previous next


AGRIMETA´TIO In the origin of a community land has been acquired either by occupying what is vacant, or by conquest from a previous occupier. The community from time to time regulates how it shall be held by its members, whether in common or by separate families or by individuals, and what transfers from one to another shall be allowed or protected or enforced (cf. Cic. Off. 1.7, 21). For such purposes land has to be defined, definition requires boundaries to be declared, and naturally brings with it measurement and land surveying in general.

1. The Roman land-surveyors (whose writings will be here referred to by the pages of Lachmann's edit.) make three great classes (qualitates) of land: one divisus et adsignatus, another mensura per extremitatem comprehensus, a third arcifinius qui nulla mensura continetur (p. 1), i. e. land divided and marked out, land surrounded by a measured boundary, and land not measured at all.

The word arcifinius was derived by Varro from arcere, and he has been followed by the Roman surveyors and by (apparently) all modern writers, Varro understanding hostes, others vicinum, others fines (pp. 6, 137, 284, 369; and Rudorff; Grom. Inst. p. 251). This derivation is certainly wrong, for it is not conformable to the laws of Latin composition, and another derivation is almost demonstrably right. Proper components of such a word are arcus and finis. A bow as seen in ancient figures was composed of two horns joined by a straight piece, each horn having a double curve, and the bow was thus a type of an irregularly curved or wavy line. Now see Balbus's words (98): Extremitatium genera sunt duo, unum quod per rigorem observatur, alterum quod per flexus: rigor est quidquid inter duo signa veluti in modum lineae rectum perspicitur: per flexus, quidquid secundum locorum naturam curvatur, ut in agris arcifiniis solet . . . Quidquid in agro mensorii operis causa ad finem rectum fuerit, rigor appellatur . . . Flexuosa linen (distinguished from straight and curved lines) est multiformis, velut artorum aut jugorum aut fluminum; in quorum similitudinem et arcifiniorum agrorum extremitas finitur. See also pp. 12, 31, and the figure in 312 referring to p. 342. Ager arcifinius is therefore land “bow-bounded,” i. e. bounded with a natural wavy line, as opposed to land bounded by the artificial straight lines of the surveyor.

2. If a conquered territory or part of it was to be assigned to Roman citizens, as, for instance, in founding a colony, the land was divided and marked off in plots according to a regular plan. The surveyor placed himself facing due west, and drew one line due east and west, and another crossing at right angles, i. e. due north and south. Along these lines were set out two wide balks (limites). The east and west balk was called decumanus; the north and south balk was called kardo. All balks parallel to the decumanus were prorsi limites, all parallel to the kardo were transversi, but were often all called decumani and kardines respectively, the two principals being distinguished as decumanus maximus and kardo maximus. The others were numbered from the centre crossing-point, all the north and south balks in front of the kardo maximus being ultra kardinem max., all behind [p. 1.85]it being citra kardinem max.; and all the east and west balks being dextrā decumanum max. if on the surveyor's right of the decumanus maximus, and those being sinistrā decumanum max. which were on the surveyor's left of the same. The combination of these terms gave a ready description of any balk or block of land. The half in front of the surveyor, i. e. the western, was called also pars antica; the half behind, i. e. the eastern, pars postica (pp. 27, 303). The halves are also spoken of as ager dextratus, sinistratus, citratus, ultratus (or pars dextrata, &c.), being respectively, according to the normal arrangement of the balks, the northern and southern, the eastern and western halves; and the quarters being denoted by a combination, e.g. ager ultratus et dextratus for the north-western quarter, &c. (pp. 247, 290 sqq.).

