the two temples h2 98 Customary Acres of of Sunium. (' Mittheil. des Deutsch. Aich. Inst./ Athen, 1884.) The older temple was probably destroyed by tbe Persians. In the time of Pericles it was replaced by another on precisely the same site and built on precisely the same plan. There were the same number of columns, and these were the same number of natural feet apart from centre to centre. If the standard of the foot used in both temples had been absolutely the same, every new column throughout the temple would have exactly covered the circular foundation of the old one. But the standard varied shghtly, and as the end of the rows of columns was approached, the new colmnns more and more overlapped the sites of the old ones. There were stiU 10 natural feet from centre to centre of the columns, and the variation in the standard of the natural foot carried out with perfect regularity was that between "246 m. of the old temple and '251 of the new one. Thus by following all through our inquiry natural measures we seem to find ourselves so far upon a sub- stratum of common and sohd ground extending far back into the past. I do not think I run the risk of provoking any material difference of opinion on the part of leading metrologists if at this stage of the limited inquiry on which we are engaged I state the following facts : — (1) The Tdiet or Egyptian aroura was 100 Egyptian cubits square. Transformed into two squares (of half the diagonal of the square) it became the Greeh medimnus, i.e. the area suitable for the sowing of a medimnus of seed. (2) The Greeh aroura of 100 Greek cubits of 6 natural Britain, Ireland, and Armorica. 99 palms turned into two squares (of half its diagonal) became tlie Roman jugerum. (3) The Greek plethron of 100 four-palm feet (Uke those of the Irish tir-cumail) put into the form of two squares (of half its diagonal) became the modius or socarion for good corn-land, suitable for the sowing of a Greek modius of seed. (4) There was also in use a square of 120 four-pahn feet, and each of the two squares on half its diagonal became the vorsus of Italy ; and the double vorsus was in use as a foor or wet-land modius. We shall have to recur to these Egjrptian and Greek units, but in the meantime for the moment these facts of ancient agricultural custom will form precedent enough to prevent the reader from throwing up the Armorican problem in despair at the outset. It may be added that these transformations of similar areas from one square to two involved precisely the same dislocation of the relations to the itinerary measure 2000 years ago as they do in the case of the British and Armorican customary acres. It may be added further that the use of the common itinerary measures and those based on the diagonals of their squares will be found to be a fact of remarkable historical interest. CHAPTEE II. THE BRITISH CUSTOMARY ACRES. I. THE CUSTOMARY ACRES. The object of this section will be to examine tte British customary acres separately from the corresponding acres of the corn-growing district on the other side of the Channel, first, because it is important to under- stand their geographical position in Britain and also because they have their own relations amongst them- selves and to the larger areas of tribute, Irish and British, already examined, quite apart from their correspondence with the Armorican acres. We have seen already incidentally how intimately the erws and customary acres of Wales were connected with the units of tribute and food-rent, and at the same time how closely they corresponded with the actual methods of ploughing even in Wales, where pastoral interests were predominant and agriculture was in its primitive form of co-aration of the waste. With the exception of the Dimetian and Gwentian erws, the form generally arrived at was ultimately that of 1 X 10. But in the two cases of North Wales, so graphically described in terms of practical ploughing, the day-work was in the form of 1 X 30 — ^the 1 X 10 acre sliced into three. Or to put it the other way, and probably more correctly, the surviving customary acres Topxcc p. 100. Ancient Scof^ish Mik 2222 Ordinary ScoftishMikiSIO The. fi'jiireA are the length of ike Jurrowt tn metre*. (There no fujiiTti are aicea the tUUule acre [Jurrow HOlm.) may haxx- been the local acre or have completely ousted the cusUrniary acre$m Fig. 16. British Customary Acres. 101 of tlie district seem in some cases to have been formed by the combination of three of the more ancient day- work in ploughing put side by side. These acres with a long furrow and a very short end present a great contrast to those in the form of a square, which occur seldom, if ever, in Britain, and which are most characteristic of the regions where the ohve and the vine are the predominant objects of husbandry. But Meitzen, in his great work, has reminded us that even square forms may have direct connection with methods of ploughing. The Roman jugerum was in two squares, with an ' actus ' or furrow of only 120 Roman or 144 natural feet. It was ploughed by a single yoke of oxen and connected with a system of cross ploughing, for which a square was obviously most adapted. The short furrow and single yoke were most adapted also to ploughing in a hilly country and custom everywhere tended to follow convenience and local needs. (' Siedelung und Agrarwesen,' 1895.) The long furrow of the 1 x 30 and 1 X 10 acres was ploughed by a heavier plough and a larger team. In Wales and Scotland and manorial England the normal team" was of 8 oxen — and there was obviously no question of cross ploughing. If we were to attempt to reason from this that the long furrow, whatever its origin, and wherever found, is everywhere a proof of a long settled preponderance of the agricultural over the pastoral element, we should be going too far. For it is precisely in the still pastoral period of Cymric tribal life in Wales that we find the long furrow in customary use with the typical plough- team of 8 oxen, which afterwards was in general manorial use throughout England. 102 British Customary Acres. It is important to notice that it was thus first connected, not with the settled agriculture of a village community, but with a tribal stage of husbandry in which the primitive system was in vogue, called ' co-aration of the waste,' consisting of ploughing up portions of the waste each year and letting them return into pasture after removal of the crop. In the hght of the Welsh Codes the long furrow and the plough-team of 8 oxen, under Cymric tribal conditions, seem to be the result rather of the necessity of co-operation on the part of the tribesmen in what ploughing was necessary than in the exigencies of a settled agriculture. It is precisely because the customary acres of Wales go back to the tribal period and to pastoral conditions that the problem involved in them is fraught with a special historical interest. For Great Britain the evidence, as to the customary acres of different districts, though not so complete as might be wished, is sufficiently striking. The Report of the Agricultural Commission of 1820-1823, imperfect as it may be, records in Engfish statute measures — yards and feet — ^the chief of the customary acres of each district, as they came under the Commissioners' notice. Other evidence, too, is to hand, which, without pretending to absolute geographical accuracy, locates these acres in districts sufiiciently marked. To give a general view of the geographical position of these acres, I have tried to put the evidence into the form of a map (fig. 16), on which I have marked roughly the geographical range of this or that acre by figures which represent in metres the length of the furrow or long side of the acre, the normal shape of the acre being in all cases like that of the statute acre 1 X 10. English Statute Acre. 103 The wide blanks on the map, where evidence of any- local acre is absent, may probably very roughly be taken to mark the districts in which the statute acre either was the ancient customary local acre, or had supplanted others so completely that they had passed out of knowledge.^ II. THE ENGLISH STATUTE ACRE. Let us take notice, to begin with, that the statute acre itself does not fit in very happily with measurement in statute yards and feet, whilst it takes an easy place amongst those based upon natural feet. Five and a half yards, or 16J feet to the rod, does not seem very happy. On the other hand, the statute acre seems to find a reasonable explanation in the system of natural measures. The Venedotian Code has shown us not only in figures but also in graphic description of actual ploughing that the short end of Dunwal's ' land ' (tqtoo of his milltyr) was 27 natural feet, and that its furrow, being thirty times its breadth, was therefore 810 natural feet in length. As already stated, three such erws put side by side would equal the English statute acre in its form of 1 X 10. Its end would be 81 natural feet, 9 rods of 9 natural feet, and well within the variations of the natural foot ; 810 feet of "248 would easily change into a round number of 800 feet of '251 and correspond with its furrow of 201 m. We have seen that it was exactly so described in the Scottish fragments, viz. as 4 X 40 of a rod of 20 natural feet, thus making 1 See supra for Map, p. 100. 104 British Customary Acres. its fuirow 800 feet of King David's higher standard of •251 m. for the foot of the middling-sized man, i.e. 201 m. When examining the Scottish plough-land we noticed that to make its 40 statute acres fit exactly into the two squares of the lesser Brehon tir-cumail, of which the plough-land was double in area, the 40 acres would have to be turned into 80 half-acres. Then 80 half-acres (1 X 10) would exactly fill it. So that the Enghsh statute half-acre (1 X 10) was more directly connected with the Brehon tir-cumail than the statute acre itself. The furrow of this half-acre would be 142 m. in length on the higher standard of King David's foot "251 — ^and would equal just yg- of the leuga of the same higher standard. Thus the half-acre, being the one the furrow of which was a direct division of the ' Old British mile,' i.e. the leuga, the furrow of the acre itself, 201 m., would be a similar division of half the diagonal of the square leuga (viz. 1609 m.), which somehow or other came to be the length of the English statute mile of 8 furrows of 201 m. Passing from the statute acre let us examine those customary acres the furrows of which are most obviously direct divisions of the Old British mile or leuga. m. THE CORNISH AND DORSET ACRES. The Cornish acre, according to the Report of 1820, has a furrow of from 219 to 220 m., i.e. the furrow is i-Q of the leuga of lower standard, just as the prevalent Scottish acre was i\, of the leuga of the higher standard. The Cornish acre is described as 4 x 40 of a rod of 18 English feet, i.e. 5-5 m. or 8 x 80 of the local gad Cornish and Dorset Acres. 105 or goad of 9 English, feet or 2 '75. Its area is 48 ares. The Cornish potato-grower still measures his ground by the ' lace ' of 4 x 4, i.e. 16 square Cornish gads. The map will show that the Cornish acre appears sporadically in South Wales. It also seems to have left intermittent traces of itself in Leicestershire and East Angha. Its historical importance Ues in the fact that con- verted into two half -acres 1^ X 10 in form it becomes the prevalent Breton arpent, and forms the most obvious and important Hnk between the British and Armorican evidence. The Dorset acre has a furrow, according to the Eeport of 1820, of 184 m., i.e. very slightly less than the stade of 185 m. It is therefore ^-g of the leuga of lower standard. It is 4 x 40 of a gad of 15 Enghsh feet and one inch. The rod is therefore 4"6 m. and the area of the acre is 34 ares. The Devon and Somerset acres with a furrow of 183 m., according to the Report of 1820, are so nearly identical with the Dorset acre as to be hardly distin- guishable from it. All these furrows are slightly shorter than what, as divisions of the leuga, they seem entitled to be, and probably the reason may be found in the description in the Report of 1820 of the Dorset acre. The Dorset gad or rod, as already mentioned, is described in the Report as measuring 15 feet and one inch, with a praiseworthy attempt at perhaps unusual accuracy, whilst the odd inch may have been omitted in the possibly less accurate information received by the Commissioners in regard to the Devon and Somerset acres. 106 British Customary Acres. This extra incli in the Dorset gad reminds us that, as in the Keport of 1820 and elsewhere nearly all the information is given in statute yards and feet without additional odd inches, absolute accuracy cannot be expected. IV. THE FOEEST ACRES. In Central England, in the Forest District of the Peak in West Derbyshire, another customary acre takes I 302 S Dimetian tref 6172 ares Rg. 17. the form of 10 X 100 Cornish gads and has a furrow of 275 m. In Sherwood Forest the length of the furrow is doubled, the acre being 20 x 200 gads — making the furrow 550 m. The areas of these two forest acres are 75*6 and 302"5 ares respectively. The half-acre of the 275 acre occurs in the form of 1 x 10 in Northampton- shire, and was apparently, though the evidence is not complete, the one in use on the Battle Abbey estates in Sussex. The Powys Acre. 107 These two forest acres are re- spectively in form and area -g^ and ■^ of the Dimetian tref . And their furrows are respectively ^ and J of the leuga of 2200 m., i.e. the lower standard. V. THE POWYS ACRE. The next customary acre claim- ing attention has a furrow of 292 m. The region over which it is pre- valent begins in Lancashire south of the Eibble, includes Cheshire and Stafiordshire, and stretches across the centre of Wales. From the position of the district covered by this acre it has been suggested that its home was the ancient Welsh district of Powys. The area of this acre is 85"264 ares. Its connection with itinerary measures seems to be that (put into the form of a square) it would be I of the square stade of 185 m., the area of which is 342 ares. In its 1 X 10 form it is divided into 4 ' cyvars ' or co-ploughings, and the ' cyvar ' is again divided into 4 ' quarts ' in Flintshire. Its furrow is \ longer than the Cornish furrow. It is 4 X 40 of a rod of 108 British Customary Acres. 7'315 m., and 3 x 30 of this rod would make the Cornish acre. This will be seen at once in the diagram. The Powys acre and the Cornish acre evidently belong to the same system. VI. THE NORTH WALES CUSTOMARY ACRE. Next we return to the important and instructive customary acre already mentioned as consisting of three of Howell's ' legal erws ' (1 x 30), three of which, when put side by side, made an acre of 1 X 10. As described in the Report, in EngHsh yards it had a furrow of 329 m. and an area of 108*241 ares. Howell's ' legal erw ' (36"08 ares) is still known in North Wales as ' the true erw.' It consists of 3 stangels, each in the form of 1 x 10. So that the 329 m. acre itself consists of 9 of these stangels, and the Cornish acre in form and area would contain just four of them. The acre of 329 m. is also divided in North Wales into 4 ' stangs,' measured 4 x 40 of a paladyr of 4*117 m., the paladyr being IJ of the Cornish gad. This ' stang ' or quarter of the acre is known in Montgomeryshire as the ' stang a dyr,' and in North Wales as the ' cyvar ' of Anglesea and Carnarvonshire. It is found again as the customary acre of Herefordshire. It has an area of 27 "08 ares, i.e. nearly exactly that of the Greek medimnus. According to the Report of 1820 two of the 329 m. acres, i.e. eight of these stangs or cyvars, in Glamorgan- shire and Pembrokeshire make an ox-land and 8 ox-lands a plough-land. The area of the plough-land would be 1744 ares and that of the ox-land 218 ares. The North Wales Acre. 109 Thus the plough-land contained in area 36 Cornish acres. It was in area half Dunwal's tref and | of his maenol, and -2^ of his square milltyr. The Cornish connection of this acre of 329 m. furrow is shown further by the fact that it can be measured 12 X 120 of the Cornish gad of 2-75 m. as well as 8 x 80 of the local paladyr of 411 m. which, as already said, was 1| of the Cornish gad. Its quarter (the cyvar of Anglesea, and the Herefordshire acre) is 4 x 40 of the faladyr, while the stangel of ^ its area is 4 x 40 of the gad. Recurring to the point above alluded to, that, owing to the absence of odd inches in the 1820 Report, absolute accuracy cannot be expected, it is fortunate that so often the survival of the local gads or rods by which the local customary acres were measured is reported. It shows that these acres were not originally based on Enghsh statute feet and yards, but upon their own local rods. VII. THE SCOTTISH AKO NORTHUMBRIAN AND CUMBRIAN ACRES. It wiU be seen on the map over how large a geo- graphical area these acres were in use (p. 100). The most prevalent Scottish acre — i.e. the Scottish acre, is described in the Report as having a furrow of 227 m., i.e. yq of the leuga of higher standard, and it may be mentioned here that the old Edinburgh mile is reported in the books to have been 2222 m. in length. This, of course, is a traditional recollection of the old British mile or leuga, which would thus seem to have 110 British Customary Acres. been at home in Scotland as in Britain and Gaul. On King David's standard of "251 m. for the natural foot this mile of 1500 paces would equal 2259 m. with which the customary furrow of 227 would nearly correspond. Moreover, this Scottish acre finds a sUghtly exaggerated counterpart on both shores of the Solway Firth and in the ancient Dabiada. It is there known as the Cunningham acre and is still in use in East Ulster. It has a furrow of 228 m. and from the fact that Dalriada extended into Scotland the historical connec- tion is obvious, whatever may be its relation to the Scottish acre of shghtly lower standard. It should be noticed also that but for the difEerence in standard this Scottish acre would be identical with the Cornish acre. A similar customary acre is reported as in use in Lincolnshire. And the similarity in area on the higher standard of this Scottish acre with that of the Roman sors of two jugera is a fact which may have a meaning, though any immediate identification might be misleading. There remains to be considered the important customary acre which whilst conspicuous in Britain as occupying mainly the Northumbrian and Cumbrian geographical position of the Celtic Brigantes, has, from its prevalence in Ireland, become best known as the ' Irish or Plantation acre.' There is a mystery about it which is not easily dispelled. (See below, p. 147.) Its area is 65"62 ares and its furrow is 256 metres. As we have seen, its half-acre in the form of 1 x 10, according to the early Scottish fragments, had a furrow of 181 m. and was important enough to give rise, according to the books, to a local or customary Scottish mile of ten times its length, viz. 1810 m. Scottish and Northern Acres. Ill It is difficult to connect the furrow of 256 m. with itinerary or natural measures ; and unless the furrow of 181 m. could be considered as a diminished stade of 185 m. it is difficult to connect it with any itinerary measure other than that to which it seems to have given rise. But if the furrow of the half-acre might be raised English statute acre in square 64- 128 Fig. 19. Square of the 256 m. furrow. to 185 m., so the furrow of the acre itself would have to be raised from 256 to 261 m. And it is significant that these figures would bring us to the further point that they would represent almost exactly rj,th of the figures of the greater tir-cumail regarded from the point of view of its dimensions on the highest and lowest standard of the natural foot. So difficult does it seem to be to get outside the meshes of the network of Celtic measures. I 112 British Customary Acres. Recurring, however, to the fact of its prevalence side by side in the north and in Wales with the English statute acre, it may further be noted on the other side that the square of its actual furrow 256 m. would contain 16 English statute acres put into the form of a square, as shown on p. 111. And, to go one step further, there might possibly be significance in the fact that whilst the English acre in a square would contain both in area and shape 8 Bimetian erws, the Irish acre itself put into a square would contain both in area and shape 9 Gwentian erws. So that the South Wales erws may after all be divisions of the Enghsh and Irish acres respectively in the form of a square. The wide prevalence of this mysterious acre is also remarkable. Though its chief geographical home seems to be that of Yorkshire and Cumberland and Lancashire north of the Eibble, it is sporadically present also in Northamptonshire, South Wales, and Cornwall. VIII. THE HALF-AGRES IN EORM OF 1 X 10. Allusion has been made already to a class of acres which are half-acres of other customary acres put into the form of 1 x 10. The half-acre of the Northumbrian 256 m. acre in use in Scotland, with a furrow of 181 has just been considered. The half -acre of the 329 m. acre also occurs in the form of 1 X 10. If we might raise this furrow from the lower to the higher standard, viz. 329 to 335, the Westniorland acre, with a furrow of 237, becomes its half-acre, whilst in the lower standard of 233 it occurs in the southern counties of England sporadically. The Half-Acres. 113 Anotlier of these half-acres also turns up in North- amptonshire. From a note on a fly-leaf of one of the MSS. of tha Record of Carnarvon it appears that contiguous districts in Northamptonshire had various customary acres, which were declared to be of larger size than that of the Priory of Fineshade. It was said that these larger acres of other districts ought to be ' set every acre for more money than Fotheringhay or the Priory of Fineshade/ whose acre was stated to be 4 X 40 of a rod of 16 feet/ i.e. with a furrow of 195 m. This Fineshade acre was in fact a half-acre of the neighbouring West Derbyshire forest acre whose furrow was 275 m. (See above p. 106.) This note is valuable as revealing that in con- tiguous districts of Northamptonshire there were three separate customary acres, viz. the Northumbrian acre of 256 m., the Cornish acre of 219 m., and the forest half -acre of 195 m.^ This accidental evidence also suggests that wherever 1 See also that of Harleston Estate Book of Henry de Bray. (Trans, of Roy. Hist. Sac., 1910, p. 123.) ' According to the measure of the perch on the gable outside the Chancel at Harlestone.' " Harleian MS. 696 ; see The Record of Caernarvon, 1838, p. xxii : — Me*, y' in the lordeship of Ape- thorp y* is nowe Williame Gruffith to hym & to his heires/and in al the townes y' weren or bene of the Kyngs holde w' in y' side of Northampton shire the polle con- tenes xxj fote by the stondart {the furlong therefore would he 256 m.] as John Rowlond told to the saide to the same Willfn, and oy] men of the saide Towne saiden y' the Polle contened xx. fote by the stondart tn quer. Also in the lordeship of ffodrjm- gay xviii fote gose to the Polle [furlong 219 m.J. And in the Prio' of ffynsehede is holde xvi fote gose to the Polle [furlong 195 m.]