The measurement was by a ten-foot rod, pertica decempeda, either substantive (pertica) or adjective (decempeda) being often used alone to denote it (e. g. Prop. 4.1, 130; Hor. Od. 2.15, 14). Twelve rods or 120 feet was an actus (i. e. one ploughing length, Plin. Nat. 18.9; Col. 2.2, 27). A square actus contained 14,400 square feet, and a jugerum was an oblong rectangle, composed of two square actus, containing 28,800 square feet, being 24 rods long by 12 rods broad. The amount originally assigned to each citizen was 2 jugera (about equal to 1 1/4 acres), together forming a square plot called heredium ( “ quod heredem sequeretur; ” comp. “hereditament” ). A hundred of these plots, forming a square of 20 actus or 240 rods each way, was a centuria (Varr. R. R. 1.10; Grom. pp. 95, 153). The century was the normal unit of division of the colony, as the jugerum was the normal unit of the centuria. Between the centuries ran the limites, those forming the principal divisions being broader than the others. The balk along the line of the meridian was called kardo, from being as it were the hinge or axis on which the heavens revolved; that crossing it was called decumanus, dividing the realm of light (south) from the realm of darkness (north). The word has been the subject of many bad etymologies (Rudorff, Grom. Inst. p. 343). But it is certainly the ordinary adjective, meaning the balk “of the tenth.” Each side of the century contained 10 heredia, and the balk bounding the century was therefore a balk “of (i. e. adjoining) the tenth” heredium. When kardo was taken for the name of the north and south lines, decumanus was confined to the east and west balks, which would come at every tenth heredium, measuring along a kardo (Nissen, Templum, p. 12, gives a more artificial explanation). Every fifth balk, counting from but not inclusive of the decumanus or kardo, was called limes quintarius: the others were called in Italy subruncivi ( “slightly cleared” ), and were used as occupation roads for the farmers to carry off their produce. The principal balks were also called actuarii, “carriage roads” (Dig. 8, 3, 1), and were usually public roads of considerable breadth (Grom. pp. 120, 175: compare the metaphor in Verg. A. 9.323). Thus in the colonies founded by the Gracchi, by Sulla and by Augustus, the decumanus max. was to be 40 feet broad, the kardo max. 20 feet, the other actuarii (i. e. the quintarii) to be 12 feet, whilst the subruncivi were to be 8 feet. This last breadth was that required by the XII. Tables for a private road (Fest. s. v. viae: cp. Bruns, p. 299; Dig. 8, 3, 18). What in Italy were called subruncivi were in some parts of the world mere lines of division, and hence called linearii (Grom. pp. 168, 169; cf. p. 212).

3. The centuries were marked by a stone at each corner. The law of the Augustan colonies, which seems to have been regarded as a kind of model for others, also prescribed that these corner stones should be of flint or millstone or the like, polished, round, a foot in diameter and four feet in length, of which 2 1/2 feet should be sunk in the ground (p. 194). Hyginus recommends the following system of inscriptions. The stones on the chief decuman and kardo to be inscribed on the front; the rest on the sides. Beginning with these two principal balks, each stone along them was to be marked with the notes of the decuman and kardo adjacent. Thus, at the corner where the decumanus maximus was crossed by the second kardo, the stone would be marked on the top with D M or D I for decumanus maximus, otherwise called primus, and K II for kardo secundus, i.e. the next kardo to the kardo maximus. The inscriptions were generally so put as to / occupy only three-quarters of the surface. The first four centuries to be marked were those immediately adjacent at each angle of the central crossing point. Each of these centuries would