. Itm one Polle of brede and xl. Polles in lengthe makes a Rode of londe. And iiij. Rode gose to the Acre/and so Apethorp londe agtte to be set eu'y Acre for more money then ffodryngay or Prio' of liynsehede londe ffoi the acres bene more &c. l2 114 British Customary Acres adjoining districts with separate customs met there might well be a conflict of rival customary acres and half-acres of which the memory has locally gone to sleep, under cover of the prevalence of the statute acre. Finally, to wind up this statement of the customary acres of Britain, we seem to be driven by the evidence to regard them, whether of the lower or higher standard, as originally based upon natural measures — actual feet and steps of consequently shghtly varying length — ■ but in nearly all cases, either in the length of their furrow or their area, aiming at a definite relation to one or other of the prevalent itinerary measures — ^the milltjrr or its divisions, the leuga or the stade — ^the actual length of these being regarded as varying within similar hmits. In days of rough natural methods of mensuration there was no standard system of metres and milhmetres by which every measure could be defined and fixed. The Koman decempeda, no doubt, during the period of Eoman rule must have provided a fixed standard, and when it came into contact with natural measures it may, as suggested, have fixed the minimum of the natmral foot at '246 m. as yj of its length. But how and where was the standard Eoman decempeda to be found in practice by which the goads and yokes could be tested ? No doubt there may have been kept in various places, in churches and elsewhere, rods and goads which in each neighbourhood may have been regarded as local standards. The bat or eglyshaw may have been one of these, and long yokes would be made of a generally recognised length in each locahty, but it would be absiu:d to expect absolute identity of standard in actual practice. Still looking at the customary acres of the Southern and Irish Measures. 115 ' West Wales ' group, it is remarkable how consistently the tendency is shown to sink rather below than above even the minimum standard. This apphes to all the acres which were measured by the Cornish goad. In a remarkable way they seem to hang together in this as in other respects. On the other hand, the furrows of the statute acre and the Northumbrian or Irish seem to be built on the higher standard. And this is a fact which amongst others it will be well to bear in mind as one from which historical inferences may hereafter possibly be drawn. IX. THE CONNECTION OF BRITISH CUSTOMARY ACRES WITH IRISH MEASURES. There remain, before we pass from the British cus- tomary acres to those of Armorica, across the Channel, two points which must not be disregarded : (1) How far were these British customary acres prevalent in Ireland ? (2) What relation had they, if any, to the two greater units, the tir-cumails ? The evidence on the first point is clear. A document in the Carew MSS. of the time of James I. describes accurately in detail what were the customary acres that the English surveyors had to deal with in .Munster. They turn out to be three acres with which we are already famihar in Britain, those with the furrows of 256 m., 292 m., and 329 m., with an additional one of 354 m., which, in area and shape, is 5 Eoman jugera put end to end. To these may be added the Cunning- ham acre in Ulster, already alluded to, with furrow of 116 British Customary Acres. 228 m., which seems to belong to the ancient Dalriada on both sides of the Solway Firth. The question of course arises whether these British customary acres, being found present in Ireland in the sixteenth century, are to be regarded as importations by English immigrants or as indigenous in Ireland, and so survivals of conditions common to both sides of the Irish Sea. The relation of these customary acres to the great tir-cumail is very remarkable. The great tir-cumail of lower standard, i.e. reckoned on the natural foot of "246 m., contains 130,356 ares. It would therefore contain 1200 of 108-63, i.e. of the 329 m. acre of 108241 ares. 1500 of 86-90, i.e. of the 292 m. acre of 85-264 ares. 2000 of 65-178, i.e. of the 256 m. acre of 65-62 ares. 2500 of 52-14, i.e. of the 228 m. acre of 52-28 ares. 3240 of 40-23, i.e. of the 201 m. acre of 40-46 ares. 4800 of 27-16, i.e. of the 165 m. acre of 27-06 ares. The fact seems to be that, with the one exception of the 292 acre, unless the measurements of all these acres have been underestimated in the Report of 1820 by description in even statute feet and yards, they correspond remarkably as divisions of the greater tir- cumail of lower standard. CHAPTEE III. THE BRETON OPEN FIELD SYSTEM. In transferring our inquiry from tlie British to the Armorican agricultural region, it will be needful before confining attention in particular to the customary acres or other land units, to refer very briefly to the agricul- tural conditions under which they were prevalent. This becomes all the more necessary seeing that our first attention is turned to the extreme Western ex- tremity of Europe, the Land's End of the Armorican peninsula. It is not enough to say that its agricultural unit, or arpent, is practically the same in area as that of Cornwall — ^the extremity of West Wales. To understand the full meaning of this resemblance we must have regard to environment and the pecuhar form of the still prevalent open field agriculture of which the Breton arpent forms the agricultural unit. I cannot pretend to be able to do justice to this part of the subject, which would be worthy of more detailed study, but I venture to hope that the facts gained from a very hmited examination on the spot many years ago may be of some use in the absence of more complete evidence. During a visit to Brittany in 1887, leaving the rail at Quimper, I made a point of visiting the wildest district 118 The Breton Of en Field System. within reach, viz. the peninsula of Penmarch on the south-west coast of the Departement of Finist^re. The commune of Penmarch hes in the corner of a blunt promontory swept as bare of trees as the Cornish Land's End by Atlantic gales. Its wide stretching monotonous plains are still cultivated on the open field system. They are dotted over with scattered hers or hamlets, each surrounded by its open fields. The hamlet of Saint Gwenole is conspicuous owing to its massive tower or belfry — ^the only portion remaining of the ancient church, forming a central landmark visible from far away. Taking an inteUigent Breton as guide and walking over the open fields, the most obvious fact at once noticed was that the arable plain was generally ploughed up into long narrow ' high-backed lands.' One noticed also in some cases the double curve in the strips, the reverse of the letter S, so common a result of open field ploughing. These high-backed lands are very narrow and piled up in the ploughing ahnost hke the mound of a grave. They were hardly broader than a long step. They are known in Breton by the same word erw as in Cornwall and Wales, and also as ' sillons.' And several of them in combination form the journal or arfent. The peasant holdings consist of so many erws or combinations of them scattered over the open fields. But the equahty of the holding, if ever prevalent, has long ago been lost. As the Roman actus, meaning the single drive of the plough or length of the furrow, became the limit of the acre, and gave its name to the square of which the jugerum was the double, so the ' arpent ' — the ancient The Breton Of en Field System. 119 Gallic ' aripennis ' — ^has been identified by Professor Ridge way and others with the Brehon ' aircean,' a word used in Irish for the end of the side of the larger agricultural unit, the ' tir-cumail.' So the Breton words ' ero/ and ' irvi/ and ' ervenn/ and ' arpent/ seem to have their first meaning in the drive of the plough to the end of the furrow, marking the length of the strip and thus falhng naturally, both in the thing itself and the name, within the hues of common agricultural tradition. In the open fields of Penmarch as elsewhere the strips he in groups, and each group has its headland on which the plough is turned at the end of the furrow. The word for the headland is talar or dalar, plural dalarou — ^hke the Welsh talar and talarau for the same thing. There is in Breton a touching use of the word, showing how completely the practice of the traditional open field husbandry is embedded in the mind and life of the Breton peasants. The right of turning the plough-teams on the headlands makes them always the last strips to be ploughed. So this last work of the jaded ploughman and worn-out oxen at the end of the season becomes the famihar symbol of the last struggle at the end of a hfe of toil. The Breton peasant nearing his end, with laboured breath, is said to be ' ploughing his headlands.' This phrase may at first sight suggest the solitary ploughing of a single peasant. But in the absence of more direct evidence of Breton co-ploughing we may perhaps connect the Welsh cyvar and the Irish comar, both used in the sense of co-ploughing, with the Breton Jcefer or kever apphed to the Armorican journal or ardent. Still further when the customary journal in 120 The Breton Open Field System. the Vannes dialect is called a hever-doar,^ it is difficult not to connect the word with the Welsh cyvardir, in the sense of a co-ploughed land. Treading one step more on treacherous ground it may not be unreasonable to claim that the Breton application of the term ' day's ploughing ' {devez-arat or deouech-arat) to the kefer or journal may suggest at least that the ' day's ploughing * of it involved co-aration. Another essential feature of the open field system is the common right of pastui'e over the fields when the crop is removed. The whole of the open field is called by the peasants the ' megou ' (pronounced mejou) — ' the fields.' The word is also applied to uncultivated or waste land, and is used to describe the part of the fields for the time being without crop, emphasising the fact that during part of the year the ploughed portion falls back into waste and becomes common pasture again. Thus, however imperfect the information gathered by a visitor, with a Breton dictionary in his hand and an inteUigent guide at his elbow, during a walk over the open fields of Penmarch may have been, the general impression gained can hardly be far wrong that this Breton open field system, both in its facts and its terms, has very close relations to that of Wales and Ireland, and shares with them to a remarkable degree agricul- tural traditions which seem to belong to an ancient common inheritance. So much for Penmarch. As we walked along the shore to the ' Pointe de la Torche ' the peasants were busy collecting the seaweed thrown up by the great Atlantic waves, and burning it for manure, exactly as 1 Douar=tir or dir, land. The Breton Of en Field System. 121 the Irish peasants do, in roughly made furnaces on the spot built up of stones from the beach. As we stood on the rocky ' pointe ' the httle columns of smoke from these furnaces rose at intervals all along the coast, just as we had seen them in the North of Ireland at the mouth of Lough Swilly. Further inquiry resulted in the fact that the primitive open field system extends to Cape Finistere and the islands off the coast. At Plouharnel and Carnac and at the prefecture at Nantes the field maps of all the country round are available. Illustrated by the pubhshed ' usages locaux ' of each commune they afiord abundant evidence of the most authentic kind of the division of the fields everywhere into the strips or erws, of the two or three-field system of rotation of crops, of the scattered nature of the holdings and of the vaine pdture or right of pasture over the strips after removal of the crops. As in the case of Saint Grwenole a prominent feature of the district of Carnac is that instead of each commune being a single village community with its open fields around it, it is in fact a collection of hamlets or kers dotted over the district. Each ker is a closely packed group of a few homesteads and farm buildings, the absence of windows outside giving the ker the look of a fort. Within its precincts each ker has generally its common well and common oven (often of sohd masonry) and its own set of open fields spread out around its walls. Viewed from the modest knoll of Saint Michel the plain stretches far away on all sides and is dotted over with these Httle kers. Carnac marked by the lofty spire of the church of St. Cornely has grown into the centre 122 The Breton Of en Field System. of a commune composed of several of these leers for the festival day of the saint, who is supposed to be a special patron of cattle ; the devout peasantry flock in crowds from the various Iters bringing their heifers to be blessed at the sacred well and purchasing halters which have received magic virtues from the blessing of the saint. One is forced to reahse how all these customs and traditions hang together and how the Roman Church, absorbing them into its system, has become the centre of the daily hfe of the peasant, as in early ages the Druidical priests may have been. Surrounded by the remains of a primitive civilisation, one begins to understand how the Romans found it to be their best pohcy to absorb silently into their provincial system of rural administration much that was of Gallic origin, as they did in the case of the Galhc leuga. Here in this region of ancient Celtic and pre- Celtic monuments Roman remains show clearly how the Roman soldier had no superstitious fear or reverence for the weird procession of stones, one branch of which in its strange monotony ten or twelve stones abreast, stretches across the plain of Carnac for hundreds of yards and then abruptly halts at Kermario, no one knows why, just when its stones are hugest and its front most imposing. Roman soldiers did not scruple to build against these stones or even to draw, some of them into the line of their ramparts. But the Breton peasant regards them still with awe, as an invading army turned into stone by his patron Saint Cornely, and old maps show that the open field husbandry brought itself up to a sudden stop with a final headland in front of the procession of stones, daring to proceed no further. Extracted by permission of M. Is Maire of Carnac from the " Tableau tlu Plan Cadastral parcellaire de la Commune de Carnao, Canton de Quiberon — 1833." The Breton Open Field System. 123 I am able to give the tracing of a map of one of the hers in the commune of Carnac, by way of evidence that the open field system in the pecuhar form described is at home in the extremity of Brittany. In visiting the district round Eedon the prevalence of the open field system may be at first sight not so apparent. The chestnut tree to the eye seems to become the ruhng object of husbandry. Seemingly, square fields are often enclosed with hedges composed of chestnut trees. But an inspection of the ofiicial map of the commune at once discloses the fact that the area within these square enclostires is in reahty divided into strips with the same scattered ownership belonging to the open field husbandry. CHAPTER IV. THE ACRES OF THE CORN-OBOWINQ DISTRICTS OF FRANCE. I. THE BRETON ARPENT. The correspondence in area of the Cornish acre and the Breton arpent taken together with the change of form from 1 X 10 to 1 X 5 is a typical instance of what will become familiar as we proceed. The Cornish acre, as we have seen, is in the form of 1 X 10, i.e. 4 X 40 of a rod of 5-5 m. or 8 x 80 of the gad or goad of 2-75 m. The Cornish peasant, as we have noticed, still measures his potato plot as 8 x 8 of this gad, and calls the square (which is ^q of the acre) a ' lace,' ' laz ' being Breton for a rod. But the Cornish peasant measures the customary acre 4 x 40 of the rod of two gads, i.e. of 5'5 m. The Breton peasant measures his arpent (1 x 5 in shape) with 4 x 20 of a corde of 7 "8 m. What is this Breton corde of 7'8 m. ? It is in length exactly the diagonal of the square of the Cornish rod of 5'5 m. (5-5 X 1-415 = 7-8). That this customary arpent is not of modern intro- duction is shown by article 263 of the ' Consuet. Brit.' The Breton Arpent. 125 quoted by Du Cange,^ in which is the statement 'Arfennis, seu Jornale apud Armoricos constat 20 chordis in longitudinem et 4 in latitudinem extensis, chorda vero 24 pedibus regiis.' The Royal standard foot of France is -325 m. and 24 of these feet = 7-8 m. There has been no change in the corde since the date of the Custumal. The furrow of 20 cordes was then, as now, 156 m. in length, and the arpent then, as now, in the form of 1 x 5, measured 31-2 x 156 m. Now, the Breton corde of 7*8 being the diagonal of the square of the Cornish rod 5*5 m., the small end of the Breton arpent is the diagonal of the square of the small end of the Cornish acre, while the long side or furrow of the Breton arpent is half the diagonal of the square of the Cornish furrow. So completely is the Breton arpent both in area and construction a transformation of the Cornish acre 1 x 10 into the two half-acres, each 1 x 10, which combine to make the Breton arpent of 1 x 5 m., or it may be the reverse. The area of the Cornish acre is 484 ares, and that of the Breton arpent 48"624 m. The small discre- pancy is sufficiently explained by the measurements being given in even feet of English and French standard. The connection of the furrows of these acres with itinerary measures is also remarkable. The Cornish furrow being ^ of the milltyr the Breton furrow ought mathematically to be ^5 of half the diagonal of the square of the miUtyr, i.e. of 4680 m. How does the Breton peasant bring the length of his furrow into harmony with itinerary measures ? At the time of my visit in 1 Sub Arpennis. 126 Acres of Northern France. 1887, 1 made inquiries on the spot, and having purchased the official ' Usages locaux ' of the Breton department of ' Ille et Yilaine ' I found to my surprise that the local ' lieue de Bretagne ' was 4677 metres, i.e. half the diagonal of the square milltyr of the Welsh Codes, as, mathematically, it ought to be. Nothing could show more clearly the close connection between Breton and Cornish itinerary and agricultural measures than this transformation of the Breton itinerary measure not into half the diagonal of the square leuga but into half the diagonal of the square milltyr, which in the Venedotian Code of North Wales was attributed to the mythical Dunwal. It may perhaps be said that the milltyr may have been imported with the acre from Cornwall to Brittany by the emigrants who fled from the Saxon Conquest. On this hypothesis it would be very striking evidence of the early prevalence of the ' milltyr ' in Cornwall, as well as in Wales. We should have to begin to beheve in the personality of Dunwal as well as in the prevalence of the measures ascribed to him. The tradition of the Venedotian Code making him the successor of the King of Cornwall by marriage with his daughter would thus find some foundation! But there is the other possibility, viz. that whUst in most cases the agricultural unit was derived from the itinerary measure, in some cases the agricultural unit may have been the one rooted most deeply in local custom, and a local itinerary measure may have been derived from it. But this ' heue de Bretagne ' can hardly have been derived from Britain, for it occurs also as the local ' lieue ' of the Province of Berry and again in the The Breton Arpent. 127 Lyonnais, as though it had its home in the South of France.^ The Breton arpent forms so typical an example of the change in the local itinerary measure to make it correspond with the change of form without change of area in the arpent, that it may be well to follow the lead it gives us a little further. The fact has abeady been mentioned that the Egyptian Met or aroura, continued in use under the Ptolemies, was in one square, but that when turned into two squares it became the Greek medimnus of Cyrenian usage. The sides of the medimnus were directly connected with the stade and the leuga of our Western districts. But with what itinerary measure was the khet in a square connected ? This question brings us back to the Breton arpent, for if, changing the point of view from the close connection with the Cornish acre, we take the other direction and ask what is the connection of the Breton arpent with the wider Continental and ancient world, we find that 1 Atlas Oiographique et Militaire de la France. R. J. Julien a I'Hotel de Soubise, 1751. Echelles de I'Etendue du Degre de Latitude evalu6e a 57,060 Toises du ohatelet de Paris. Laeues communes de France et de Normandie 25 (to degree) . ....... i.e. 4444 m. Petites Lieues de France, ou Lieues de Paris et de Sologne 28|: Grandes Lieues de France ou Lieues Marines Lieues de Beauce et de Gastinois de Gascogne et de Provence . de Bretagne et d'Anjou de Berry ..... de Bourbonnois .... de Lyonnois .... de Picardie et d'Artois . Milles d'Angleterre ...... 70 K 28i 3932 „ 20 5554 „ 33| 3291 „ 19 5846 „ 25* 4539 „ 25| 4702 „ 23 4829 „ 23i 4776 „ 25J 4400 „ 70 1572 „ 128 Acres of Northern France. its furrow of 156 m. is only one metre less than three times the length of the side of the Egyptian khet in its original square form. The square of the Breton furrow would therefore contain 9 Egyptian khets. Was then the side of the khet also a division, not of the leuga, but of the diagonal of its square ? And if so in what geographical region were the diagonal itinerary measures in use with which it was connected ? Is there any evidence of their prevalence anywhere in the Eastern world ? This is not an altogether wild or useless inquiry. We know from Herodotus and others that next in importance to Egypt as a corn-growing country was the plain at the mouths of the Danube and other rivers which flow into the north-western corner of the Euxine. The reader will hardly need to be reminded that Athens drew most of her corn supply from this region, and it was studded with Greek cities, whose commerce was of great historical importance. Moreover, this district, hardly less than Egypt under the Ptolemies, was under Greek, and, specially, Macedonian influence. The corn from this rich corn-growing centre was measured in Greek medimni. It is natural therefore that we should seek for the itinerary measures connected with the khet and medimnus in the district from whence the corn supply of Athens and other Greek cities was derived. It is true that direct contemporary evidence as to the local itinerary measure of the Danubian district fails us, but we have seen enough to warrant a reasonable trust in the antiquity and permanence of customary measures. Turning, then, to Martini ^ for information 1 Angelo Martini, Manuale di Metrologia, Torino, 1883. The Breton Arpent. 129 as to the customary itinerary measures of this region what do we find ? The ancient measures of ' Bucharest ' and of ' Jassy ' are thus described : — (1) The Posta = 2 mil = 15,697 metres. (2) The Mil = 7848 metres. (3) The Lega di Rumania = 3924 metres. (4) The Lega di Moldavia (Jassy) is described as practically of similar length — 3962 m. Thus the Posta of Rumania is a round 100 furrows of the Breton arpent, i.e. 5 diagonals of the square leuga. The mil of Rumania is 50 furrows of the Breton arpent, i.e. 5 half-diagonals of the square leuga, whilst the Rumanian and Moldavian Lega is 25 furrows of the Breton arpent, or 2| half-diagonals of the leuga. i Not that these Danubian itinerary measures so obviously Breton at first sight had anything directly to do with the Breton arpent. We may come to the conclusion that a more probable explanation may be found in the wide prevalence of a common tradition extending from the Euxine into Western Europe. Be this as it may for the moment, the fact remains — whatever the explanation of it may be — that while the long side of the two squares of the medimnus (viz. 74 m.) is ^q of the leuga, the side of the khet or medimnus in one square (viz. 52'4 m.) is g^, of the half-diagonal of the leuga. The customary itinerary measures of the great corn region of the northern plains of the Black Sea are found after an interval of 2000 years, mathematically speaking, to be multiples of the side of the Egyptian khet, and therefore of the Breton arpent, and not of the Greek medimnus in two squares. 1 Of. the ' Lieue de Paris et de Sologne,' 3932 m. k2 130 Acres of Northern France. So tliat the seeming coincidence hid in the fact that the Rumanian Posta is exactly 100 furrows of the Breton arpent may be simply the historical result of common relations of the Breton arpent and the khet to the diagonal of the square kuga. In any case the fact itself is important in the illustration it gives of the persistence of the relation between agricultural and itinerary measures. But to return to the Breton arpent. If we would bring it into touch with what we have seen of Breton methods of ploughing we must examine its structure more minutely. As already remarked, walking over the erws they are found to be ploughed into very narrow high-backed strips sometimes scarcely more than a long stride in width. How does this comport with the Breton arpent ? The narrow strips evidently were not arpents. They were called erws, or sillons, by the peasants, but evidently several of them went to the making of the arpent. In the OflS.cial Tables^ for the extreme Western departments of the Cotes du Nord and Finistere the arpent of two half-acres with a furrow of 156 m. is described as consisting of 120 raies or single fucrows about i of a metre in width— 6 of these raies being grouped in a sillon 1-56 m. in width. The sillon there- fore probably corresponded to the narrow high-backed land, 20 such sillons in this district making the arpent. The sillon, or high-backed land, was the same width in the Department of lUe-et-Vilaine. But in Morbihan, round about Carnac, the arpent is said to be divided into 60 sillons, which would make the width of the sillon only -52, little more than half a metre. 1 See reference below on p. 137 note. To fact p. 131. ''■■■: -_.-^■•" V-■^..A.. ..,.•••••■; 1..../ v- \.- '■'? Kf /'■■■ 'i... 78 78 \ ! 78 ■y K 78 s :.■•;. J--V-. .-■■ i "■■-. ,:>-i V ' 78 \ 78 (■] FiQ, 20. The Breton Arpent. 131 In the ' Recueil des Usages Ruraux ' in the Canton de Seiches, near Angers, the ancient practice of cultiva- tion in high-backed sillons of ' one or one and a half ' metres wide is spoken of with strong disapprobation as still general in the canton. How far further eastward it extended I am not able to state. But in the introduction to the Eedon Cartulary it is stated on the authority of local archives that at Rennes in the fifteenth century there was a journal of 16 sillons each of 6 rayes. These variations in the structure of the prevalent arpent and the further fact reported that there were in one or two districts variations of the arpent itself, hut without alteration of the length of the furrow, are also interesting. Thus there is mentioned an ancient journal of Vannes and Rennes and ' petit journal ' of Morbihan with an area of 36468 ares and its double with an area of 72-936 ares, still having the same furrow as the arpent — the fact being that the latter is simply three of the half -acres instead of two. This method of retaining the furrow of 156 m., and of increasing or diminishing the area of the arpent in certain cases by altering the number of sillons, brings into notice the further fact that in some districts there was also a customary arpent in a square. It was a division of the square of the furrow of 156 m. into 4 apparently square arpents with 78 m. to the side and with an area of 60-78 ares. This square arpent was probably in use for other crops than corn, for it occurs sporadically much further afield than the 1 x 10 or 1 x 5 arpents distinctive of Brittany and the corn-growing district of Northern France. The accompanying map (fig. 20) will show the extent of the prevalence both of the normal arpent of Brittany fading away towards the East, and also of this apparently square arpent 132 Acres of Northern France. passing southwards into the region of the vine, where the 1 X 10 acre disappears. These facts, standing alone, might or might not afford sufficient ground for historical inference as to origin. But they bear within themselves at any rate some significance. We may probably state the inference thus : — That the square of the Breton furrow of 156 m. contains without alteration in shape 9 of the Egyptian khet ; and that both the Breton furrow and the side of the khet find their direct connection with itinerary measures, not with the leuga, but with its diagonal, whilst the furrow of the Cornish acre (similar in area to the Breton arpent) is a direct division, viz. one-tenth, of the leuga itself ; these are facts which point to connection with the khet and the medimnus rather than with the Greek aroura and the Eoman jugerum. II. THE NORMANDY CUSTOMARY ACRE. Passing on to Normandy the most prevalent and typical acre — the ' Normandy acre ' — is in the form of 1 X 10. It consists of 160 perches of 7'15 m. and so has a furrow of 286 m. and an area of 81 '715 ares. It is therefore in area the double of the English statute acre put into the form of 1 x 10. An English half-acie 1 X 10 (of 142-143 m. in furrow) thus becomes the rood or verge of the Normandy acre. And further (as if following the Breton example), the Normandy verge or rod of 7*15 m. is almost exactly the diagonal of the square of the English statute rod of 5-025. To face p. 133 --^-'■"'i 286 f , 43 143 71 Fio. 21. The Normandy Acre. 133 With regard to the relation to itinerary measures it must be specially noticed that it is the Normandy furrow which is in direct connection with the itinerary measure, being \ of the leuga of higher standard, while the furrow of the EngKsh statute acre, 201 m., is \ of half the diagonal of the square leuga of higher standard — eight furrow lengths, i.e. 1609 m., making the English statute mile. Thus the relation to the itinerary measure is maintained on both sides the Channel in the same way as in the case of the Cornish and Breton acres, but by a reverse process. In passing from the Breton-Cornish pair of acres to the Normandy-British pair of acres, we seem to have passed from a probable connection with the hhet and the medimnus to quite a fresh one, viz. with the Greek aroura and the Roman jugerum. The Normandy verge or rod of 7'15 m. is almost exactly ^ of the Roman actus and ^^ of the long side of the two squares of the jugerum, viz. 71*1 m. It is of slightly higher standard. The result follows that the square of the British furrow of 201 m. (J of the half-diagonal of the square leuga) would contain within it 16 Grreek aroura or 36 Greek plethra in their proper shape as squares, whilst the square of the furrow of the Enghsh half-acre in the form of 1 x 10, viz. 142-143 m., would contain without change of form 4 Eoman sortes or 8 Eoman jugera, but of shghtly higher standard. But how far these facts taken alone would give ground for an inference of Greek or Roman origin, and, if of either, of which, must be left to a later stage of this inquiry. We may have to go back to an influence wider than both. In the meantime the map (fig. 21) will show the 134 Acres of Northern France, range of the prevalence of the Normandy acre. On the whole it does not seem to travel very far from its special home. But its verge, the Enghsh half-acre (1 X 10), wanders much further. The Chartrain setier, in the form of 1 x 5, with a slightly diminished standard, 4 x 20 of a rod of 7"04 m., and an area of 39 "628 ares, can hardly be regarded as other than two English half-acres or verges of the Normandy acre, while the Lorraine arpent or ' jour ' — 1 X 10 with an area of 20"44 ares — ^regains the full standard of the single half-acre. Finally, the prevalence of the same half-acre as the ' morgen ' of Frankfort-on-the-Main, 4 x 40 of a rod of 3'55 m. (h) of the side of the Eoman actus), having an area of 20"25 ares, seems meant to remind one that it was at home in a thoroughly Romanised region, and that, even though sporadically traceable back into once German territory, it may not be of German origin. III. THE 329 M. ACRE IN TWO HALE-ACRES. This acre is represented on the Continental side of the Channel ahnost solely by its half-acre in the 1 x 10 form, with^ a^ furrow of 233 or 234 m. and occasionally with a higher standard of 238 m., being 4 x 40 of a rod of from 5*82 to 5'95 m., and having an area from 54-702 to 56-746 ares. In the Departement of the Seine et Oise it occurs with a furrow of 234 m. measured with 3 x 30 of the Breton rod of 7-8 m. I am not aware that it occurs in any other form than that of the half-acre 1 x 10. Its interest hes mainly in the fact that on its lower To face p. 135 232^: 232 116 233" / V Fio. 22. The Stade Acre. 135 standard its area equals 2 Greek medimni or 5 Greek modii of corn-land. Nor can we overlook the fact that in its higher standard the whole acre — ^the ' faltasce' — occurs again in the corn-growing country at the mouth of the Danube (Jassy), where the two half-acres, 1 x 10, make an acre in the form of 1 x 5. It there is reported by Martini to consist of two half-acres, each 12 x 120 stingeni of r98 m., or 4 x 40 ' predjine ' of 5 "94, the rod of the French half-acre, and with an area of 113 ares. Its appearance in this region may have a direct historical meaning, but there can be little doubt that it owes its existence to the fact that it contains an area of 4 Greek medimni and that at the same time in its form of 1 X 5 it contains 10 modii of corn-land in their normal form of two squares. The square of its furrow, moreover, would contain 16 Egyptian khets in their normal square, just as the square of the Breton furrow would contain 9 such khets. Its relation to the Breton arpent rather than to the Normandy acre seems thus to be distinctly indicated, both by its connection with the khet and medimnus, and its occurrence in the faltasce of the Danubian district. In the meantime the map (fig. 22) will show the range of its somewhat sporadic appearance among the other associated acres in the corn-growing districts of France. IV. THE DORSET OR STADE ACRE. This important acre, as might be expected from its direct connection with itinerary measures, was widely spread over France, sometimes in its natural form as a 1 X 10 acre with a furrow of one stade or 185 m. in 136 Acres of Northern France. length, at other times in the form of 1 x 5, i.e. two half-acres 1 x 10— with a furrow of 131 m. — and more rarely in the form of a square. It sometimes doubles itself like the Normandy and Breton acres by putting 4 of its half-acres together and so regaining the form of 1 X 10 with a furrow of 262 m. The area of the true stade acre is 34-333 ares, and that of its double is 68-666 ares. From the map (fig. 23) it will be seen that in its form of 1 X 5 it keeps mainly to the North, though reappear- ing in the South of France, whilst in its natural form, 1 X 10, it becomes prevalent in what we may call the Burgundian district on both sides of the Jura, in- cluding Switzerland, as we shall presently see. It also occurs, as it naturally would do, outside the corn district, in a square form over a considerable region. The stade was so prominent an itinerary measure, being ^-g of the leuga and | of the Roman mile, that its far wider prevalence than that of the other Armorican group of acres cannot be surprising. V. THE GROUP OF ACRES. One cannot help noticing that whilst these customary associated acres spread over different parts of the corn- growing districts of France they seem to accumulate in the Departements of Calvados and the Seine- Inferieure, as if gathering for the purpose of crossing the Channel together into Britain ; or rather as if this were the point of passage from either shore. In those Departements not only are all the four above-mentioned acres in evidence, but also the Jersey To face p. 136 130 131 rsf -- 131 7 1 31 131 D 185 "v . -■'" 130 188 188 185 i ( (88 185 a 131 > "31 ) Fia. 23. The Group of Acres. 137 acre, with its furrow of 268 m., appears in its half-acre form 1 X 10 with a furrow of 190*4 m. and an area of 36-08 ares. In the Departement of the Seine-Inferieure the group of associated acres has the further addition of the West Derby 275 m. acre without change of shape and varying in furrow from 275 to 279 m.^ We miss only on the French side of the Channel from the British and Armorican group of associated acres the Northumbrian or Irish acre with the 256 m. furrow, perhaps the most interesting because most mysterious of the British and Irish group. ^ But it occurs, and without change of form, in the Channel Island of Guernsey, with the additional information that 5 Guernsey acres make a manorial bouvee, and 12 bouvees a corvee of 60 acres with an area of 3933 ares. The position of the so-called Irish acre of 65*62 ares is so prominent among British customary acres that it is worth while to examine carefully the Guernsey example of it and the larger agricultural unit derived from it. It is very easy to see that the area of the corvee of 60 acres, viz. 3933 ares, is equal to 144 Egjrptian khets or Greek medimni. 1 Tables des Rapports des Anciennes Mesures Agrains avec les Nouvelles, par P. Gattey, 3""^ ^fidition, Paris, 1812. ' Ces tables sont extraites des tableaux de la comparaison des anciennes mesures de la France avec les nouvelles, dressdes par les Commissaires nomm6s h, cet effet dans chaque departement,' " See below, p. 247. CHAPTER V. THE ASSOCIATED GROUP OF ACRES LIMITED TO ITS OWN DISTRICT AND ITS INDIVIDUAL MEMBERS TRACED EAST- WARD ALONG THE LINE OF THE LEUGA INTO THE CORN- GROWING REGIONS AT THE MOUTHS OF THE PO AND THE DANUBE. I. THE RHENISH 1 X 10 ACRES AND THEIR GERMAN NEIGHBOURS. The question now arises how far tlie group of these associated acres common to both sides of the Channel can be traced further East. The extent of the prevalence of the group is most important, but it is worth while to follow the individual acres as far as we can. A reference to the maps will show that East of the boundary of Normandy, that is to say, of the Departe- ments of the Seine-Inferieure and Sure, the 1 x 10 acres seem suddenly to disappear, and to give way to those in a square. In Picardy and Artois — the Departements of Pas de Calais, Somme, and Oise — ^they are conspicuously absent. Happily we have still the guidance of the French Official Tables, as at the time when they were made the geographical area of the French departments extended to the Rhine. The Moselle Valley and the districts comprised in the Departements of the Rhine and Moselle, La Sarre, La Meurthe, des Vosges, and Mont-Tonnerre, cover the ground from Treves to the Rhine and the Main — a thoroughly Romanised district, and the meeting- The Rhenish Acres. 139 place of Roman and German influence, and the scene of the great battle-ground between them. It also is peculiar in the fact that whilst in the Roman itineraries the distances between the stations on the main road from the south as far as Mainz are nominally stated (although really in leugse) in Roman miles of 1000 paces, those from Cologne to Coblenz, and across country as far as Treves and Rheims are stated openly in leugce. And there is a passage in the Notitia dignitatum^ quoted by Mommsen and Meitzen, in which it is stated that at that date the Romans had extended their pos- sessions for 80 leugse beyond the Rhine. ' Nam Ixxx leugas trans Renum Romani possederunt.' So that, as Mommsen observes, the Roman occupation or influence must have reached eastward as far as Fulda. Thus, reckoning directly in leugse in this district was prevalent in the fourth century. Throughout the district of which Treves was the centre the stade acre, varying in furrow from 184 to 188 m., and 1 x 10 in form, was the most prevalent. The verge or quarter of the Normandy acre, as already mentioned, occurs however as the Jour of Lorraine. It is found again in the Departements of La Meurthe and the Vosges, and crossing the Rhine, according to Martini and Doursther, it occurs again as the Feldmorgen of Frankf ort-on-the-Main with a furrow of 142*4 m., and area of 20"25 ares, in company with a Waldmorgen (also 4 x 40) with a furrow of 180 m. and an area of 32 "55 ares. The Waldmorgen is in area very near the half-acre of the Northumbrian and Irish and Guernsey acre with the 1 Seeck, p. 253 ; Meitzen, i. p. 390. 140 The Grouf of Acres. furrow of 256 m., but it may possibly be a stade acre witli a slightly reduced area and length of furrow, as the English form of it also may be. The Morgen of Homhurg, according to Martini, was 4 X 40 Ruthen of 3-45 m., and therefore had a furro.w of 138 m. and area of 19-06 ares. It seems, therefore, to be a rood of the British 275 m. forest acre. ^ 40 of 3 -33 m . iB-4373res 3 6 -874 ares Tageiverke of 2 Morgen Fio. 25. being its \ acre in the form of 1 x 10 — if it is so, it is an exaggerated form of it. The GoTHA WMacker is practically the stade acre of 185 m. of slightly lower standard, i.e. with a furrow reduced to 184 m. (See fig. 26). The Fddacher of G-otha equals 140 Feldruthe of The Rhenish Acres. 141 4-026 m. with an area of 22-7 ares and does not seem to belong to the 1 x 10 acres. In Lippe-Detmold, before 1857, the customary acre was the ' SchefEelsaat ' of 4 x 40 Ruthen of 4-63 m. with an area of 17-165 ares. Its furrow was a stade of 185 m. It was therefore a half of the stade acre longitudinally divided. In Oldenburg the ' Altesjuck ' is reported to have been 4 x 40 of a Ruthe of 5-91 m., making a furrow of 236-4 m., with an area of 56-03 ares. This is again the half-acre of the 329 to 335 m. acre of higher standard. When we have added the Wiirzburg and Nuremberg acres with a furrow of 146 m. (half that of the British 4 X 40 oF 4 60 m The Waldacker of Gotha. Fig. 26. 292 m. acre) and that of Saxe-M.einingen with a furrow of 170 m. we seem to have exhausted the sporadic 4 X 40 or 1 x 10 acres reported as in use in North and Middle Germany. It must be confessed, however, that no longer having the guidance of the French Official Tables the informa- tion as to the 1 x 10 acres of Germany may be incom- plete. Any question how far they were really at home in North Germany must be held in suspense. There are several 1 x 10 acres reported in Switzer- land. The Zurich acre for gardens, 6 x 60 of a rod of 3 m., with a furrow of 180 m. and area of 32-405 ares, seems to be the half-acre of the Guernsey acre, or possibly p, diminished stade acre. It occurs again at Basle with 142 The Group of Acres. a furrow of 183 m. and area of 33*39 ares, very near to the stade acre. There are others which though in the form of 1 x 10, do not seem to occur elsewhere, e.g. Berne 6 x 60 of a rod 293 m. with a furrow of 176 m. and area of 31 ares, the rood or quarter of which occurs at Strasburg. Before passing from the geographical range of the associated 1 X 10 acres of the French Tables and the imperfect German authorities, the reader must be re- minded of the fact that the great corn district of France through which they stretch was surrounded by other districts in which the vine was the predominant object of culture, and in which the agricultural unit naturally takes the square form. We have already noticed that the Breton furrow of 156 m. was connected with a series of apparently square acres with sides of half that length. So also the furrow of the Normandy acre (286 m.) was connected with a series of square arpents with J of the Normandy furrow (71*5 m.) to a side and with an area of 51*072 ares. This was an arpent prevalent enough to be adopted nationally before 1792 as the ' Arpent d'ordonnance des eaux et forks ' of Paris. It was, moreover, practically the same area as that of the Roman sors of two jugera. Again, a square of J of the furrow of the 233 m. half- acre was known up to 1793 as the ' arpent de Paris ' with an area of 34*189 ares, sometimes with slight increase to 35*466. This was the area of the 1 x 10 stade acre with a furrow of 185 to 189 m. put into a square. We therefore end this subject with the reminder that the long furrow of the 1 x 10 acres which seems The Armorican Acres. 143 so readily to be explained as a division of tlie leuga or stade or of the diagonal of their squares may have no monopoly to this distinction. The shorter furrows of the square arpent of the surrounding districts may often be divisions of prevalent itinerary measures. And thus as both may have direct relation to the same itinerary measure it would not do to conclude from the presence of both in the same dis- trict that the one form has been directly derived from the other, though that is quite possible in some cases. The form taken by the prevalent agricultural unit must have been mainly fixed by the character of the land and chmate of the country and the dominant crops for which in the long course of many generations and centuries it has been found to be adapted. But at the same time, whatever the form of the unit in its actual dimensions, the relation to the prevalent itinerary measure may have been just as naturally preserved. II. THE LINK BETWEEN ARMORICAN 1 X 10 ACRES AND THOSE OF THE PO VALLEY AND THE DELTA OF THE DANUBE. Having no longer the help of the French Official Tables, on which we rehed for knowledge of the custom- ary acres in use before the French Revolution in the various Departements of France, and having attempted imperfectly to follow the 1 x 10 acres into Germany by means of such information as the more general metrological treatises of Martini and Doursther afiord, the reader may naturally ask for information as to the antiquity of their prevalence in the corn-growing districts where they seem to have been at home. For this we are thrown more directly upon historical records. L 144 The Group of Acres. Apart from historical evidence it might be argued that even if it were granted that the French customary- acres belonged to an early system of tribal agriculture like that in Wales in the tenth century, it would not alter the fact that they must have been again and again submerged under the convulsions and economic changes involved in mihtary conquest and poHtical action. These convulsions, it might be argued, may have wiped older agrarian customs from the slate and sub- stituted others for them. And in some cases it may have been so. The French Commission of Inquiry which produced the valuable Tables, was appointed for this very object. Yes, but in spite of revolutionary legislation the force of custom in each commune has remained so strong that the network of the open field system with its scattered strips and its vaine pdture and customary acres still remains on the ground. It is not only in Brittany that it is so. As already said, you have only to climb the tower of Chartres Cathedral and to look across the vast open fields of the rich corn country all round, and then to call at the mairie and see the map of the commune, and spend a franc in buying a copy of the ' Usages Locaux ' of the district to be convinced of this, and to receive an impression never to be forgotten. The same may be said of the stretch of the open field enclosed by the bend of the Ehine as viewed from the hiU behind the old Koman town of Andernach. And when we consider that this open field system of husbandry in holdings of scattered strips, in one form or other, has been the common possession for ages not of one race only, but of Celts and Germans, and Slavs and Russians, to say nothing of the Farther East, we need not altogether despair of finding in The Armorican Acres. 