have the decumanus maximus on one side and the kardo maximus on an adjoining side, and three corners would thus be on these principal balks. The fourth corner was called the angulus clusaris, “closing corner;” it was diagonally opposite to the centre crossing; and on this was inscribed the full technical description of the century. Thus, the north-west corner of the north-west century was marked with a stone inscribed D D I V K I, i.e. dextra decumanum primum, ultra kardinem primum, being the century on the surveyor's right hand and in front. The south-west corner of the south-west century was marked S D I V K I, i.e. sinistra dec., &c.; the south-east corner was marked S D I K K I, i.e. sinistra dec. primum, kitra kardinem primum, being the century behind the surveyor on his left. The north-east century was marked D D I K K I. The next century east, and on the right of the decumanus maximus, would be beyond the kardo secundus; and its name would be on the stone at the north-east corner, i. e. the corner furthest from the principal crossing. These names of the centuries were put not on the tops but on the sides of the stones, because they would often be of considerable length, e. g. D D LXXXXVIII V K LXXV. (p. 173), i. e. a remote century in the north-west region on the right of the 98th decuman and beyond the 75th kardo. The inscriptions had to be written on the stone downwards (p. 195; see however p. 111). This systematic marking of the centuries by the number of the adjacent decuman and kardo, the numbers all starting from the centre crossing, enabled a surveyor on examining a series of these stones to say in what region of the district he was, and by comparison with the heavens, [p. 1.86]whether the decumanus and the kardo had their normal directions respectively of east and west and of north and south (p. 290). Sometimes stones were put only on the main balks and oaken posts on the minor ones (p. 112). Augustus directed all the century marks to be of stone, and the marks put to distinguish one man's allotment from another's to be oak (p. 172).

4. The earliest allotments were two jugera (Liv. 4.47; Grom. p. 153); later we read of 2 1/2 jugera (Liv. 6.161), 2 3/4 and 3 (ib. 8.11); of 37/12 (ib. 5.24), of 5 (ib. 39.55), of 6 (ib. 44), of 7 (ib. 5.30), which last was often regarded as a standard amount (Val. M. 4.4, 11; Plin. Nat. 18.18), though larger amounts are also mentioned, e. g. 8 and 10 (Liv. 29.55), even in one case of 51 1/2 jugera (ib. 41.13); and Hyginus (p. 199) speaks of one-third of a century (=66 2/3 jugera) as if it were not unusual. The quantity assigned as one lot was not necessarily continuous: e. g. of the 66 2/3 jugera, 6 2/3 might be in one century, 15 in another, and 45 in another (p. 204). The quantity depended primarily on the extent of the divisible land, and the number of the recipients; but the rule of equal distribution was not always followed. In the assignments to soldiers army rank was regarded, a common soldier receiving one lot, and higher ranks 1 1/2 or 2 lots, and even in republican times at Frentinum 20 jugera were given to each foot-soldier, 40 to each horse-soldier. In the Latin colony of Bononia the amounts were respectively 50 and 70 jugera (Liv. 35.9; 37.57). At Aquileia the foot-soldiers had 50 jugera, the centurions 100, the cavalry 140 each (Liv. 40.34). The quality of the land was also regarded, its goodness being sometimes compensated by a reduction of the quantity (pp. 156, 159; cf. Suet. Aug. 49).1 In the Augustan rules (no doubt following older precedents) the measurement and assignment of the land was to be carried “where the bill-hook and plough had gone” (qua falx et arator ierit, pp. 112, 201), i. e. so far as it had been cleared and ploughed (see Cicero's remarks, Rull. 2.25.67). Hyginus held that this did not necessarily confine it to land actually cultivated, but that some woodland or pasture might usefully form part though not the whole of an allotment (p. 203).

The straight lines of the Roman limitation were carried out regardless, so far as possible, of natural obstructions. Hence a century would often include part of a river, and thus water as well as land fall into an allotment. This was indeed necessary where the total territory was small, and convenience often compensated the person to whom it was assigned. But with large rivers like the Po, spreading from winter torrents far beyond their banks, it was found better to except from this distribution not only the bed of the river, but a considerable margin along it also (pp. 51, 120, 157). In the same way the balks were sometimes included in the centuries, i. e. half the balk in each adjacent century. Sometimes where the balks were broad, these (and often public highways also) were excepted from the distribution (pp. 120, 158). But, whether excepted or not, there was a right of road along them either for the public, or at least for the neighbours. In some cases, as the land distributed had been previously occupied by the conquered people, the rigorous line of the balks would pass through a farm-house (villa), and the owner then put gates and a porter to let the people through. But if he did not mind giving up more land, he might apparently deflect the public right of way, provided the circuitous road was not inconvenient (pp. 121, 158, 159).