145 historical records traces not only of its existence but also of the route of the travel of agricultural details connected with it from the East to these Western corners of Europe. To begin with, one element of permanence has no doubt been the fact that for more than 1000 years, during the long reign of the manorial system both in England and on the Continent, the services of the peasantry under the open field system of agriculture was intimately connected with the ploughing and sowing of a certain number of strips or acres in the open field or on the lord's demesne. A typical and potent instance in point is the wide prevalence in manorial records of what in the Saxon phrase of the ' Rectitudines ' was the ' gafolyrth,' i.e. the special service of both free and semi-servile tenants in the ploughing, sowing, and reaping of so many acres and the carrying of the produce to the manorial barn afart from the ordinary weeJcworh on the lord's demesne of the more servile class of tenants. In the eighth century Polyptiques of the Abbeys of St. Germain des Pres and of Prum, and more definitely still in that of Rheims, there is evidence enough of the necessity to define not only how many acres were to be ploughed and sown by the tenants, but also in some cases what the size and shape of the acres were to be. Especially needful was it in the case of the Abbey of Rheims, because in the pecuhar district over which the estates of the Abbey were scattered, we know from the French Tables that there was an unusual mixture of customary acres of all shapes and areas. As in the Enghsh case of Northamptonshire (above p. 113) it be- came needful to describe the acres of each district hable 1,2 146 The Group of Acres. to certain customary payments, so in the case of the Abbey of Kheims it was especially necessary to describe for each property the acre by the ploughing of which the services were to be rendered. One point comes out clearly. The Abbey could not impose upon its tenants a uniform acre of its own. Custom was too strong to admit of such a course. The Abbey had in each case to adapt its arrangements to the customary measures of the district in which the property happened to He. The first entry in the Polyptique of St. Eemi, describes the services of the tenants of the estate of Adenaius on the Marne. Hrotmannus ingenuus tenet mansum ingenuilem. (1) Arat ad hibernaticam sationem mappam 1. Continentem in longitudinem perticas xl in latitudine perticas iiii. Ad estivatioum similiter. . . . The mappa was in this case an acre of the typical shape of 1 X 10, measured by 4 x 40 perches. Of the other tenants of this manor it was enough to say ' tenent mansum similiter.' On six estates of the Abbey the ' mappa ' was of 4 X 40 perches. On five it was 6 x 60. Thus on twelve estates out of twenty-eight the acres were of the 1 X 10 form. But on six other estates the mappa was 4 X 100 and on the rest were other varieties. The Polyptique describes the estates and the services as they existed about the middle of the ninth century. But St. Kemi died a.d. 530, and there can hardly be a doubt that the customary acres described went back to the time when the several estates one after another were at different dates given to the Abbey, carr3T.ng their local customs with them. Following further the traces of the ' Gafol-5Tthe ' in other records, the area to be ploughed was known The Armorican Acres. 147 sometimes under the name of tlie ' anzinga ' or ' ande- cena.' We find it under the latter name in the services of the free tenants of the Church as defined in the well- known passage in the Bavarian Laws of the seventh century. It will be noted that the itinerary measure is still the leuga. (S. 13.) This is the tribute for arable according to the estimation of the judge. Legal andecenas (the pertica being of 10 feet) 4 pertica in breadth and 40 in length [he is] to plough, to sow, to fence, to gather, to carry, and to store. For spring crops each acedia is to prepare for two modia of seed and sow, gather, and store it. . . . Let them give posthorses, or go themselves wherever they are bid. Angaria (carry- ing services) with waggons as far as 50 leugce. They cannot be compelled further. It is well worth while to examine the andecena of this passage carefully. It was to be 4 x 40 of a pertica of 10 feet. But of what feet ? If we regard the district as part of the Roman Agri Decumates we should guess that the pertica of 10 feet was the Roman decempeda, and the foot the Roman foot of -295 or -296 m. And when we turn to the customary measures of Bavaria the guess is confirmed by the statement of Martini that the customary Ruthe of Munich, hke the rod of the andecena, was still of 10 feet, the Munich foot being '292 m. ; whilst the foot of Augsburg was •296, and that of Salzburg -298 m. We can hardly be wrong therefore in regarding the foot of the andecena as having been the Roman foot and the pertica as having been the Roman decempeda. The length of the furrow of the andecena would thus be 118 m., or if the foot were to be taken at •292 m., the furrow would be 117 m. 148 The Group of Acres. The andecena is therefore no stranger. It is the rood or quarter of the half-acre of the 330 m. acre in the form of 1 x 10 with a furrow of 233 m. which we have seen to be one of the associated customary- acres in use on both sides of the English Channel, and also found again in the great corn-growing district on Andecena [■iof233haJFgcre] 13 924 or 13 -43 ares Fig. 27. the northern shore of the Black Sea. Its area is the half-medinmus in the form of 1 x 10. The andecena is not the only historical acre of this Bavarian district. In a twelfth century MS.^ or fragment of a chapter of Bavarian Law, quoted by Pertz is a description of 1 Pertz, 495. The Armorican Acres. 149 the jugemm of 6 x 30 pertica of 15 feet, i.e. two 1x10 half -acres each of 3 x 30, making a 1 x 5 acre, as in so many Armorican instances. Taking again the foot as •292--295 m. the rod would be 4-38-4-44, making the jugerum in form and area as below — 2 half acres 1 x 10. 34- -33 ares OPSS 48 ares Pig. 28. This is in fact the Dorset or stade acre in the shape 1x5 assumed by it so often on the Gallic side of the Channel. These two Bavarian acres of seventh and twelfth century records are thus known to us independently of modern customary usage. But they have nevertheless left traces behind them. According to Martini the customary Juchart of the kingdom of Bavaria as fixed in 1809 (including Munich and Augsburg) was 34-0727 ares and in form a square of 400 rods of 2-92 m. Doursther mentions an ancient Jauchart of Augsburg of 16,000 square ancient feet, and with an area of 14*0366 ares. The ancient foot (Werkschuh) of Augs- burg is stated to be the Eoman foot "296, so that the furrow would be 118-4 m. (See Fig. 29.) If we could trust Doursther's ancient Jauchart to be in actual customary use and not merely as a tradi- tion of the andecena, we might say that both the historical acres had survived to modern times, the 150 The Group of Acres. andecena remaining after 1200 years tlie same in area and shape as of old, tte other stade acre retaining its old area but (since 1899) turned into a square. Ancient Jauchart of Augsburg i4-036eares Fig. 29. Having got so far as this, we may fairly, with some confidence follow the evidence of the customary acres further in trying to answer the question whether we may trace the trail of these Bavarian acres further down the Danube Valley or over the Brenner Pass into Italy. Perhaps both ways may be open. If we follow the faithful guidance of the leuga we shall turn up the valleys of the Inn and the Enns on our way over the Brenner Pass into Italy. From the Urkundbuch of the Enns Valley (1, 74) we learn that the leuqa was the itinerary measure of the Alpine district in the eighth century. In the customary measures of Innsbruck we have the familiar foot of *334 m. and Ruthe (of 10 such feet) of 3-34. The areal measures are stated by Martini to be : — Square rods Ares. Grabe 80 8-93 Starland . . . . 100 11-16 Jauch . . 360 40-183 Tagmat . . . . 400 44-648 Stochiaoah . . 800 89-296 The Stochiacah of SOORuthen would naturally take the form of two squares, and the furrow of its long side would then be of the same length as that of the Bavarian twelfth century historical stade acre of full Eoman standard. And the Grabe would be \ of it with the same furrow. The Armorican Acres. 151 The two Tagmat remind us that hay is the ruHng crop. Then, as now, all through the Alps the amount of hay possessed by the peasant rules the number of cattle which can be kept through the winter. esB star/and 100 sq rods ^9^ 6° Grabe 80 rods i e J>' (P" 2x40 of 3-34 Stochiacah SOOsq.rods Tig. 30. C/. Aranzada of Madrid, 400 square Estadals of 3-34 m. i.e. 44-72 ares. The Starland is the area for ploughing and the sowing of the ' corn-star ' of seed, and it is of the same area as the Greek corn-land modius of higher standard. The Jauch is identical with the EngUsh statute acre, both in area and in its 1 x 10 shape, and by its presence here in the Tyrolese Alps we reaUse that its area is equal to three of the andecena reckoned on the foot of •292 m. Jauch €"60^ 20 04- 13-39 square rods of \3-34- Fig. 31. Divided longitudinally into 3, the Jauch would make an andecena in the form of 1 x 30, i.e. 66'8 x 200'4 m. 152 The Group of Acres. in both area and shape almost the exact counterpart of Dunwal's ' tir/ 1000 of the short end of which made his milltyr. The Meile of Tyrol is recorded as 10,691 m., i.e. 32,000 feet of -334 m., i.e. 80 of the furrows of the Grabe. Beyond the culm of the pass, on the Itahan side in the valley of the Adige lies Botzen or Bolzano, where the agrarian measures seem to be the same as those of Innsbruck, the rod being 3-34 and the foot -334 m., and the mile 10,691 m. and the Jauchart 35-99 ares (4 Grabe) making the acre 1 x 5— two half-acres of the stade acre of higher standard. Lower down the valley hes Trent, and here the twelfth century Bavarian jugerum appears again under Italian nomenclature but identical both in area and shape. The pio of Trent is described by Martini as 720 tavole or square rods of 2167 m. and an area of 34-84 ares. Doursther gives the area as 33-827 ares. P/o of Trent two I 'U, i.e. 10 x 100 ghebbi of 1*565 m. 0^' t\0 Migliaio di C/7ebd/ (paces) o^ / 565 z^ 4-S6 ares Fig. 35. Here we have repeated in area and form and length of furrow one of the two half-acres forming the Breton arpent. Here also it is connected with a local itinerary measure. The mile of Modena of 1569 m., i.e. 10 of the Venetian and Breton furrows, was half the diagonal of the square of the leuga of 2220, and -^^ of the Danubian posta. (3) The ' cawbfo di 840 tavoW with the same 280 Tavo/e Kg. 36. furrow of 156-9 m. with an area of 36*566 ares, being f of that of the Breton arpent. This is almost the exact counterpart in area and structure of the ancient ' journal ' of Vannes and Rennes, like it retaining the furrow of the Breton arpent but having only f its breadth. (4) The migliaio di passi containing 10 x 100 paces of 1-738 m. and an area of 30-229 ares of the same 156 The Group of Acres. area as that of Verona but in the form 1x10 instead of 1 X 5. Campo v.- — Hes. tt. n4\e6poi'. 196 Mediterranean Land Units. In Iliad xiii. 706 : ' In fallow land two wine-dark oxen, with equal heart, strain at the shapen plough, and round the roots of their horns springeth up abundant sweat, and naught sunders them but the pohshed yoke as they labour through the furrow till the end of the furrow brings them up.' And on tie shield of Achilles many ploughers are pictured at work ; but they seem to have driven their single yokes, ploughing in strict order of rotation, following one another as in the case of Elisha, who was ' ploughing with twelve yokes of oxen, and he with the twelfth.' In no other way can Homer's wonderful picture of the ploughing be fully realised as true to life. (Iliad xviii. 541-548.) Furthermore he set in the shield a soft fresh-ploughed field, rich tilth and wide, the third time ploughed : and many ploughers therein drave their yokes to and fro as they wheeled about. Whensoever they came to the boundary of the field and turned, there would a man come to each and give into his hands a goblet of sweet wine, while others would be turning back along the furrows fain to reach the boundary of the deep tilth. And the field grew black behind. Instead of the larger team, each single ploughman with his yoke ploughed his own single furrow to be followed in close succession by the next ploughman, else the field would not have grown black behind the ploughers, and each single ploughman received his goblet of sweet wine before he wheeled about his yoke to the other side of the field, to commence the return journey of the plough. He had accomphshed his plethron and turned to complete his diaulos. Now the doubt arises whether the single yoke could plough a long unbroken furrow of a stade of 600 feet in length, when the Roman agrimensores considered 120 feet as long enough for the single yoke. Even if we make allowance for its being a ' third time ploughing,' so that ' the clod should yield before the ploughshare,' the emphasis laid on the 'clean Homeric Ploughing. 197 unbroken furrow ' seems to forbid a rest before the end of tbe stade would be reached. But if, taking advantage of the scholiast's remark, ' the gues has two stades,' we turn the gues or stade into the form of two i stades, each 1 X 10, without altering its area, as we find so often done elsewhere, the day-work of the plough becomes more possible. The longer furrow of the sweating yoke is lessened, but still a long one for the single yoke, while the area ploughed in the day seems still too large. Is all this conjecture ? Here let us leave the Homeric evidence and turn to that of a remarkable fragment which seems to belong to a district and period when Phoenician or other Eastern measures had not become wholly merged in the Egjrptian or the Greek. The -n-epl iiirpoDv yrjf is given at length by Hultsch (' Met. Script.' i. 56), though he confesses that he is unable to understand its measures. It contains, how- ever, information which may influence our view of the Homeric evidence whilst, inter alia, throwing some interesting hght upon the day-work of the plough and the Homeric problem involved. There may be after all the possibility of a long furrow for the single yoke in an already well-ploughed plain of alluvial soil, notwith- standing the Roman tradition that 120 feet was long enough for the single yoke. The measures of this fragment up to a certain point seem to be Greek. Cubit = 24 fingers = •498--504 m.i Palm = 4 „ = •082--084 m. Span = 12 „ {3 palms, i.e. natural foot) = •246-252 m. Podi8mus= 16 „ i.e. 4 palm foot . . = •328--336 m. 1 If based on the natural foot. 198 Mediterranean Land Units. The natural foot of 3 palms has, it seems, already been supplanted by the artificial 4-palm foot. The natural foot is here called a sfan, as we shall find it elsewhere in the Mediterranean basin, notably in the districts at the mouths both of the Danube and the Khone, where the natural foot becomes the fan in one case and the folma in the other. But we seem to pass beyond Greek fines when the jugon of 5 arouras is described as containing 30 satias, i.e. 6 satias to each arouxa. The aroura, moreover, is not the usual Greek one of 100 cubits s(iuare but ' it is of 130 culits.' Why of 130 cubits ? Now 130 Greek cubits of 496 to '504 m. would make an aroura of from 64-48 to 65-52 m. to the side. It /v*.6'^ ^ Fio. 45. can hardly be an accident that the side of this aroura taken at the higher standard of -504 m. to the cubit Homeric Ploughing. 199 would be 65-52 m., i.e. just half the furrow of the ' stade ' acre in the form of 1 x 5 (131 m.) and its diagonal (92-7) just half the length of the stade of 185 m. This ' aroura of 130 cubits ' is clearly in area intended to be \ of the square stade of 185 m. Allowing for variation within the hmits of the natural foot, its area would be from 41-6 to 42-9 ares, and that of the jugon of 5 such arouras would be 208-0 to 214*5 /euga 8 ares Fig. 46. ares. As it contained 30 satias the area of the satia or saton would be 6-93 to 7-15 ares. The next statement of the fragment is that ' The yoke of oxen ploughs in a day 2\ satibas.' This would make the day-work 17*32 to 17-87 ares, which brings us again to the stade, for this is the area of half a stade acre. So that the day-work of the fragment is just half the Homeric day-work of Professor Ridgeway. There is a further statement that ' the plethron has a length of 26 cubits and in breadth the same,' i.e. from 200 Mediterranean Land Units. 12-9 to 13-1 m. square, being thus -^ of the half stade acre or day-work in the form of 1 x 10. The fragment contains one other statement. Whilst the jugon is said to be of 30 satibas the kouria {? Cor) is said to be of ' 13 small jugera.' The Phoenician hor as a measure of capacity con- tained 30 sata, so that the kor and the jugon of seed- land should correspond. The area of the jugon of 5 arouras of 41 "6-42 '9 ares being 208-214 ares, the 13 ' small jugera ' it contained would be from 16 to 16 '5 ares. This is the wet-land modius of two vorsus of higher standard. The long side of the two vorsus on the diagonal of the wet-land modius would be -f5 of the leuga. (Fig. 46.) Within the hmits of natural variation of standard the statements of this interesting fragment thus seem to be consistent with existing measures otherwise known and so far as they go to represent reahty. Besides the Homeric interest in the connection of the day-work of ploughing with the stade, they are useful as showing how Phoenician, Egyptian, and Greek measures were entangled together and gradually blending in a substratum of common custom. The kor and the saton were not G-reek measures but yet they seem according to the fragment to be connected with the stade and the plethron and to seek to estabhsh relations also with the wet-land modius and the vorsus. The diagram (fig. 47) represents the scheme of measures of the fragment keeping within the Limit of difference of standard between 185 m. and 189 m. in the length of the stade and reckoned upon the natural foot. The division of the aroura of six satibas into 25 Homeric Ploughing. 201 plethra of the \ stade acre (1 x 10) probably indicates a conflict and reconciliation of measures — but without absolute exactness. Nevertheless the self-contained consistency of the fragment is remarkable. Lastly, before passing from it, let us note, so to speak, by-points which have been gained. Heretofore we FiQ. 47. have been led to regard the stade as distinctively an itinerary measure. So it became when the length of the furrow in the ploughing was found so often to be a division of the leuga. 202 Mediterranean Land Units. The leuga contained 12 stades of 185 m. of the Eomano-Greek standard. But we have been led by the Homeric evidence, confirmed by that of the fragment, to regard the stade as originally the length of the furrow in the ploughing, the travel of the plough, rather than as the unit of travel on the road. Another point of interest has arisen, viz. the importance of the day- work of ploughing as a unit of measurement of land. And, further, we have learned that the notion of the Roman agrimensores that a yoke of oxen could not at a breath draw the plough more than 120 feet was a local one, and not to be regarded as belonging to widely extended custom. Finally, putting together the Homeric evidence and that of the fragment, it is something to have learned that the stade of the ploughing had already set the standard of the stade of the leuga at or near 185 m. CHAPTER II. THE EGYPTIAN LAND UNITS. After the Macedonian entry into Egypt and all through the succeeding Ptolemaic period, Egyptian and Greek measures were naturally still more entangled. This is clearly shown by the evidence of the various papyri recently published. How far they were already natu- rally connected as simply sharing in the common substratum of widely spread agricultural custom be- longing to an already ancient and advanced civihsation, it may not be easy to determine, but there are facts which seem to point to it. Let us take them for what they are worth. To begin with modern customary measures. Martini describes the ' antiche misure ' of length at Cairo as follows : Oassab . . . 3"85 m. Peddan {20 Cassabeh) . 77' m. Dereghe (4 Peddan) . 308- m. Malacah or Ora di marcia (16 Dereghe) . . 4928- m. He describes the superficial measures thus : Feddan = 24 Chirat or 400 Cassabeh 59'29 ares. Feddan dalle contribuzioni . . 44"69 ares. And Doursther states that the feddan of 400 cassabas is still in use by the Arabs. Recurring to the fragment we should say that this 204 Mediterranean Land Units. feddan of 400 square cassabs, except in the fact of its lower standard, was 4 of the ' smaller jugerum ' or wet-land modius of 2 vorsus. It was in fact a square of 8 vorsus. The feddan has a further Greek connection in the fact of its lower standard following the standard of the khet and medimnus and the stade of 184'9 m. Its diagonal is -^^ of the leuga. If this feddan might be thrown back into ancient times when the Arabs on the eastern side of Arabia were in close contact with Babylonia and on the west with Egypt, then, in spite of its diminished standard, we could hardly fail to recognise in the feddan the quarter of the square of the Breton furrow, which, as we have seen, contained 9 Eg3rptian khets. The quarter of this square had a side of 78 m. and an area of 60'7 — i.e. of 8 vorsus of 7 "6 ares. So far, then, as to the Greek connection with the feddan still in use in Egypt. But was it in ancient use in Egypt ? I do not wish to mix up metrological facts derived from the examination of ancient buildings with the strictly agricultural measures under consideration. But Professors Petrie and Griffith agree that the length of the side of the Great Pyramid of Gizeh was 9069 English inches— i.e. 230-5 m. The area of its base would there- fore be 531 ares. The feddan of Cairo, being a square of 20 x 20 ' cassaheh,' had sides of 77 m. and an area of 59-29 ares. The side of the feddan is thus as nearly as possible ^ of the side of the Great Pyramid, and its area i of that of the Pyramid base. I mention this merely as possibly pointing to the antiquity of the feddan. Egyptian Land Units. 205 When closely examined, moreover, the feddan, though apparently of Arab and possibly of ancient Egyptian origin, has another Greek connection in the cassab of 3'85 m. Though of lower standard, this 7^ m o /j/o A^ Vor / i5-027ares\ SUS ,6V f / ^ / / \ \ Salma oF740-433ares\ Fig. 48. cassab nearly equalled two of the canna of Rhodes, described by Martini as of r968 m., a measure we shall find at home throughout the Mediterranean basin. The cassab of the feddan and the canna of Rhodes were both of the lower standard of the Greek and Roman stade of 184"9. In tracing the spread of these 206 Mediterranean Land Units. measures by Greek colonists, we must be prepared to allow for legitimate variation of standard. But with this caution the reader need not be surprised to find land units equal in area to the feddan and rods equal to the cassab, so to speak, naturally sailing away west- wards with Greek or other colonisation of perhaps still earlier date. Thus it was from Rhodes and Cnidus that the Lipari Islands received Greek colonists in the sixth century B.C. according to Diodorus. And the customary measures of the Lipari Islands as reported by Martini very nearly correspond with the Egyptian feddan. They seem to confirm the connection of the Egypto-Greek measures with the Itahan or Ligurian vorsus, for in the Lipari Islands the tomolo becomes exactly in area two vorsus of the normal standard 7'5 ares, and the length of its sides is 10 Egyptian cassabs. (Fig. 48.) CHAPTER III. TRACES OF GREEK COLONISATION IN MAGNA GRMCIA AND SICILY. In any attempt to distinguish the traces remaining in customary local measures of those introduced by Greek colonists, it is needful to recur to the specially Greek division of the common itinerary measure of the stade fixed at 184'8 into 600 artificial feet of '308 m. instead of 750 natural feet of "251 m. Allusion has already been made to this as the basis of the North- Eastern European measures built so steadily on the medimnus, and upon the higher standard of this par- ticular artificial foot. It is worth while therefore to examine closely the measures of the early Greek colony of Cyrene, because we owe our accurate knowledge of the medimnus, and the striking connection of the medimnus with the stade of the apparently fixed lower standard of 184"9, to the information of the Roman agrimensor Hyginus. The measures he describes are those of the colony when handed over to the Romans by the last of the Ptolemies (b.c. 95) and still in use in his own day (103 A.D.). Founded (about B.C. 631) by Dorian colonists from the island of Thera, the colony had from time 208 Mediterranean Land Units. to time been reinforced by colonists from otlier Greek cities at later dates. Hyginus describes the great land measure of the Ptolemies as having been the flintli — a square of 6000 feet of "308 m. to the side. This foot was the Greek foot of 6OT of the stade of 184-8 m. Fig. 49. Thus the side of the plinth was really 10 stades of 184-8 m. Its area was therefore hterally a centuria of 100 square stades. Hyginus goes on to state that the plinth contained 1250 ' medimna,' and that the ' medimnon ' contained 1356 Roman square feet, in two squares Uke the jugerum. We thus learn exactly the measure of the medimnus. Magna Grcecia and Sicily. 209 Its stort side was } of the stade or ^ of the leuga of lower standard. (Fig. 49.) The square stade contained thus 12| medimni. This description of the Cyrenian phnth of 100 square stades shows that for agrarian areas on a large scale the stade was the usual Ptolemaic measure and probably that of the Greek settlers. This is confirmed by the description of the Lille papyri of the third century B.C. of a great square of 10,000 arouras laid out in 40 plots of 250 arouras each ; for if we take the aroura to be the Egjrptian khet, y ^ y ^^ i.e. the medimnus in a square, the area of the great square of the LiUe papyri would exactly equal that of eight of the Cyrenian phnths. The divisions into the 40 lots were to be made by 3 dykes across in one direction and 9 in the other. The plan given on the papyrus fixes the point that it was to be in a square and the text accordingly de- scribes its circumference as 400 schoinia, and each of the 40 lots as 10 x 25 schoinia, i.e. 250 square schoinia. This makes it clear that the schoinion in this case is equal to the side of the square aroura, i.e. of the Egyptian khet and not of the side of one of the two squares of the medimnus. The schoinos of the land measurers in this case at any rate was the side of the khet or 52*30 to 52*40 m. p2 Fig. 50. 210 Mediterranean Land Units. The object of the document and its plan had reference mainly to the contract for the making of the ditches and therefore measurement directly by the larger measure of the stade would have been out of place. But had the great area been in two squares instead of one it could have been measured by 20 X 40 stades. Each of the 40 plots of 250 arouras was equal in area to 20 square stades. And the long side of the plots may be recognised at once as 10 times the furrow of the stade acre when in the form of 1 x 5, with which we are famihar as the half-diagonal of the square stade. / O/' BA-oee ..>^ •^Ooo ee-/32 7 e£ -ISA FiQ. 51. Passing on from Gyrene to Magna Grsecia and leaving Heracleia for separate consideration we find the custom- ary measures reported by Martini as ' abusively ' in use at Brindisi and Tarentum to be again closely con- nected with the stade of the same standard as that of Gyrene. The tomolo of Brindisi of 2500 passi of 1"848 m. is exactly J of the area of the square stade ; that of Tarentum, an early colony of Sparta (b.c. 708), of 2000 Magna Grcecia and Sicily. 211 passi or two stade acres, i.e. ^ of the area>^f the square stade. It will be noticed that this tomolo of Brindisi is in area exactly two of the aroura of 130 cubits of the fragment put into a square. Tomolo W'7I8 / / e5-743\ / 4-2 -8 71 '^ 7 / I?i-fa7 ares / i. e. half square Stade oF343-ares Fia. 52. Passing on to Sicily. Catania was a Chalcidian colony of about 730 B.C. The salma of Catania of 171 "487 ares is half a square stade in area, i.e. a square of half the diagonal of the square stade. It contains (at a sUghtly higher standard) two of the Brindisi 212 Mediterranean Land Units. tomolos. And we again recognise in it 4 of the aroura of 130 Greek cubits of the fragment. The Cataman tomolo is hardly other than the Greek plethron of 100 x 100 four-pahn feet of lower standard of "328 which equalled in area the seed-land modius. ,.8 Tomolo il-ias 44-740 / /' 178-96 -t square stade of 35 8- Fio. 53. Messina, another Chalcidian colony, had nearly the same measures, but of the higher standard, the side of the square of the sahna being 133*6 and its area 178"960 ares. The tomolo was the plethron of the Greek four- Magna Grcecia and Sicily. 213 palm foot at its higher standard of '334 and with an area of 11"185 ares. There are also in Sicily clear traces of the medinmus. The CaUanissetta tomolo abusively in use was a square of 52 1 m. and had an area of 27 '124, and but for its lower standard it was the khet or medimnus in a square. The salma of 16 tomoh was of 434 ares. Tom olo / Aroura or Khet of /ow standard 27IZ4 I'M. 54. It therefore must be admitted that, besides the evidence of the common substratum of agricultural measures covering an area wider than that of merely Greek influences, there are traces of the influences of particular Greek colonies, left behind them by the colonists and maintaining their ground in the habits of the people, in spite of Punic wars and conquests and reconquests of their territory by Carthage or Rome. CHAPTER IV. THE AQRIGVLTVEAL UNITS OF ITALY. If we regard Italian agricultural units simply from the evidence of the customary areas according to Martini, ' abusively ' continued in use in the several districts notwithstanding legislative prohibition, we shall be struck by the fact that the classical Imperial measure —the Roman jugerum— seems to have left so small an impression on the usage of the Italian peasantry. Martini, himself an Italian, naturally gives full place in his metrological work to Italian customary measures. He gives the particulars of those officially recognised and also of those ' abusively ' in use in no less than 102 districts, and quite imconsciously to himself the results of his figures are as significant as they are unexpected. Happily in this case we have not to rely only upon the customary measures ' abusively ' in use. We can also go back to historical evidence and compare the results. To begin with, Varro (b.c. 116-28), writing his ' De re rustica ' in the retirement of his old age, probably at one of his villas at Cumae in Campania, or Tusculum, close by Rome, makes the following statement : ' We,' in agro Romano ac Latino, measure in jugera In Campania they measure in versus.' And after stating Italy. 215 that what is called the jugum is what a yoke of oxen can plough in a day, he describes the difierence between the versus and the jugerum to be that the versus is a square with sides of 100 feet, whilst the jugerum consists of two squares or actus, with sides of 120 feet. In Latin, he says, the ' actus ' is called an ' acnua/ There is nothing in this statement of Varro taken alone to suggest that strictly Roman measures were extensively in use beyond what was acknowledged as strictly Roman ground. His words ' afud nos in agro Romano ac Latino,' distinctly suggest the contrary. The Latins of Latium, which lay between his two villas, had not yet absolutely lost their separate identity. The rich plain of Campania in which his villa at CumcB was situated, was not strictly Roman ground, and there the ' versus ' was in use in his day. Varro does not mention the fact which later writers do, that the versus was prevalent in the regions occupied by Oscan and Umbrian tribes before the Romans came. That he hmited his statement as regards the prevalence of the versus to Campania did not mean that it was in use nowhere else. He was making no wider statement than what was under his own eye as a resident at Cumse. He was not thinking of a wider horizon.^ In the first century a.d. in a fragment attributed to Frontinus (at one time Roman Governor of Britain) is contained a passage which widens the horizon and ascribes the use of the versus or vorsus to the Oscans 1 There is a passage in another of his works {De lingua Latina, V. 34, 35) which is important. He describes the ' actus minimus ' as contained in 4 X 120 feet, i.e. a slice of the actus making a strip of the form of 1 X 30, with an area of '420 ares. If Boman measures were widely prevalent in Italy this unit might well turn up here and there in the customary measures ; but apparently it does not do so. 216 Mediterranean Land Units. and Umbrians generally, thus covering the ground occupied by these ancient tribes and by no means confined only to Campania. The writer states that the first ' modus agri ' was the square of 4 equal sides of 100 feet ' quod Grseci plethron appellant, Osci et Umbri vorswm.' Lastly, at about the same date Hyginus states that he had found in many places other measures than the jugerum in use, with other names ; for example what is called the ' versus ' in Dalmatia. And he proceeds to describe its contents exactly in Roman feet, making it a square with sides of 100 of the Itahan foot of "275 m. And he adds that to avoid confusion, when the vorsus is in use he gives both measures — ' jugera tot, versus tot.' AH this seems to show that at the time of these early agrimensores the Romans kept their jugerum very much to themselves, and that beyond their own ground the vorsus of the Oscan and Umbrian tribes was widely spread and left untouched, though we have no right to press their silence too far as to the prevalence of other measures than the jugerum and vorsus. One other point may be noticed. The Roman centuria consisted of 100 sortes or 200 jugera, in one great square. But Frontinus states that there are those who call other areas centurise, as at Cremona, where there are 110 jugera in the centuria. We may note that whether they recognised the fact or not the centuria of Cremona of 110 jugera would contain 100 Greek medimni, and therefore might very properly be called a centuria. So also in the ' Liber Coloniarum ' ii. (Lachmann, 1. 210 and 261) there is mention of a centuria of 240 Italy. 217 jugera as in use in Apulia. This would contain 800 vorsus of 7*55, and may be significant, as we shall find the customary measures of Apuha to be generally multiples of the vorsus. So much for the evidence of the Eoman agrimensores. It has throughout emphasised the importance of the vorsus as probably the agricultural unit of the Oscan and Umbrian people, who preceded the Greek and Roman colonies. Before passing to the customary measures, it may be well to recur to the position of the vorsus in the Eastern system of measures. It has been explained already how in the Heronian tables it was connected with one of the two ' modii ' which measured the area for a modius of seed on two different kinds of land. That for good corn-land was the Greek plethron of 100 four-pahn feet square, having 32"8-33'5 m. to its side and an area of 10"76 to 11 '23 ares. That for inferior land had 120 four-palm feet to a side, with an area of 15'29 to 16'19 ares. The two squares of half its diagonal had each a side 27 '5 m., and each was a vorsus of 7 "64 to 8 "09 ares. We may approach the hmits of variation in the vorsus from another direction. The length of the side of the vorsus as described by Hyginus was 27 '5 m., i.e. -^ of the leuga of lower standard. But if the vorsus varied within the same hmits as the leuga, its area would vary between 7'56 and 7"97 ares. As already said, Martini describes the customary measures of 102 locahties in Italy. In 79 cases out of these I have been able, with some confidence, to recog- nise of what known unit they seem to be multiples. 218 Mediterranean Land Units. Most of them will be found marked on the accom- panying map. The analysis of the 79 cases is as follows : — Ares. 49 are multiples of the vorsus, i.e. of one of the two squares of the wet-land modius . . . 7'66-7'97 15 of the com- or seed-land modius of . . 10'76-11'233 7 of the Greek stade acre ^\ of the square stade 6 of the Roman actus 2 of the Greek medimnus The occurrence of the vorsus, i.e. one of the two squares of the wet-land modius of the Heronian tables, side by side with the corn- or seed-land modius, was natural, as they belonged to the same system of measures, whilst the connection of each with the same itinerary measure — the leuga— is also suggestive. The side of the vorsus was -^ and that of each of the two squares of the corn-land modius yoo' of the leuga. The predominant fact seems to be that the state- ments of the early Roman agrimensores as to the preva- lence of the vorsus in Oscan and Umbrian lands is fully confirmed and extended by the evidence of the custo- mary agricultural measures as given by Martini. But before considering further the distribution of the vorsus unit on the map, I am anxious that the reader should be fully aware of the nature of the evi- dence to be drawn from the multiples of the vorsus unit, i.e. how the unit is arrived at in places where the local measures are not actually divided into units equal to the vorsus, as well as in cases where they are so divided into units of the area of the vorsus, though no longer so called. By way of illustration we may take the case of the soma of SinigagUa (near Ancona on the Adriatic). 756 [To face p. 218 7-66 NOI/ARA 7 6 TURIN BRESCIA 7 62 7 98 MILAN Italy. 219 Martini describes it as having an area of 124*79 ares, and as a square of 400 cannas of 5-58 m. (10 local ' feet ' of '558) . This is the information given by Martini. It is easy to see that the local ' foot ' of "558 is two old Itahan feet of •279, 100 of which made the side of / 4> £qua/ to the Versus 7B 31-20 ares 1 24-79 ares FiQ. 55 the vorsus of higher standard ; so the canna is really 20 of such feet. A square of 20 x 20 cannas of 5*58 m. would make the side of the soma 111 "6 m. {i.e. -^ leuga). It would divide, in the usual way of 4 by 4, into 16 vorsus of 27"9 m. to the side and 7*8 ares in area. So that here 220 Mediterranean Land Units. we get at once not only the vorsus but also its connection with the leuga. Fmther down the coast at Ascoli almost the same measmres recmr. The ruhUo of 123-13 ares contains 400 square V /' Quarte C oF (7 69) Z Vorsus) 15 39 6- 0*- 123 13 Fig. 56. canne divided into 8 quarte of 15*39 ares, i.e. the wet- land modius of 2 vorsus of 7 "69 ares. Inland, but stiU east of the Apennines, the local customary measures are again almost a repetition of those of Sinigagha. At Foggia and Barletta the versura Italy. 221 is a square of 122-63, i.e. 16 versus of 7-66 ares, whilst over the Apennines and ahnost in Varro's Campania the tomolo of Benevento is a square of 30 "659 ares equal to 4 vorsus of 7 "66 ares. And amongst this nest of apparent multiples of <^- <^' 0^ // ft // 30 653 / y 122 63 ares Ko. 57. the vorsus unit we may note another form of the combination of 4 vorsus, this time put into two squares. At Bari on the coast there are reported to be two measures in customary use identical in area but different in measurement and name. (Figs. 58 and 59.) 222 Mediterranean Land Units. Here obviously we have 4 vorsus of 7"82 ares con- tained in two squares. These cases I have treated in detail by way of illustration, broadly speaking, all of them from the Tomolo oF 800 palmi of I- 98m 3 1 28 ares Pio. 58. •§> 1 Aratro of 1250 square pass/' of l-58m. 31 18 ares Fig. 59. Apulian district in the south, hovering round Varro's Campania. But similar illustrations can be taken from North Italy. Thus at Novara and Varallo in Piedmont the cus- tomary unit is again the square of 30'66 ares — the moggio — divided into four pertiche of 7"665 ares, each of which though called a pertica is of the area of the vorsus. Italy. 223 At Pavia also the pertica of 7 "7 ares is the unit in customary use. So also at Nice, the eminate of 7*72 ares, though not so called, is again the vorsus unit, and the starata / Per "tea \z7 Sm / Nr^ / \^ / *%...> / \ \ °7 \ / \/ x'" / \fo y \ y33- 02m B.iolca of Cuastalla 72 Ti vole 30 525 55m 7a04'n Fig. 60. of 15'444 ares is its double. At the same time the hiolca of Guastalla is as nearly as possible the moggio of Novara turned into two squares. In Piedmont there seems to be a cluster of customary Q 224 Mediterranean Land Units. measures in use five times tlie vorsus unit instead of four, with an area of 38 '104 ares. At Turin the customary unit (before 1818) was the giornata di Piemonte of 100 square tavole of 616 m., or 400 square trabucchi of 3-08 m. (10 Greek feet), and the local itinerary measure is the mile of 800 trabucchi or 4 of the sides of the giornata. And these measures seem to be common to Mortara and Voghera and are reckoned as containing 5 of the pertica of Pavia, and so we may hardly be wrong in regarding them as multiples of the vorsus unit. :<<^ 6\' .^^ 38. ares Fig. 61. There are a few cases in which it may be at first sight doubtful whether the unit in use should be reckoned as a multiple of 3 vorsus or two of the corn-land modius. Thus the tornatura of Cento is a square with an area of 22-63 ares, which might be either 3 vorsus of 7-55 of lower standard or 2 corn modii of 11-36 of higher standard. So also the tornatura of Faenza with an area Italy. 225 of 23-02 might be either 3 vorsus of 7-66 or 2 corn modii of 11 '5 ares. But there are not many such cases. I have been anxious, as aheady said, by these illus- trations to make quite clear on what kind of evidence I have marked on the map the presence of the vorsus unit in the customary agricultural measures recorded by Martini as ' abusively ' in use in various locaUties. Unhappily the evidence is absent for Savoy ; but, passing from the Itahan evidence of Martini, it may be helpful to cross over the Alps to take note of what can hardly be other than an overflow of the vorsus unit into the French districts contiguous to Savoy. Here we have the evidence of the French Tables with their much greater details of the measures of every conunune. In the Departement of the Hautes Alpes, e.g. : In 9 cantons or communes the sHeree or iminke is reported as . . . 15-19 ares (= 2 of 7-59) In two others as ... . 15-65 „ {= 2 of 7-82) 15-69 „ (= 2 of 7-84) In two others the eminee . . . 7-59 „ (= the vorsus) 7-97 „ In the south-eastern side of the Departement of Ain, which is contiguous to Savoy, in the arrondissement of Belley, there is similar and still stronger evidence of traces of the vorsus unit. In 8 communes the ouvree for vines is . 3-799 (= J vorsus of 7-6) and the journal or seytive is In a cluster of 10 communes the ouvree is . and the seytive ..... In another of 5 communes the bichette or mesure is .... . and the journal ..... In another of 6 communes the seytive is In two others the ouvrte is . . . q2 22-792 (= 3 7-601 (= 1 30-390 (= 4 5» „ 7-6) „ 7-6) „ 7-6) 7-598 (= 1 22-792 (= 3 22-792 >> Si „ 7-6) ,. 7-6) 3-799 (= I it „ 7-6) 226 Mediterranean Land Units. In the absence of direct information from Savoy, this evidence of overflow of the vorsus miit into French territory is remarkable and confirmatory of that of Piedmont. There is nothing but the Alps between the arrondissement of Belley and the Itahan district of Turin and Novara. And we may note that the French evidence confirms the point on which we might other- wise doubt, viz. that the Itahan units of 22'63 and 23"02 may fairly be identified as triple multiples of the vorsus unit. Now at last, taking a general view of the apparent prevalence of the vorsus unit as shown by the map, the result seems to be that it was prevalent over a far wider area of Italy than the Roman agrimensores might have led us to expect, and perhaps than they knew. It seems to pass the bounds of ancient Oscan and Umbrian occupation into the region of the ancient Ligurian tribes which have always been regarded as belonging to an altogether difierent race. The ethnological question thus raised is outside our present object and must not be dwelt upon here. But there are points connected with the wider extension of the vorsus which have a bearing upon the possible travel of agricultural custom and methods, as shown by customary measures, pointing towards Armorica and Britain. Before we pass westward along the Ligurian coast, or it may be over the maritime Alps to the district round the mouths of the Rhone where the vorsus unit under other names makes itself prominent again in the customary measures of Provence, we may pause to consider how far we may already have taken some steps towards our end. First we have clearly found in the customary areas. Italy. 227 consisting of 16 of the vorsus unit, connection with the leuga. The sides of these squares are ^o of ^^^ leuga. The smaller squares of 4 vorsus units have sides of half that length and are -^ of the leuga. In those cases in which the area of 4 vorsus is found in two squares, it follows that the sides of the squares must be -^ of the half-diagonal of the leuga. Accordingly, in Modena, when the unit is in two squares but with a somewhat reduced standard, the area being only 29'22 and 28*36 instead of 30 ares, the mile is described as 1569 m. at Modena and 1592 at Reggio de Modena. The itinerary and agricultural measures seem to have come down separately, but the mile of 1569 to 1592 m. is nevertheless clearly the half-diagonal of the square leuga and 40 times the length of the sides of the square of the area of the 4 vorsus in two squares. Thus if we took the biolca of Guastalla as our pattern, 40 times the side of its squares — 39*02 m. — ^would make a mile of 1561 m. ; whilst if we took the Bari example, 39*6 X 40 would make a mile of 1584 m. And that these Modena miles from 1561 to 1584 m. are really half- diagonals of the leuga, we may be reminded by the fact that they are practically just 10 times the length of the Breton arpent. This brings us to the last point which need be noted at this moment, viz. that the long side of the two squares of the 4 vorsus unit is just half the length of the furrow of the Breton arpent ; so that a square of the area of 8 vorsus would become identical with the square of 78 m. to a side which we found to be cropping up on the map between Brittany and the Mediterranean. We arrive at the fact that, consciously or not, the square of the Breton furrow contained 32 vorsus. CHAPTER V. THE OVERFLOW OF THE VOBSUS INTO THE LIOUBIAN DISTRICT TO THE WEST OF THE MARITIME ALPS. We have seen from the clear evidence of the French Report that the vorsus unit overflowed across the moun- tain region of Savoy (as to which unfortunately evidence is wanting) into the closely bordering district of the arrondissement of Belley in the Departement of Ain. We have now to examine the continuation of the Ligurian region extending by the coast to the mouths of the Rhone, or perhaps overflowing the maritime Alps into G-aul. The district in view extends beyond Provence. It very nearly coincides with the eight Departements in which the oUve is grown. Such a region would naturally be tempting to early Phoenician or Greek colonisation. Heyn in his interesting chapter on Ancient Ohve Culture favours the view of its introduction by Greek colonists. Apart from colonisation, its proximity to the Ligurian district of Italy would make it natural that its agri- cultural units should resemble those of northern Italy. The problem of its customary measures has therefore an interest of its own. The statements of the Roman agrimensores describing the prevalence of the vorsus in Italy do not directly apply to this district, but in the absence of direct early evidence an examination of CHE Ft NIEVRE ...