Considerable respect in some ways was shown to the former proprietors and their arrangements. Any sacred places, groves, temples, tombs, public waterworks, public drains (fossae), or drains between neighbours, roads, &c., were allowed to remain subject to the same rights and use as before (pp. 120, 157). The existing proprietors sometimes, from their rank or influence or friendship (the case of Virgil is known), were allowed to retain their own lands, or else, if a man's lands were scattered, a larger continuous area was given him in one place to make up for the separated parts which were taken for distribution (pp. 155, 117, 130). A whole town indeed might be locally within the boundaries of the limitation, but be exempted and left in its former condition (p. 118; cf. Cic. Fam. 13.4). Barren parts, woods, &c., were often not assigned, but either reserved for subsequent disposal or granted (concessa) to the colony as a whole or to individuals, frequently to the neighbouring proprietors, to be held as common land for pasture (ager compascuus, pp. 196, 201). Such places would rarely or never coincide exactly with the area of a century, and the remainder of that century was an oddment (subsecivum). Other oddments were made at the extremity of the limitation, the ground not allowing of complete rectangular centuries (pp. 7, 155). In the colony of Emerita, in Lusitania, a very large district was surveyed and marked out, but not fully assigned. A number of veterans were settled at the extreme border, and others at the colony itself and near the river Anas. The intervening district received two additions of colonists subsequently, but there was land yet left, which was utilised apparently by such of the colonists as chose. This land also came under the class of subseciva (pp. 51, 201). But if land within the extreme bounds (fines) of the district was not wholly divided by limites--as, for instance, if it had not been cultivated and was therefore unfit for assignment--it was called abandoned and shut out (relictus et extra clusus, p. 55) from the regularly assigned district. In such a case Hyginus recommended that the bounds should be surveyed and fixed by a system of right angles, and boundary stones be erected. In certain places stone altars should bear on one side the name of the colony, on the other the name of the neighbours abutting (adfines); where the bounds made an angle, the altars should be triangular. If the boundary was rock, marks and (if possible) inscriptions should be put upon the rocks themselves in lieu of stones (p. 198).

5. As a general principle, land was divided by balks in order to be marked out to persons (ad signatus) as their property; but division [p. 1.87]was one thing and assignment was another (p. 54). After the survey was completed, sometimes it was found that there was more land than was wanted for assignment, and that some of it, being barren, was not suitable for assignment. On the other hand, the land might prove not enough for all the intended assignees, and then occasionally other land described, but not divided by balks, was assigned to additional citizens outside of the regular limits (p. 160). And sometimes part of a century, or even two or more complete centuries (lati fundi), were restored to their old proprietors on their declaring the quantity of their former holdings (pp. 156, 160). Neither the land assigned nor the land restored was subject to a ground rent (vectigal). But such a rent was imposed on land which was not assigned in private ownership, and was evidence of its remaining public. Thus the oddments and the land, in excess of that required for the number of colonists, were (if not left to chance) subjected to a rent payable by the occupiers to the owner, i. e. either to the Roman State, or, if the oddments had been granted out to the colony (concessa), or restored to the borough, then to the colony or borough chest (p. 117). The lands of the Vestal Virgins or of priests were also subject to a ground-rent (p. 117). In other words, the lands not assigned were usually let for a year, or five years (lustrum), or a hundred years. The actual occupiers were often the adjoining landowners, but they did not always hold directly from the lord, but only as sub-lessees of a manceps, the state generally in this, as in other matters, availing itself of the agency of a large contractor (manceps), who would give proper security both in person and specified estates (praedes ac praedia) for the due payment of the ground-rent. When the lease was for 100 years, the words buy and sell (emptio, venditio) seem to have been thought by the surveyors more appropriate than take and let (conductio, locatio), though Gaius gives the other as the better legal view (3.145). See under AGRARIAE LEGES