-;;■..; ■■■■: JURA ,'■ A LLI E R SAONE et LOIRE "■■:. 331 C R E us E \l-787 ;• ■', A I N J I 867 { RHONE':.,, 3 797 {LOIRE \ LYONS •■•■■■•' '-, IpUY DE dome '•-..2 -563 '■■■■•. ., ( /SERE S "V 1-678 .,.;'■% ,.,. .,-.■■ ■.,„„^„ji„y •., ...v' .... 2-046 ( CORr'eZE ...••• 1 99 1 ■'/ { I HSA VOIE • ■; 339 ■J'"'--. SAVOIE 341 CANT A L I 7 8 1 .. 1.1-673:' '\ H LO I R E j \ :-^2-982'^ 1-95 .:■"•■■ 22 • 792 / to ■2 003 -2071... 7-597/ 7-990 2.-046 .......^ 3 3 3 ■■■■-■:, ■-■" HALP€ S .J 1-989 -2 20 AVEYRON TARN \ LOZERE :> 1-992 ..ARDECHE ; ■DROME ( l-90-2;106/ ( D 3-94- -,td \ALP£S ■■.„ 1-97-2-10 L.I993" Xh^ARITIMES ■,VAUCLUSE I., B ALPE5 ,,.,r-'- \ 785 ■ PYRENEES OR The Vorsus. 229 the customary measiires given in the French Tables from which we have learned so much at once presents a similarity suggestive of continuity, which can hardly be misleading. Looking simply at the extraordinary variety of measures prevalent in this district of olive cultivators as they are given in the Tables, at first sight, it may well seem hopeless to reduce them into any orderly system. But there is at the basis of them all the lineal measure of the canne of 8 pans of 8 thumbs or inches. The length of the canne is remarkably constant within a variation of from 1"97 to 2'1 m. Taking the Departement of the Bouches du Rhone as typical, in 35 out of 39 communes the canne varies between r97 and 1*99, and in only four communes is a higher standard prevalent. The area of the square of 100 (10 x 10) square Cannes would vary between 3"89 and 3'95 ares, i.e. half the vorsus unit of from 7'78 to 7'90 ares. And as in this Departement the larger measures are generally multiples of the square of 100 square Cannes, they become multiples of the vorsus unit. In order to put the evidence in the simplest shape I have marked on the map the length of the canne of each Departement belonging to the district. It wiU be seen that the canne varies from about 1'97 to 2*1 m. The canne is of 8 fans, and taking 1"99 m. as its average length the pan would be "249 m. The pan, therefore, is the natural foot (•246-'251 m.), but the natural foot or pan is here divided into 8 thumbs or inches instead of the usual 9. Why is this ? The reason seems to be found in connection with the Italian vorsus-foot of '275 m., which passes over or through this 230 Mediterranean Land Units. region to become the customary foot of Spain. The vorsus-foot divides into 9 thumbs and 8 of them make the pan of "249 m. The length of the cann£, of 8 pans is 8 natural feet, or 6 of the Greek four-palm foot of "332 m., the decempeda of which becomes the estadal of Spain. The four-pahn foot, moreover, extends northward and becomes the customary foot of the Burgundian district ; and further north Drusus used it in laying out the conquered lands of the Tungri, where it became known as the Drusian foot. The larger agricultural areas are generally multiples of a unit of 100 square cannes, and the area of this unit is just half that of the vorsus. Four of the squares of 10 x 10 cannes equal in area the wet-land modius. The favourite multiples are those of 1600 and 1800 square cannes. The 1600 multiple brings us again to the quarter of the square of the Breton furrow but of higher standard. If we turn to the Alpine Departements the canne of the Basses-Alfes is still 1-986 m., and the journal 500 square cannes — i.e. 5 half- vorsus. In the Alpes-Maritimes, the seteree is 15*445 ares, i.e. two vorsus of 7 '72 ares. In the Hautes-Alpes the canne is 1-95 m., but the measures generally are too various to make it worth while to dwell upon them. It is more to our purpose to recur to the canne and its pecuhar structure, its division into 8 pans, equal to the natural foot of 3 pahns or 9 thumbs, and the division of the pan into 8 thumbs instead of the natural nine. In this Ligurian district we seem to have a system of agricultural measures built upon the pan or natural The Vorsus. 231 foot and producing a unit equal in area to the vorsus in two squares and a wet-land modius of 2 vorsus making one square. Now turning back once more to the Danubian district (Jassy), we cannot fail to note a striking coincidence amounting to practical identity in the details and method of the construction of the agricultural unit. Answering to the pan of the Ehone Valley is the Danubian falma of "248 m., i.e. the natural foot, and as we have seen, both are divided into 8 thumbs instead of the natural nine. As already said, we may even see a reason or excuse for this, for adding a ninth thumb to the natural foot of "248 the addition at once makes the "279 foot of the vorsus known to the agrimensores and the Heronian table as the Itahan foot. Nor is this all. For the Danubian ' stingeni ' of 8 palms (r98 m.) answers exactly to the canne of 8 pans of Provence. Finally, coming to the larger units the result is a multiple in the one case of the corn-land modius and in the other of the wet-land modius of two vorsus. AVhether the original route of importation was by the Danubian route, or round by the sea, the result is the same ; there is the common tradition and possession of a system of measures and methods in agricultural matters engrained in the minds and habits of the people by long-continued custom in the lower valleys of the Rhone and of the Danube. The prevalence of the Oscan and Umbrian vorsus in Italy over a wider region than could be easily explained by Greek colonisation points to these local customary measures — especially the vorsus — being an earUer com- mon possession on both sides of the Maritime Alps, allowed to continue by the Phoenician and Grecian intruders into the region of the old Ligurian inhabitants. 232 Mediterranean Land Units. The fact that in both districts they are based on the natural foot, though disguised as the fan or the palma, points to the early absorption of these land units in the common substratum of agricultural usage over a wide region of Western Europe. CHAPTER VI. THE SPANISH AGRICULTURAL UNITS. Pursuing the inquiry into Spain, the customary units of agriculture seem to belong to two types crossing one another, closely connected and yet representing apparently two separate influences. First, we note that the ItaHan vorsus-foot of "274 to "279 m. has become the national foot of Spain, and the estadal of 3 "28 to 3 "35 divided into 12 of these feet is widely prevalent. Here evidently we come once more upon the Ligurian vorsus, for 10 x 10 of the estadal would make a square with 32 "8 to 33 "5 m. to the side, i.e. the Greek plethron^ one of the two squares of which when put into that form is the vorsus as described by Hyginus. Secondly, besides these connections with the vorsus, we seem to get more or less direct traces both of the Breton arpent with its 156 m. furrow and area of 48 •624 ares and also traces of the square of half its furrow with an area of 60"78 ares. The first set of measures seems to have become the Castihan or national ones, whilst the second seems to keep pretty closely to the Mediterranean coast, as though the result of maritime or colonial enterprise from the East. As the estadal is the rod of the Castilian group and 234 Mediterranean Land Units. is of the higher standard, so the canne of the coast districts becomes the rod of the units which seem to aim at the Breton arpent, and is of the lower standard. Thus, to begin with the maritime district, the canne of Barcelona was 1-555 m. in length, i.e. xoo of the Breton furrow. The customary agricultural unit of Cuartera 24- 48 48-96S ares Mojada of Barcelona Fio. 62. Barcelona was the mojada of 2025 square Cannes forming a square of 45 x 45 cannes, or 69 "975 m. to a side. Thus the square of the mojada was very nearly equal in area to the Breton arpent (48*624 ares) and the canne very nearly y^ of the Breton furrow— 1-555 m. (the Breton furrow being 155-9 m.). The mojada of Spain. 235 Barcelona was thus practically the Breton arpent in a square, and being divided into two cuartera, each of the latter equalled in area the Breton half-arpent. Between Barcelona and Valencia Hes the coast town of Tarragona, which gave its name to the Tarra- conensis Provincia. Its agricultural unit was the ' cana de rey ' reported to contain 2500 square cannes. Vorsus 7-6 Cana 2500 = 8 Vorsus de rey square canas of 7-6 ares 60-34 ares Fio. 63. The 'cana' was again r56 m., i.e. -j-oo of tt^ furrow of the Breton arpent. 50 X 50 of this cana made a square of 78 m., i.e. half the Breton furrow and containing an area of 60"84 ares. This quarter of the square of the Breton furrow is identical with that marked upon the map showing the extent of the prevalence of the Breton arpent, which 236 Mediterranean Land Units. was seen stretching southwards from the Breton district to the Mediterranean (see p. 131), Inland from Tarragona at Saragossa the hraza or estado of 1-54:3 (2 varas of -771) seems to be a reduced example of the Tarragona cana. The cuartal of 4 ahnudes of 100 square varas, i.e. 20 x 20 varas, had an area of 2'38 ares, i.e. -^ of a ' canxi de rey ' reduced to 59"50 ares. The tendency seems here to be to lower the standard beyond even the ordinary limit of variation. In Valentia the customary measures are peculiar. The cuerda or corde is of 20 X 20, i.e. 400 square brazas /^ Fane gacia i? S3, / /6-62 49 86 ares Fio. 64. of 2'038 m., and the fanegada consists of two squares each of 100 square brazas. It is therefore half the square cuerda (see fig. 64). Here again we have repeated, with a slightly exaggerated standard, in the hraza, the canne of the French Ligurian district ; in the fanegada, the vorsus in two squares ; and in the square cuerda, the wet-land modius — ^all of higher standard. But the Valentian measures present another feature, connecting them again except in standard with those of Barcelona. Six Valentian fanegada made a cahizada. Spain. 237 with an area of 49 "86 ares, thus shghtly exaggerating the area of the Barcelona mojada. So here again we are reminded of the area of the Breton arpent, but in this case put into three squares instead of forming one as at Barcelona, and with exaggerated standard. Further a still larger agricultural area is reported, viz. the yugada of six cahizadas with an area of 299 ares and in the form of two squares. There may perhaps be traces here of a conflict between the two influences and two standards and an attempt to bring them roughly into harmony. Passing still southward on the coast we find reported both for Malaga and Cadiz that they shared more or less in the ancient measures of Castile. But for Malaga an agricultural unit is reported in a fanega of 60'370 ares containing 8640 square varas. This fanega is again J of the square of the Breton furrow Hke that of Tarragona, but it is described in an odd number of Castilian varas. For Cadiz on the other hand Martini reports that the ancient agricultural measures were those of Castile. We are thus left much in the dark as to the original customary measures of the ancient Bsetica. Passing now from the customary measures of the maritime districts to those of Castile, as reported in use before the year 1800, we come back to the estadal of 3'34 m. divided into 12 of the famihar ItaHan vorsus feet of higher standard, viz. "279 m. The chief Castihan agricultural unit, according to Martini, was the ' fanegada de tierra,' a square of 24 x 24 estadals with an area of 64"39 ares and divided into 12 celamin of 48 square estadals. (Fig. 65.) Consciously or not, this ' fanegada ' contained 4 of 238 Mediterranean Land Units. the wet-land modius or 8 of the Ligurian vorsus, but of higher standard. .(<^- 6-66 m. Celamin oF\2-24Est3da/s^^^ ^^^^ 1609^^ 1 64 39 ares Pio. 65. There was also an aranzada for vines, a square of 20 X 20 estadals with an area of 44-72 ares. (Fig. 66.) And this, whether so recognised or not, was equal to 4 corn-land modii again of higher standard, each modius being a square of 100 Greek four-palm feet to a side, so that both the wet-land and the corn-land modius were present or hidden in the Castile measures. Martini also mentions yet another fanegada or fanega of 4900 square varas of Castile. The vara seems to have been that of Burgos, "835 m. in length, Spain. 239 70 X 70 of which would make a square of 58-52 m., with an area of 34-238 ares. Here we have clearly, consciously or unconsciously to local knowledge, the stade acre, i.e. yo of the square stade in area, here put into a square. The same agricultural unit is reported as in use on the coast of the Bay of Biscay at S. Sebastiano, where > (<^- Corn land modius II- la 44-72 ares Tio. 66. the vara, hke that of Castile and Burgos, is -837 m. and the fanega of 70 X 70, i.e. 4900 varas, becomes a square of 58-59 m. with an area of 34-328 ares. So much we learn from Martini of the smaller agricultural customary units of Castile. He reports, however, two larger units : (1) The Castihan yugada of 50 fanegadas forming two squares (401 X 802 m.) and containing an area of 3220 ares. This yugada would R Mo Mediterranean Land Units. thus equal in area 200 wet-land modii or 400 vorsus of higher standard ; (2) a still larger unit, the cahallaria of 60 fanegadas, with an area of 3864 ares, thus con- taining 240 wet-land modii or 480 vorsus of higher standard. We may note here also that the vara of Castile or Burgos was connected directly with the Spanish itinerary measure — the lega real. This ' lega ' was 8000 varas or 2000 estadals in length, i.e. 6687 m., being thus of the same length as the Welsh milltyr of 1000 Welsh land-ends, but measured by the higher standard of the natural foot and equal to three leugee of 2229 of 12 stades of 185 to 186 m. We note at once that the side of the aranzada (20 estadals) is exactly ^ho of the length of the ' lega real.' And what is still more striking is that the small end of the celamin of 2 X 24 estadals was y^g^ of the ' lega real,' thus possibly suggesting a common tradition with that involved in the 1000 land-ends of the Welsh milltyr. But it is not so with the yugada. To bring it into direct connection with the lega real and the leuga, its two squares must be converted into one square, and then, within the normal limits of variation in standard, its side (567 m.) would be equal to \ of the leuga or yj of the Welsh milltyr and the Castilian lega real, all of the higher standard. In whatever form, in one square or two squares, consciously or not, the area of the yugada of Castile was ^ of that of the square leuga. We may note also that it was just double in area that of the lesser Brehon ' tir- cumail.' That is to say, one of the two squares of the yugada would equal the tir-cumail on the lower Spain. 241 standard of the natural foot. It may be remembered that tlie Brebon tir-cumail, like tbe yugada, was built up upon the foot of 4 natural palms. On the lower standard of the natural foot the tir- cumail was exactly as below. On the higher standard ,(<' .^ Yy/C g / X. 3320 ares Fig. 67. it would be 289 x 579 m., but though measured in Ireland in natural feet of lower standard its sides are divisions of the leuga of higher standard. It may further be noted that the yugada put into a square of two tir-cumails would contain almost exactly 128 Koman jugera in the normal shape of two squares and 40 Normandy acres 1 x 10. In its present shape of two squares the yugada would contain 80 Enghsh e2 242 Mediterranean Land Units. statute acres in their proper form of 1 x 10. The meaning of aU this is a part of the problem to be explained. If we may take the customary measures of Lisbon as representing those of Lusitania the agricultural unit was the geira of 4840 square varas of 1-09 m. or 1210 square hraqas of 4-36 m. The area of the braga being 4-778 and that of the geira 57*816 ares. We get httle guidance from this except that in form the geira was 1 X 10 and that in area it would contain 5 corn-land modii of 11-56 ares, or 2 medimni of 28-91 ares, in either case of too exaggerated standard, if standing alone, to be relied upon for any inference of origin. It should be noted, however, that the furrow of the geira is of the same length as the long side of the Valentian yugada. The estadio of Lisbon is almost identical with the length of the furrow of the Irish and Northumbrian acre, viz. 256 m., though it does not seem to enter into the agrarian measures. Before we pass from the Spanish customary land units as given by Martini, the reader should be reminded that the evidence has been confined to that of the Spanish customary measures reported from the modern point of view as ' abusively ' in use in spite of modern legislation. It may be worth while to consider whether there may not be some direct historical evidence in support of or against their antiquity. Unfortunately it hes in very small compass. Speaking of Bcetica Columella (' De re rustica,' V. ix.) mentions an ' actus ' of 120 feet square which the rustics called an agnua. If it had been identical with Spain. 243 the Roman actus he surely would have said so. We may take it therefore that it differed from the Roman actus. Now the Spanish estadal, being 10 Greek four-palm feet in length, divided in Spain into 12 of the vorsus-feet, though of higher standard, seems more Ukely to have had a Greek or Phoenician than a later origin. It may probably even have belonged to the common substratum of still earher agricultural tradition. And if we may take the 120 feet of the side of the agnua as so many Spanish customary feet, i.e. of '279 m., then we bring the agnua back at once to a square of 10 X 10 of the estadal, i.e. of 100 Greek four-palm feet to the side, which we have seen to be the customary Castihan aranzada for vines containing 4 corn-land modii of higher standard. Columella also on the same page mentions another measure called the ' porca,' which he describes as 30 feet in breadth and 180 in length. Assuming as before that the feet were the Spanish customary feet of '279 m. the area of the porca would be from 4-083 to 4'192 ares, i.e. half the vorsus of higher standard, or one of the two squares of the ' fanegada ' of Valentia. It would be very nearly ^e of the Castilian 'fanegada de tierra ' of 64"39 ares. There is not much rehance perhaps to be placed on this scant evidence, but so far as it goes it seems incidentally to confirm the antiquity of the customary agricultural units. Now, if on a review of the Spanish agricultural customary measures we were tentatively to divide them into the two sets — (1) the Castilian measures pointing back to the vorsus or wet-land modius, and (2) those, chiefly on the maritime border, pointing 244 Mediterranean Land Units. towards the Breton arpent, we should have to recognise the significant fact that the Castilian itinerary and agricultural units seem to be consistently based upon the higher standard of natural measures, whilst the Maritime units seem to be as consistently based upon the lower standard. The variations in standard being within natural limits might perhaps be overlooked if they were not so constant, and if there were not a more or less obvious reason why the maritime measures like the Breton measures to which they seem so clearly to point should be based on the lower standard. We have aU along seen reason to suspect a connection between the Breton furrow and the Egyptian khet. The Breton furrow of 157 m., being three times the length of the side of the khet, naturally perhaps could not well wander away too far from the Egyptian standard or that of the Cyrenian medimnus. And the fact remains that it did not. On the other hand, if we turn to the Castilian land measures with their estadal of 3 •34 m. and foot of '279 m., both of the higher standard culminating in the yugada, and note the connection in area with the smaller Irish tir-cumail whose sides were divisions of the leuga, of higher standard, we cannot avoid connecting it further with the Normandy and British statute pair of acres, both of which are of the higher standard. The closeness of this connection is shown by the fact that the short side of the yugada of Castile is as nearly as possible two furrows of the British statute acre, whilst the diagonal of one of its squares is consequently of the same length as two furrows of the Normandy acre. So that whilst the maritime land measures point Sfain. 245 to the Breton arpent, the Castihan land measures point to the Brehon tir-cumail and the Normandy acre and its connection with the Roman jugerum. The strange consistency in the matter of standard forbids our ignoring it. If we start with the Itahan vorsus as described by Hyginus, it is a square of 27"5 m., 100 feet of '275 m. The side of the vorsus was thus g-Q of ^^6 leuga of 2200 m., i.e. the lower standard. If Spanish customary measures had kept to this standard the foot would have been "275 m. and the estadal 3 "30 m. The Castilian ' lega real ' of 2000 estadals would have been 6600 m. — a milltyr of 3 leugas of 2200 m. — and the whole stream of measures, Castilian and Maritime, would have fitted on, so to speak, at both ends with the Italian vorsus at one end and the Breton and Cornish measures at the other end. But the facts seem to conspire in showing that it was not so, but rather, that some stream of influence or two separate streams had been continuously at work making and perpetuating the same difference of standard both between the Maritime and Castilian measures in Spain and also between the Breton and the Normandy measures in the north of Gaul, not even stopping there but crossing the Channel and becoming just as prominent (as we have seen) in Britain. This is an important element in the problem which, as far as the Itahan and Spanish evidence goes, we carry with us in returning to the consideration of the group of Armorican and British customary 1 x 10 acres which form the special object of this inquiry. CHAPTEE VII. THE LAND UNITS OF THE DISTRICT BETWEEN THE LOIRE AND THE GARONNE. There is one more district in Graul requiring special attention, viz. that occupied in the time of Caesar by the Pictones and the Santones between the mouths of the Loire and the Garonne. We have akeady noticed that on the Spanish shore of the Bay of Biscay the stade acre in a square was a fairly prominent land unit, whilst it was not so in other parts of Spain. Crossing to the Gallic side of the Pyrenees in the Departement of the Landes the stade becomes all at once almost the only and certainly the predominant land unit. But the sides of the squares which are in area divisions of the square stade are not divisions of the stade itself but of its diagonal, just as the Breton furrow is not a division of the leuga but of its diagonal. There may be a significant reason in this. The arpent of ' St. Sever,' according to the French Official Tables, is prevalent in 134 communes of the Departement of the Landes. Its area is 42-208 ares,^ i.e. 1 of the square stade of lower standard ; and in 52 communes its double, 84-170 ares, is prevalent, being I of the same square stade. 1 The same area and standard as the feddan but probably with no direct connection. Land Units between the Loire and the Garonne. 247 The new point, however, is the appearance of a customary acre or journal of 65 "950 ares still more at home between the Loire and the Garonne. This journal can hardly be other than that of two stade acres, i.e. J of the square stade somewhat below the usual lower standard. Its area is almost exactly that of the Irish or plantation acre with a furrow of 256 m., but here it is in a square. We have again and again been puzzled by the absence from the associated acres of this specially important member of the group, and we have more than once been tempted to consider that it might be after all a double stade acre of unusually low standard. As we find its equivalent here in the form of a square between the mouths of the Loire and the Garonne in the midst of stade acres or divisions of the square stade, notwith- standing the lower standard and the change of form, we can hardly hesitate any longer to identify both with the double stade acre. (See above, pp. 110 and 137.) Thus the typical Irish acre and the journal of the Gallic Pictones are brought into connection independently of the question of connection between Gallic and British and Irish Picts. As the mouth of the Garonne is approached, in the Departement of Charente and of Charente-Inferieure, where the stade acre (yq of the square stade) is strangely predominant, it resumes its usual area of 34189 ares instead of the Pictish 33 ares, and is a square of 58"50 m. to the side. The real home of the acre of the Pictones is obviously in the district near the mouth of the Loire, from which maritime intercourse with Ireland would be most hkely to take place. 248 Mediterranean Land Units. Now if by this identification of the acres of the Pictavi on the Loire with the typical Irish acre we may be led to look upon the latter as pointing to maritime connection with the Picts of Ireland and Britain, we also seem to find a way of escape from the naturally misleading inference from geographical position that the so-called Irish or Plantation acre might be of Cum- brian, Northumbrian, or Brigantian origin. We may possibly look upon it as a Pictish acre, if after all it be not simply a double stade acre, and therefore colourless as to origin and spread over Britain. In either case, considering its prevalence so near to the Roman wall and the district of the struggle with the Picts and Scots, the stade acre or its double would become quite naturally at home in that region as an Irish acre. But if without wandering too far outside the evidence of the acres themselves we follow Professor Eidgeway's interesting conclusion that in the Cuchulainn Saga there is proof of connection in chariots and other things with the La Tene stage of civihsation, then we are led away to Ulster, close to Armagh, for the goal of early Galhc maritime enterprise into Ireland. For it was there that Conchobar and Cuchulainn fought their battles from their chariots, and it is worth marking also that it is in Yorkshire, where the Irish acre is so prevalent, that the remains of burials with chariots are foimd.^ So the evidence of the La T^ne influence brings together the regions on both sides of the Irish Sea, and especially of the Solway Firth. Nor need we be surprised that St. Patrick's and St. Columba's labours should be centred in this region ' Ridgeway's Date of the First Shaping of the Cuchulainn Saga, p. 6. Land Units between the Loire and the Garonne. 249 and that in fact the Solway Firth became the centre of Irish and British monasteries and missionary spirit. However this may be, it is enough at the moment to note that there may be in the practical identity in area of the acre of the Pictavi with that of the typical Irish acre some hint at least of the existence and direction of maritime intercourse between the mouth of the Loire and the Channel, during the late Celtic period. Returning to the Galhc acres of the Pictones it is important to notice how acres of the higher standard seem to be intruding into this home of lower standard acres. In the Departement de la Sarthe, where the journal corresponding in area with the Irish acre is at home and lower standard acres are prominent, there is a journal of 80 perches with an area of 39"629 ares. What is this but two British half-acres forming half of a Nor- mandy acre of lower standard ? And side by side with it is the square arpent of 71 "4 m. with an area of the exaggerated Roman sors 51 "072 ares in area, instead of the real Roman sors of 50'296 which, aU the while, was present in lower standard in an arpent of 70'41 m. to the side and with an area of 49"536 ares. We seem thus to mark here an overlapping of the higher standard over the lower, without, however, eradicating it. And, our eyes thus opened, as we pass gradually eastward, we seem to see more and more evidence of the acres of lower standard having once had a wider spread westward, and then at some later period having been submerged by the influence bringing in the higher standard. To verify this fact the reader has only to recur to the maps showing the spread of the associated acres individually to satisfy himself that 250 Mediterranean Land Units. something like this must have happened. Almost every one of the acres appears in the double relation of lower and higher standard in different co mm unes without variations between. But the point to which attention must be drawn at this moment is the fact that the region on both shores of the Solway Firth, where the Irish Plantation acre was so prominently at home, was also the home of the acre of Dabiada with a furrow of 228 m. and that further north and covering most of Northern Scotland, with a furrow of 226 m., these customary acres have held their own to the present day and been adopted as the Scottish acre. But what is this Scottish acre, with its area of 51 "423 ares, but the Cornish acre of somewhat exagger- ated higher standard ? It is so closely identical in area, though not in form, with the enlarged Roman sors, whose intrusion in the Pictavian district of lower standard round the mouth of the Loire has attracted our notice, that one is tempted to aek whether it may be the enlarged Roman sors put into the form of 1 X 10. PART VI. THE PEOBLEM OF THE BEITISH AND ARMORICAN ACRES. We have returned at length., after a round of extended inquiry, to the main object in view, viz. the considera- tion of the problem involved in the historical position of the Armorican and British customary acres, on the evidence of the acres themselves. The evidence may be shortly summarised as foUows : — They he, speaking broadly, in a group by themselves in a corn-growing country. Nowhere else have we found a similar group, except perhaps in the Venetian end of the Po Valley. The acres are pecuhar in their adherence to the 1 x 10 form. Even when divided into two half-acres the half-acre must be in the form of 1 X 10. The 1 X 10 form may occur sporadically in isolated instances elsewhere, but nowhere else does it seem to dominate a group. The fact that their furrows are mostly divisions of the itinerary measure common to Gaul and Britain or of the diagonal of its square is another important feature. It gives a distinctive character to the closely associated group. Above aU, the intimate connection of these customary acres with" a remarkably similar form of the open field system of husbandry forming the basis of the village community binds together, so to speak, into one traditional system the methods and 252 The Problem of the custom of Armorican and British agriculture. It lends interest and confirmation to the historical statements of Pytheas and Caesar as to the early importance of the corn crops of both countries. Not that the common features of their open fiield systems belong to them exclusively and appear nowhere else. Far from it. The inquiry into the agricultural units of other countries has led us to recognise that the peasantry of the British and Armorican corn- growing districts shared with the rest of Europe in the common possession of an ancient substratum of agricultural custom by no means confined to the geographical area of the group. We have not been able to trace the hne of travel of the group of customary acres, as a group, into its Armorican position. But by following the hne of the common itinerary measure we have foimd members of the group sometimes left stranded here and there as it were along what may have been the hues of agricultural advance westward from the great corn-growing plains on the Euxine or from the Mediterranean basin, or even crossing the Alps between the valleys of the Danube and the Po. But that stragglers away from the group might hnger along the hues of travel is one thing. What the move- ments of population may have been which pressed the people owning the group of acres to settle in the Armorican ' Land's End,' and what forced an overflow of them, with their acres, across the Channel into Britain and Ireland, are quite another thing. In order more closely to reahse the extent of this pressure of the associated acres into a group it may be well to turn once more to the French Tables for the single Departement of Calvados. It may help us to a British and Arniorican Acres. 253 conclusion wtether it be worth while to follow further the indications in the Spanish evidence of a separate origin of the Breton and Normandy acres. The Table for the Departement of Calvados first mentions the ' toise of Paris ' of 1"949 m. This is almost identical with the canne of the Ligiirian district, and the stingeni of Moldavia and the Euxine plains, and the Egyptian cassab, which were practically ahke, and aU of which based their customary itinerary measures not on the leuga but on the diagonal of its square. Next it is mentioned that there are two distinct /eei in use in the Departement. 1. The Paris foot of 12 thumbs .... '325 m. 2. The foot of 11 thumbs '298 m. The latter is obviously an enlarged Roman foot, i.e. •2955 m. enlarged to '298 m. Then follows a hst of perches founded upon these two feet. AU these rods are reported as in use in the Departement of Calvados. Of feet of 12 thtunbs (-325) Of feet of 11 thumbs (-298) Perohe of 24 feet 7-796 m. 22 „ 7-146 m. 18 „ 5-847 m. 16 „ 5-197 m. Perohe of 24 feet 7-152 m. 22 „ 6-551 m. 18 „ 5-360 m. 16 „ 4-764 m. Here surely is confusion and comphcation enough to bewilder the peasantry of a single Departement. But mark how the confusion vanishes when attention is turned to the acres. The acres are as follows in their 1 x 10 form of 4 X 40 perches. (1) 4 X 40 of the perch of 7*15 (22 feet of -325 or 24 of -298) make the Normandy acre. (2) 4 X 40 of the perch of 7-796 (24 feet of -325) = two Breion acres. 254 The Problem of the (3) 4 X 40 of the perch of 5-85 (18 feet of -325) = J a. of 329 acre. (4) 4 X 40 of the perch of 5-197 (16 feet of -325) = i a. of 292 acre. (5) 4 X 40 of the perch of 4-764 (16 feet of -298) = J a. of Jersey 268 acre. (6) 4 X 40 of the perch of 6-55 (22 feet of -298) = 2 stade acres. Thus in the single Departement of Calvados nearly all the associated acres (common to Britain and Armorica) find a place. And all of them, with the exception of the Normandy acre, and British half-acres and the stade acres, are of the lower standard and belong to the Breton group. Besides these there are three acres reported in the form of a square, viz. : (1) The quarter of the square of the Breton furrow = 60-780 ares. (2) The arpent of ' waters and forests ' (10 X 10 perches of 7-146) = 51-072 ares. (3) The arpent of St. Sever = 42-915 ares, and one or two stray acres on which we need not dwell. Now though the Table describes these acres in all the details of difierent feet and perches, it, of course, tells us nothing of what may have been their original object and importance. It does not tell us that the 329 m. acres contained 4 Greek medimni or Egyptian khets, and 10 corn-land modii. It does not tell us that the 292 m. acres contained 8 corn-land modii. Nor does it teU us that the Normandy acre and the British half-acres into which it was changed in crossing the Channel, were strangely entangled with the wet-land modius and vorsus of higher standard and with an enlarged Eoman sors of two jugera. Probably at the date of the Table these facts of original meaning and British ayid Armorican Acres. 255 importance had already for one or two thousand years been absorbed unconsciously in the common substratum of agricultural custom. But however hidden and for- gotten, still these facts have remained embedded all the time in the structure of the acres just as they were in their origin. Notwithstanding, however, that in the construction of so many of the group of 1 x 10 acres the presence of the Greek corn-land and wet-land modius may be referred to the pre-Roman substratum of custom, it would be absurd to assume that centuries of Roman rule may not have left their marks upon them. We have already seen traces of it. And it may be well to consider them more carefully. We have already seen that, according to the Table of Calvados, all the acres, except the Normandy acre, its British half-acre and the stade acre, are distinctly of the lower standard. These three alone are of the higher standard. We have seen that the Normandy acres were measured by two difierent feet — the ordinary French foot of "325 m. and the enlarged Roman foot of "298 m. And if we were to mark on the map the geographical area over which this enlarged Roman foot is reported as stiU in use, we should hardly doubt that its home had been in the Roman province of which Treves was the capital city. We could hardly refuse to recognise that from this centre of Provincial Government it had ex- tended without intermission from Treves to Normandy. That in some communes the peasantry should still prefer the enlarged Roman foot of '298 m. to the ordinary French foot in measuring the acres suggests a Roman survival with a meaning worth examination. 256 The Problem of the It is very significant that kad the Paris forest acre been a square of 240 Roman feet of '2955 m. instead of -298, or 10 x 10 rods of 711 m. instead of 7-146, it would have exactly corresponded to the Roman sors of two jugera. Its area would then have been 50*29 ares instead of the exaggerated area of 51 '072. And the relations to it of the furrows of the Normandy and British acres would have adapted themselves accordingly. But as the Breton region seems to have been one of persistently lower standard, so the Normandy region seems to have been one of higher standard. This is seen in the continuance of the higher standard in the overflow of the Normandy acre in its two or f ou r half acres across the Channel. The British statute acre- retains the higher standard of the Normandy acre. We are surely deahng with a Greeco-Eoman measure in a district of higher standard deeply rooted in local custom previous to Roman occupation. For the inference to be drawn from the rise of the Roman foot from -2955 m. to '298 in the measurement of the Normandy acre is that the Romans raised their measure to make it correspond with local standards too deeply rooted in custom to be easily ignored. The Provincial Government of Treves found itself apparently in a region of measures higher than the Roman standard, and, following the usual Roman habit, adopted it. I think that the annexed diagrams will explain how the Greek aroura and plethron fit, so to speak, into Roman measures. (Figs. 68 and 69.) They will also show the relation of the furrows of the Normandy and British half-acres respectively to the side of the Roman sors and the diagonal of its square and the relation of the sors itself to the Greek aroura. The British and Armorican Acres. 257 square of fig. 68 is of 4 Greek arource. The side is half the furrow of the statute acre and its diagonal is the furrow of the half-acre in form of 1 X 10. The square of fig. 69 is of 4 Roman sortes. The side is the furrow of the British half-acre in form of 1 X 10 ; and its diagonal is the furrow of the British statute acre, in its usual 1 X 10 form. These remarkable relations of the Normandy acre Half the furrow of trie ststoce sere o/^e/< aroura (OO'IOO GreeKSu/6/ts of 2 naturaJ feet FiQ. 68. and British half-acres (1 X 10) cannot be accidental. Had the standard used been the Roman one, whether of Greek or Roman origin, the British acres would have followed on the lower standard. But it was not so, the Roman foot had to be raised to meet a probably existing higher standard ; and the forest acre in the same way had to be raised to the enlarged area of the Roman sors to meet local long-estabHshed custom. s2 258 The Problem of the Whence did the higher standard come ? However imperfect our information of Central and Eastern Europe may be, the broad fact was apparent that in the region of the rival stunde or parasang not only was the stade itself of higher standard and Furrow of British half acre (/ "/OJ f^oman Jugerum / V I^Q. 69. tending apparently to become of higher standard still, but the medimnus of which the acres were mostly midtiples was also of excessive standard. We noticed that at the same time in the Baltic region the contrast in standard was as clearly marked as in the Breton. British and Armorican Acres. 259 The acres were no longer multiples of the medimnus but of the corn-land modius of as low a standard as the Irish acre of 2 stade acres, and even still lower. What movements of population may have caused the passage westward of the acres of higher standard, whether the pressure may have been of Brythonic over Goidehc races, I shall not venture to suggest. But as regards Armorica, the facts remain for anyone to examine for himself not only of the separate existence side by side, or one on the top of the other, of acres of the two standards, but also of the gradual preponderance of the higher standard, especially in the case of the Normandy acre and its British half-acres, as they are found in the act of crossing the Channel. For in spite of the general higher standard of both Normandy acres and British half-acres there are traces here and there of the earher prevalence of the lower standard, but we must not forget the legitimate variation in natural feet, &c., and so strain the test of the standards unduly. The setier of the Chartrain, for instance, was a British half-acre of 39"629 ares instead of 40'85, and several other similar cases might be quoted. On the other hand, other instances occur of acres, belonging to the Breton lower standard group, being raised in standard as they come geographically under the influence of the higher standard. Still, on the whole, it will be difficult not to regard the Normandy acre and British statute acres as interlopers bringing with them the higher standard into a region once very possibly of the lower standard. We turn now to the Breton agricultural units of the lower standard of natural measures. 260 The Problem of the Quite consistently with the general conclusion as to the common substratum of custom over the wider area of North and South Gaul, we have to pass both in Armorica and Britain from the region of the higher to that of the lower standard of natural measures. As aheady suggested, it may perhaps be possible to discriminate by the test of standard between two more or less separate subsidiary currents of influence. In the agricultural units of the maritime districts of Spain, indications appeared of approach to similarity to the Breton arpent both in formation and area and standard. The Breton arpent, as we have seen, seems to have sprung rather from the Egyptian khet and Greek medimnus than from the Greek aroura and Koman jugerum. The Breton furrow was three lengths of the side of the khet. The length of the side of this Eg3rptian unit, as described by Dr. Griffith and Professor Petrie, was 100 Egyptian cubits or 52-4 m., and three times this equalled 157 m. The furrow of the Breton arpent was 156. The side of the khet was one-thirtieth and that of the Breton furrow one-tenth of the diagonal of the square leuga of lower standard. We found these figures in use in the customary itinerary measures of the lower Danubian district which were based on the diagonal of the leuga, and were multiples of the side of the khet. The Breton arpent was of the same area as the Cornish acre, the furrow of which was -^ of the leuga of lower standard. We noticed further that the Cornish gad or goad of 2-75 m. (8 x 80 of which made the acre and which was the decempeda of the vorsus-vimt foot) entered into the measurement of a series of other British British and Armorican Acres. 261 acres showing that they belonged to a common system. Moreover, they are based upon the lower standard of the leuga, of which their furrows were divisions. They were in fact the British counterparts of the Calvados acres of lower standard. These acres had adhered to the lower standard notwithstanding their change of form in crossing the Channel. Moreover, these acres of lower standard measured by the Cornish gad of 2"75 m. were mostly prevalent in the western and most pastoral and tribal region of Britain, whilst the statute acre seem^ed to have spread over that eastern and central portion of England which, according to GaUic tradition as reported by Csesar, had received a strong infusion of Belgic emigrants who claimed that they had introduced agriculture into Britain. In this agricultural rather than tribal and pastoral portion of the island it is probable that settled villages with their open fields around them were already estabUshed, whilst in the western region of the acres of lower standard tribal and dominantly pastoral methods still prevailed, such agriculture as pastoral conditions required being carried on by ' co-aration of the waste ' and the eight-oxen plough — the length of the furrow being a twelfth, or a tenth, or an eighth, or a sixth of the leuga of lower standard, as the case might be. There is another fact which may have a meaning in distinguishing the two pairs of acres. The leuga being the common itinerary measure in both cases it is curious to notice how in the case of the Breton and Cornish acres it is the furrow of the Cornish acre that is directly the tenth of the leuga and the Breton furrow that is a tenth of the half -diagonal of the square leuga ; whilst 262 The Problem of the in the case of the Normandy and British statute acres the change is reversed. It is the Normandy furrow that is directly the eighth of the leuga and the British furrow that is the eighth of the half-diagonal. Whether any or what inference might ultimately be drawn from these singular facts I know not. Whether pressure westward of the higher standard may have had any connection or not with the supposed displace- ment of a Goidehc by a Brythonic people long before the Koman occupation of Gaul and Britain I dare not argue. The two languages were, after all, so nearly related that it is hard to think that the displacement of one by another can have represented a great move- ment of population. But whatever the movement of races may have been, whether amounting to anything like a continuous immigration or conquest in proportion to the amount of pressure inwards and displacement, it might well have resulted in a counter-movement — ^in an overflow of forced emigration of the older race from Western Britain into Brittany, something hke what happened under the stress of Saxon invasion. The reversal of the change to a shorter furrow might then possibly in some mysterious way that one cannot understand mark the direction of the hne of ingress and egress involved in a simultaneous movement of population into and out of the island. I confess that I should be more interested if light could be thrown upon the movements and pressure of population which seem necessary to account for the close relationship between the British and Breton peoples akeady existing before Caesar's invasion. The common possession of the peasantry on both sides of the British and Armorican Acres. 263 Channel of the remarkable group of acres under con- sideration does not stand alone. It is but one indication of the possession of common traditions. In their common maritime skill, in their concerted action when in danger, and even in the more or less vague recognition of the importance of Britain — as well as of Chartres — as an ancient centre of a common Druidical rehgion there surely is indication that under cover of a substantial identity of race and language they had long been closely related. All these things confirm the evidence of the acres that the closeness of their intimacy was not severed by the Channel between them. These are not matters, however, which can be dealt with further in this chapter, and once more I beg the reader to make due distinction between the facts and any suggested inferences from them. The mention of the latter may, however, not have been irrelevant to the point under consideration, viz. — whether the influences which moulded the agricultural traditions of Brittany and Western Britain may not go back to different and perhaps even earher sources than those which affected the agricultural measures of the Belgic district of Gaul and of Eastern Britain, Turning once more to the specially British aspect of the problem, and crossing the Channel with the suspicion that the higher standard of the Normandy and British statute acre was an intrusion into what may have been originally a country of lower standard, the temptation can hardly be resisted to follow the clue of Eoman influence from Gaul into Britain. Why else should the furrow of the acre which became the Scottish acre be raised in Scotland from the s8 264 The Problem of the Cornish length of 219 or 220 m. to 226 m. and 228 m. unless indeed to raise its area to the higher standard of the exaggerated Roman sors ? Or to put the point in a still more direct way, how are we to account for the similarity in area between the Scottish (1 X 10) acre and the French forest acre in a square, both of which we have seen to be equal in area to the Roman sors raised to an exaggerated standard by raising the Roman foot from "2955 to '298 m. The area of the French forest acre in a square was 51*072 ares and that of the Scottish acre is reported to be 51 "423 in a 1 x 10 form. The facts regarding this Scottish acre are suggestive. Its prevalence both in Ayrshire and Ulster, thus forming a connecting hnk between the Scottish and Irish portions of Dalriada, and leaving to this day the Cunning- ham acre with its furrow of 228 m. as the customary acre of East Ulster of similar area and form, cannot be over- looked. It seems to carry its presence back to the sixth century and to touch a real movement of population. It is again a curious fact that deahng with a Scottish customary acre which is really of the Cornish type, but with exaggerated furrow, we should find the district where it was present in Scotland occupied according to Ptolemy by a tribe of the Damnonii, by whom so much of the Cornish peninsula was also occupied. Mr. Skene, without thought or knowledge of the similarity in acres, seems to accept the presence of a ' Cornish element ' involved in the tribal name, without hesitation, on Ptolemy's authority. Speaking of the 'Kingdom of the Britons' of Alclyde after the departure of the Romans he writes! : — 1 I. 236. British and Armorican Acres. 265 The population of this Kingdom seems to have belonged to the two varieties of the British race — the southern half including Dum- friesshire, being Cymric or Welsh, and the northern haU having been occupied by the Damnonii who belonged to the Cornish variety. . . . The Kingdom of the Britons had at this time no territorial designa- tion, but its monarchs were termed Kings of Alcluith and belonged to that party among the Britons who bore the pecuhar name of Romans and claimed descent from the ancient Roman rulers in Britain. In another place ^ he speaks of ' the great nation of the Damnonii extending as far North as the river Tay . . . They possessed 6 towns — ^three south of the Firths and three north of them.' And again: ^ ' The fertile plains from the Tay to Galloway were entirely possessed by one great tribe, the Damnonii.' He speaks ^ of the Roman Wall between the Firths of Forth and Clyde as constructed through the heart of the territories of the Damnonii, thus dividing the nation into two parts, one of which was included within the province and subjected to Roman government, while the other remained beyond the boundary of Roman Britain ; three of the towns mentioned by Ptolemy being within the province and the other three north of the wall. The existence of this tribe of the Damnonii in the North is accepted by Elton as well as by Skene, though probably neither of them knew, as certainly Ptolemy did not, that the Cornish acre seems to have followed the Damnonian name, though with an exaggerated furrow and area, accounted for possibly by Roman surroundings. Thus, following the evidence of the acres in con- firmation of what Mr. Skene arrived at from quite other evidence, the acres seem to have passed through 1 I. 72. a p. 127. 3 p. 128. 266 The Problem of the the Roman period like the leuga unchanged, except in standard, and landed, so to say, in a most unexpected combination of customary acres side by side in the district which, having been the very centre of the struggle of the Romans with the Picts and Scots, had become so fully Romanised that on the Roman departure it formed a new Romano-British centre of influence from which came the Cumbrian invasion of Wales by ' Cunedda and his sons ' with semi-Romanised usages and notions. What customary acres do we find left still on this classical ground ? (1) The Cornish acre of Dalriada has become, with an exaggerated furrow of 226 m., the prevalent Scottish acre to this day. (2) The strangely contrasted so-called Irish acre, with its area of two stades of very low standard stiU cHngs to the Yorkshire and Cumberland hills and extends in Lancashire as far as the Ribble. Where else do we find this curious proximity of these customary acres ? We have already seen that the Dalriada and Scottish acres are exaggerated and perhaps Romanised Cornish acres. But where shall we find the so-called Irish acre ? We have seen that it lies in the form of a square, but with a remarkable identity of area as a double stade acre in the land of the Pictones side by side with the Brittany arpent. Shall we find once more in the similarity of name and of the area of the Irish acre evidence of tribal movement of a portion of the Pictish tribe from the mouth of the Loire up the Irish Sea to the scene of the warfare which cost the Romans so much and made the region of the walls the limit of the Empire ? I dare not ofier a suggestion. It is enough to have mentioned the fact. British, and Armorican Acres. 267 In concluding this chapter there is, I think, a lesson to be learned, for which this remarkable Scottish acre might well be taken as the text. We have spoken many times of the substratum of common agricultural custom underlying these customary acres. But where does it lie ? It cannot be the result merely of hnes of commerce or intercourse passing over the surface of the common substratum. We have suggested that it must, so to speak, represent the stage of agricultural advance already reached in the course of evolution by the population of Western Europe, a common possession brought with them in their wanderings or learned from one another during ages of occupation of the country in which they are found permanently settled. But besides the common substratum of civilisation of Western Europe as a whole, there are distinct traditions belonging to particular tribes just as in the case of the Bamnonii, if the recurrence of the name in Ptolemy's map in Scotland as well as in West Wales may be trusted. Why did the tribe take the Cornish acre with them when an offshoot of them moved from the Cornish peninsula to found a new detachment of their tribe in Scotland ? All we can say is that the force of custom seems to have been strong enough in the case of the customary acres of Britain and Armorica to account for it. From these cases of frequent emigration of tribal offshoots, bearing with them their old tribal name and their customary acres we seem to learn the further lesson that above the substratum of common traditions in Central and North-Western Europe there was a tribal independence and restlessness which was a 268 British and Armorican Acres. feature of tribal conditions not to be overlooked. The fact that Csesar found a portion of the Helvetii on the move right across Gaul to the land of the Santones at the mouth of the Garonne is a striking case in point. It must confirm as well as throw Kght on Ptolemy's instances of emigration of offshoots of the tribes carrying their tribal name with them into new geographical positions. Whether this can have any bearing upon the problem of the remarkable similarity in the cus- tomary acres of the Veneti of Brittany and the Veneti of the Po Valley, I dare not venture an opinion. Finally, coming back to the intention adhered to throughout this inquiry to keep the facts regarding the acres distinct from inferences from them, let not the reader think for a moment that the substratum of common tradition so often alluded to was confined to the acres. It doubtless contained a thousand other elements with a long past behind them gathered up in the experience of hfe and the course of economic evolution. The study of the British and Armorican group of acres is but one hue of inquiry separately followed, because the time has not come yet for generahsations which become legitimate only when the facts of each separate line of inquiry have been sufficiently examined on their own evidence. So by the labours of many fellow-workers may we hope some day to understand better than we do what is involved in the toilsome path which humanity has had in the past and it would seem has still to tread towards the goal of civihsation— the art of Hving together in civiHsed society. INDEX. Aherbrothoc, Abbey of, 58 Acre, unit of cultivation, vii ; width of, as cricket pitch, vii ; length as range for long-bow practice, vii ; acre-strips in open field hus- bandry, 2, 118; North Wales, 15, 31, 108 ; acre-ends as dis- tances, 22 ; English statute acre, 30, 61, 103 s??., 258; South Wales, 33, 34; Irish, 110, 137, 242, 247 ; Cornish, 64-67, 104, 124 ; customary acres of Britain, Armorica, and Ireland, 59, 95- 116, 251 ; Dorset, Devon, and Somerset, 105, 135 ; Forest, 106 Powys, 107 ; Herefordshire, 108 Scottish, 61, 109, 110, 250 ; North umbrian and Cumbrian, 59, 110 Westmorland, 112 ; Northamp tonshire and West Derby, 113 Ulster, 110, 115,264; Brittany 124 ; Normandy, 132 ; stade acre 135, 199, 218, 239, 246 sg groups, 136 ; Prance and Channel Islands, 136, 138 ; Khenish, 139 Bavarian, 147-163 ; Venetian, 153 ; Po Valley, 158 ; Northern and Eastern Europe, 173 ; Sax- ony, 176 ; Baltic, 178 ; Low Countries, 181 ; Italy, 214 ; Spain, 233 Actus or agnua, 101, 118, 158, 215, 242 sqq. Ager, Cornish, 64 Agricola, 23 Agricultural Commission (Weights and Measures), 36, 59, 102, 106 Aillts or tenant-farmers, 16 Aircean, 119 Alexander, King of Scotland, 56 Almudes, 236 Altesjuck, 141 Ancillce, 40. See Cumal Andecena, 147-155 < Anglesea oyvar, 108, 109 ' Annals of the Four Masters,' 45 Antonine, Itinerary of, 88 Anzinga or andecena, 147 Arab measures, 203-6 Aranzada, 151, 238 Aratro, 222 Aripennis, 119, 125 Armorican acres, 124 sqq., 143-153, 251-268 ; furrow, 96 Armra, 98, 133, 198, 209-213, 257 Arpents, 95 sqq., 117-132, 142, 244-245, 254-260 Attic foot, 167 Augsburg jauchart and werkschuh, 149 Aussaat, 175-179 Avesta, 189 sqq. Babylonian agricultural units, 97 ; foot and palm, 192 BcBtica, 242 Baile, 45 sqq. Bais, 43 Baltic acres, 178-181 Barleyccn-n, 22, 28, 39, 42 Baronia, acre in, 59 sqq. Bat, 36, 114 Bavarian acre, 148 sqq. ; plough, 161 Belgic foot, 93 Bichette, 225 Bioka, 223 Bohemian measures, 175 Bonnier s, 184 270 Index. Bouvie, 137 Bovate, 58 note Brata, 242 Braza or eatado, 236 Brehon areas of tribute, 39, 60, 92, 240 ; tracts, 20, 39, 49 Brenhin or chieftain, 15, 16 Breton arpent, 95 sqq. ; 124-132, 260 ; ero, irvi, ervenn, 119 ; corde, 124, 125; furrow, 96, 128- 132; laz, 124; lieue, 126 sqq.; open field system, 117-123 Bretts and Soots, Laws of, 58 Brigg, Mr. J. J., on old Yorkshire milestones, 81 note Bucharest measures, 129 Burgo, acre in, 61 Byzantine donum, 175, 177 Caballaria, 240 CcBsar on British agriculture, 73, 261 Cahizada, 236 Cairo measures, 203 sqq. Calvados acres, 136, 252 sqq. Campo, 153-155 Can, 56 Cana de rey, 235 Canghellorship, 16 Canna, 205, 219 Canne, 229 sqq. Cantref, 15, 16, 38 ' Caption of Seisin,' 66 Carew documents, 39, 115 ; ' History of Cornwall,' 55, 66 Camac open fields, 121 ; arpent, 130 Carnarvonshire cyrar, 108 Carucates and Carruca, 19, 64-67, 159-161 Cassah, 203 sqq. Cattle in common herd, 18, 52 Celamin, 237-240 Celtic plough-team, 5, 75 ; units of tribute and the hide, 68-77 Co-aration, 17, 22, 39, 45, 75, 102, 120 Columella, 242 Comar, 119 Conveth, 57 Corde, 124 sqq., 236 Cornish acre, 64-67, 264 sqq., 104, 113, 124, 125, 132 ; gad or goad, 36, 105 sqq., 114, 124, 261 ; lace or square gad, 124 Corn-star, 151 Corvie, 137 Cotes-du-Nord, 130 Cow as unit of value, 40, 52, 68-77 Cuartal, 236 Cuartera, 234, 235 Cuchulainn Saga, 248 Ctierda, 236 Cuidichie or food-rent, 55 Cumal or female slave, 40, 68 Cumbrian measures, 23 sq., 59, 109 Cunedda, 23, 266 Cunningham acre, 110, 115, 264 Cymric traditional units of tribute and known reahties, 21, 68 Cymwd or commot, 16 Cyrus and the postal service, 188 Cyvar or co-ploughing, 5, 19, 107, 108, 119 Cyvelin or goad, 36 Damnonii, 264-267 Danish and Norwegian measures, 169 Danuhian acre, 143-157; itinerary measures, 128-129, 168-169, 173 Davach, Scottish ploughing area, 60 David, King, of Scotland, 24, 25, 58, 59, 79, 104 Decempeda, 25, 97, 114, 147 Deer, Book of, 54 Denbigh Extent, 17, 20 Derbyshire (West) acre, 106, 113, 137 Dessiatina, 177 Devon acre, 105 Devonshire, Extent of Beri, 65 Diagonal of square, some measures related to, 96, 257, 261, et passim Diamed and the boar, 24 Diaulos, 194, 196 Dimetian Code, 16; erw, 34, 112; tref, and tunc pound, 35-38, 71 Domesday Survey, 64, 70, 83 Ddrpfeld, Dr., and Sunium, 97 Dorset acre, 105 ; gad, 105, 106 ; stade acre, 135, 136 Dovraeth, 16, 57 Druidical and Roman Catholic customs, 122 Dunwal, 'the great measurer,' 15, 21, 25-33, 35-37, 69, 103 Dutch acres, 182-184 Index. 271 Earls, Seven, of Scotland, 56 Eglyshaw, 36, 114 Egyptian khet, 98, 127-132, 137; land units, 203-206 Eminate, 223 Eminee, 225 English mile, old, 93, 133 ; statute acre, 61, 103-106, 111-114, 133 Ero of Brittany, 119 Erw or acre, 15, 27-33, 79, 108, 118 Estadal, 233, 237 sqq. Estadio, 242 Exon Survey, 64-66 Faltasce, 135, 153, 173 Familia or tribal group, 76 Fanega or fanegada, 236-243 Farthing land, 66 Feddan, 203 sqq. Feldacker, 140 Feldmc/rgen, 139 Fddruthe, 140 Female slave as unit of value, 39^1 Ferling, 65 Fineshade Priory, acre, 113 Finistire, 118, 120, 130 Finn, thumb of, 24 Flintshire ' quarts,' 107 Food-rent, Welsh, 15 sqq. ; Irish, 39 sqq., 45-52 ; Scottish, 54 sqq. Foot, 22-27, 79-93, 97, 114; Scottish, 24, 25, 79, 104 ; Welsh, 25, 26 ; Irish, 39, 43 ; Roman, 26, 83, 114; English, 82, 93 ; Belgic, 93; Greek, 98; French Royal standard, 125 ; Calvados feet, 253 ; natural foot, limits of, 25, 98 Forest acres, 106, 113, 256-264 Fostre or food-rent, 67, 76 Fotheringhay acre, 113 French heues, 127 ; Royal standard foot, 125 ; acres, 130-137, 142 ; open field husbandry, 143-146 Furlongs, 82 sqq. Furrow, 29-31, 96, 100 map, 101- 118, 128-133 Gad or goad, 36, 105 sqq., 124-132 Oafolyrth, 145-147 Oavael, 28 Oeira, 242 German acres, 139 sqq., 165 sqq. Giomata, 224 Oiraldus Cambrensis, 18 Glamorganshire plough-lands, 108 Goad, 29, 36, 105 sqq., 124-132 Gosgordd or retinue of chieftain, 16 Grabe, 150-152 Greek measures, 22, 25 ; stade, 26 ; modius, 29, 79, 93, 99, 133, 167, 173 ; socarion, 99 ; medimnus, 98, 127-137, 173; aroura, 98, 132, 133 ; plethron, 98 ; foot, 167 ; colonisation in Magna Greecia and Sicily, 207-213 Guernsey acre, bouvee, and corvee, 137 Gues of Homer, 194 sqq. Gwelegordd or family group, 16 Gwely or family group, 18 Gwentian Code, 16, 33 ; erw, 33, 112 ; tref and tunc pound, 35, 36, 71 sqq. ; plough-land, 72 Gwestva or food-rent, 15 sqq., 20, 27, 34, 70 ' Harleston Estate Book,' 113 Harrison's ' Description of Britain,' 83 Hathra or hasar, 190 Henry VIII. and Leland's survey, 90 sqq. Herefordshire acre, 108 Herodotus, 128, 188-192 Hides and carucates, 19, 64-77' Hochdcker or high-backed lands, 162 HoUnshed's ' Chronicle,' 83 Homeric ploughing, 193-202 Howell, Laws of, 15, 20, 21, 27, 35, 37, 71, 72, 108 Hyginus on Belgic foot, 93 Ine, King, 67, 76 Innes, Cosmo, 55 sqq. Innsbruck measures, 150-153 Irish or Plantation acre, 110, 112, 137, 247 sqq. Italian agricultural units, 214-227 Itineraries of Britain, Antonine's, 88 ; Harrison's, 87 sqq. ; Leland's, 90 sqq. ; of William of Worcester, 92 ; Ogilby, 84 ; Roman, 139 Itinerary measure, 15, 22, 25 sqq., 35, 167-172, 186-192 272 Index. J assy measures, 129, 135, 153, 173 Jauch, 150, 151 Jauchart or juchart, 149-152, 184 Jersey acre, 136, 137 Joch, 174 Jour or quarter-acre, 139 Journal or arpent, 118, 131, 225, 247 Juck of Kalenberg, 178, 179 Jugerum, 98, 101, 110, 115, 118, 133, 158, 214 aqq., 258 Jugon, 198 sqg. Kdbellangde, 169 Kehrpflug or swing plough, 161 Kermario, 121 Ker3 or hamlets, 118-123 Kever or kefer, 119 Kilda, St., 55 Lace, 124 Lanatz of Belgrade, 174 Land (tyr, tir, or grwn) or land-end, 22, 27, 79, 103, 240 Land measures of Welsh Codes, 20 sqg. ; milltyr, 21, 22 ; barley- corn, thumb, palm, foot, pace, leap, land (tyr or grwn), land- ends, or acres, 22, 27 La Tine civihsation, 157, 248 Laz, 124 League, 82 sqq. Leap, 22, 27 Lega, real, 129, 240 Leland's ' Commentaries of Britain,' 83, 90 sqq. LeuccB, 82 Leuga, Gallic, 25, 26, 42, 47, 67, 68, 77 sqq., 93, 104, 138-163 Libere ienentes, 65 Lieue de Bretagne, 126 sqq. ; French, 127 Lismore, Book of the Dean of, 24 Llath, 36 Loire and Garonne, land units between, 246-250 Low Countries, acres of the, 182-185 Moat, 184 Mactigems, 9, 10 Maenol, unit of food-rent, 15, 16 19, 21, 27-38, 71, 72 Maership, 16 Magna Orcecia, Greek colonies in, 207-213 Mairtis, 55 MaiOand (Prof. F. W.), 'Domes- day Book and Beyond,' 70, 73 Mai or tribute, 15, 35, 37, 69 Malcolm, King, 56 Man's thumb, foot, and step, 22- 27, 79-93 Mappa, 146 Martini (Angelo), 128, 139 sqq. Medimnus, 98, 128-136, 208-213,254 Meer or meering, nine steps of the, 23, 29, 32 Migou or open field, 120 Meile of Tyrol, 152 Meiizen, Prof., 101, 139, 166 MerMands, 64 sqq. Migliaio di ghebbi and di passi, 165 Mil, 129 Mile, Roman, 26, 26, 48, 87-89, 97, 170; Welsh, 15, 22, 25, 27, 30 sqq., 79 ; old British, 79-93, 104 ; English statute, 79, 88, 93 ; old English, 93 ; old London, 93 ; Scottish, 69, 110 ; Modena, 165 ; Venetian, 156 ; Courland, 172 ; Sweden and Ejnland, 172 Milestones, old, 80 sqq. Milk in common chum, 18 Milliaria, 92 Milltyr, great itinerary measure, 15, 21-27 ; square, 30-37, 51, 79 sqq. ; 97, 103, 240 Modena mUe, 165, 227 Modius, 29, 99, 231, 238 Moggio, 223 Mojada, 2Z4i: Moldavian measures, 129 Montgomeryshire stang, 108 Morhihan arpent, 130 Morgen, 140, 178, 184 Munster acres, 115 Natural measures, 22-27, 79-93, 97. 98, 114, 115. See Foot Normandy acre, 132-134, 264 sqq. ; furrow, 133 ; verge, 132-134 North Wales stang or cyvar, 108 Northamptonshire halt-acre, 106 NorthunArian measures, 23 sq., 59, 112-114 ; acre, 109-114, 137 Ogilby's survey, 84 sqq. Old Extent, 54-60 Index. 273 Oldenburg, 141 OUamh Fodla, King, 45 Open-fidd agriculture, 1 sqq., 18 sqq., 117 sqq., 145 Ordlach or orJaxih, 39, 42 Orgrj/ia, 191, 195 Ouvrie, 225 Oxgang, 59 sg?. Pace, 22-27, 79-88 Paladyr, 36, 108, 109 PaZnj, 22, 28, 39, 93 Pan of Provence, 229 sqq. Parasangs, 25, 165-169, 186-192 Pembrokeshire plough-land, 108 Penmarch, open fields at, 118, 120 Perch, the King's, 82 sqq. Perche, 253-4 Pertica, 147, 223 Petrie (Prof.) on weights and measures, 93 Phoenician measures, 200 Pictish acre, 248 Pictones (Pictavi), 246 sqq. Planaratum, 159 Plantation acre, Irish, 110, 112, 137, 247 sqq. Plethron, 98, 194, 196 Plinth of the Ptolemies, 209 Pliny on ploughs, 159-161 Ploughing in Homeric times, 193- 202 Plough-lands, Irish, 39, 45 sqq., 70 ; Scottish, 55-63, 69, 73 ; Cornish, 65 ; Welsh, 108, 109 Plough-team, 28, 69-77, 158-163 Plouhamel, open-field system at, 121 Posla, 129 Pound-paying units, 70 Pounds of auld extent, 54 sq. Predjine, 135 Price, Sir John, 37 Po Valley acres, 143-157 Podismus, 197 Polish measures, 176 Porca, 243 Portuguese measures, 242 Powys acre, 107, 108 ; furrow, 107 Prussian mile, 169 ; joch, 174 Ptolemies, measures of the, 127, 208 Pyramids, measurement of the, 191, 204 Quadriga or quadrijuga, 159 QuarantencB, 82 sqq., 92 Quarte, 220 Quarts of land, 107 Raie or furrow, 130, 131 Randir, 28 Bhcetian plough, 159, 161 Rheims, Abbey of, acres of, 145 Rhenish acres, 138-143 Rhys (Sir John) on Welsh units of tribute, 23 Ridgeway (Professor), 119, 194 Roman actus, 118, 215 ; decem- peda, 25, 97, 114; mile, 25, 48, 87-89, 170 ; jugerum, 98, 101, 110, 115, 118, 133, 158, 214 sqq. ; stade, 25, 26 ; vorsus, 99, 214^ 232 ; sors, 110, 133, 249, 250 Round, Mi., 64 Rubbio, 220 Rumanian measures, 129 sqq. Russian units, 179 ; werst, 170 ; dessiatina, 177 Ruthe, 140, 178 Salma, 205, 211-213 Satia or saton, 198 sqq. Satiba, 199-201 Saxon ceorl's holding, 76 ; hide or hiwisc, 76 Schoinos, 209 Scone, Monastery of, 56 sqq. Scotland, Acts of Parliament of, 25, 58 ; islands of, 54 sqq. Scottish acre, 109, 110, 250, 263, 264 Scottish Kings, Laws of, 58 Scottish mile, 109 Seine-Infirieure, grouped acres in, 136 sqq. Skerke, 225, 230 Setier, 134, 259 Seytive, 225 Sherwood Forest furrow, 106 Sicily, Greek colonies in, 207-213 Sillons or high-backed strips, 118, 130 Skene (Mr.), 24, 55 sqq. Soma, 218, 219 Somerset acre, 105 Sors, 110, 113, 249, 250 Spanish agricultural units, 233-245 274 Index. Stade, 26, 44, 48, 79, 82, 107, 135, 165, 167, 189, 199, 207, 218, 239, 246 Stade acre, 135, 165, 199, 218, 239, 246 sqq. Stang and Siangel, 108 Starata, 223 Starkmd, 150, 153 Statute mile, origin of, 82, 88, 93 Stingeni, 135, 231 Stochiacah, 150 Stunde or parasang, 165-169, 175 Strich Aussaat, 175 Sulung or swulung, 71, 73 Sunium, temple of, 25 plate, 97 Swiss acres, 141 sqq. Taeogs or servile tenants, 16, 17 Tagewerke, 140 Tagmat, 150, 151 Tervyn, 23 ' Testa de Nevill,' 64 sqq. Tetraguon of Homer, 195 Thumh, 22, 28, 253 Tir, tyr, land, or land-end, 22 Tir-oumail, 41 sqq., 60 sqq., 68 sqq., 92, 104, 111-116, 240, 241 Toise, 253 Tomolo, 206, 210-213 Tornatura, 224 Trabucchi, 224 Trc/ or irev, unit of tribute, 15 sqq., 28 ; .area occupied by a pastoral group, 17 ; summer and winter trefs, 18 ; alternately pasture and plough-land, 19 ; Irish treab, 20 ; changes in meaning, 20 ; Dimetian tref, 34, 38, 71 ; Gwentian tref, 35, 37, 71 ; division of square milltyr, 37 ; traditional size of trefs compared with actual area of Wales, 37 ; Venedotian tref, 38 Trefgordd, payers of food-rent, 16, 35 ; traditional meanitig of, 17, 46, 70 Trent pio, 162 Trichaced, 45, 51 Tunc found, unit of tribute, 16, 16, 33-36, 52, 70 Tyddyns or homesteads, 16, 28 Tyrol meile, 162 Uchelwyrs or free tribesmen, 16 Ulster acre, 110, 116, 248, 264 Vaine p&ture, 3, 121, 144 Valvasone campo or acre, 152 Vara, 236, 248 sqq. Venedotian Code, 6, 15 sqq., 19, 21, 23, 27, 80, 93, 163 ; erw, 32 ; maenol, 27, 35-38 ; milltyr, 36, 51, 69, 79 ; tref, 27, 38, 79 Feraeiiara acres, 153-157 ; mile, 166 Verge, 132-134 Versura, 220 Vienna joch, 174 Virgate, 66, 67 Vorsus or versus, 99, 214-232 Waldacher, 140 Waldmorgen, 139 Wales, actual area of, 37 Walter of Jtienley's league, 82 sqq. Wedders or widderes, 55 Weehwork on lord's demesne, 76, 145 Welsli, Codes : Venedotian, 15, 21, 23, 27, 30, 35, 126; Gwentian and Dimetian, 16 Werkschuh or Augsburg foot, 149 Werst, 170-172 Westmorland acre, 112 Wheel in mensuration, 84 Wheeled ploughs, 157 sqq. Worcester, evidence of old British mile, 91 Xenophon's stade or day's march, 189 Yardland, 4, 5, 66, 72 Yoke, 28, 32 sq., 114, 159 sqq. Yorkshire milestones, 80, 88 Yugada, 237-244 PBINTED BY SPOTTISWOODE AND 00. LTD., COLCHESTER LONDON AND ETON