6. The colonies in the early days of Rome were Roman garrisons and outposts in a foreign, and, it might be, still hostile country (Cic. Font. 3, § 5; Agr. 2.27.73), garrisons to watch the boroughs, outposts to repel the enemy on the border (Grom. p. 135). In the days of Augustus and his successors they were settlements of veterans in strangers' territories. Alike in both cases they were isolated intrusions, and while they had jurisdiction within their own bounds they had none outside of them. Nor was their jurisdiction co-extensive with their bounds. Only that which was given and marked out (datum adsignatum) was subject to their control. Such of the former owners as were allowed to retain their own lands within the bounds of the colony retained as a rule their allegiance to the neighbouring borough or other community to which they belonged before; and the same was sometimes the case with any oddments (subseciva) or vacant centuries (pp. 117, 163). Occasionally the former boroughs were deprived of all jurisdiction over land outside of the borough walls (p. 164). In the Augustan colonies some specially deserving veterans had their estates (fundi) exempted (excepti) from all duty to they colony, and deemed to be part of Roman soil. Other veterans had larger estates granted than the terms of the foundation allowed. These estates were therefore specially marked as concessi (p. 197).

The whole ground assigned to a colony was called a rod (pertica, p. 26), and was divided on one system and from one central crossing. If the land originally given up to the colony proved insufficient for the number of colonists, some adjoining territory was taken and surveyed, but on a separate plan, and with its own decumanus maximus and kardo. Such an adjunct to the colony was called a praefectura, probably as being under a praefect sent from the colony (p. 160). Sometimes the land of one borough was seen from the first not to be enough for the intended colonists, and the lands of several boroughs were taken and thrown together to form one pertica (p. 164). The case of Mantuan lands being taken to make up what was wanted for the veterans established at Cremona is well known from Virgil and his commentators (ad Ecl. 9.27; Georg. 2.198).

7. The distribution of the plots among the individuals was determined by lot. The individuals were grouped according to the number of allotments contained in a century. If 20 jugera were allotted to each veteran, groups of ten (decuria) were made; if the allotment (sors or accepta, i.e. sors: cf. Ov. Am. 2.9, 19; Cic. Fam. 11.2. 0) was a third of a century, the groups would contain three (conternationes). Slips bearing the names of all the sharers were put into the urn, and the first ten (or three) that came out of the urn formed the first group, and that group took according as their respective names came out. Then lots, inscribed with the exact description of ten (or three) plots, were put into the urn, and the first that came out was for group 1, the next for group 2, and so on. The description was thus--“ Sors prima: D D I et secundum et III. et IV. citra kardinem illum, ” i. e. “First lot: on the right of the 1st decuman and 2nd, and 3rd, &c., on the hither side of such and such a kardo.” Sometimes the composition of a group was settled by private agreement, those in the same group having presumably neighbouring plots (pp. 113, 200).

8. A chart (pertica, typus, forma, centuriatio, &c., p. 154) was made of the whole pertica, showing its bounds and abuttals, all the balks, the position and limits of the oddments, common pastures and woods; plots assigned to individuals in fee (data adsignata); plots granted (concessa), e.g. as corporate property or under special circumstances; plots specially exempted from the colonial jurisdiction (excepta); plots given back to the ancient occupier (reddita veteri possessori), which sometimes were of large extent, and filled several centuries (hence specially registered redditum suum lati fundi, p. 157); plots given to an ancient occupier in exchange for his own (reddita commutata pro suo); the district assigned to a river, sometimes quaintly described as a veteran or ancient occupier (datum adsignatum ut veterano, or redditum suum veteri possessori flumini Pisauro, tantum in quo alveus, p. 157). Initials were often used to denote these frequently occurring categories: for instance, C V P for concessa veteri possessori (p. 203). When part of an allotment consisted of woodland adjoining, or mountain pasture at a [p. 1.88]distance (ultra quartum forte vicinum, pp. 204, 226, 15), this was joined to it on the chart by a wavy line (commalleolari2 debet, p. 204 ; and see figs. 196, 197). Besides the chart, there was a register of the allotments, grants, declarations (professiones) of the ancient owners, &c., setting forth the local description of each century, the names of the persons to whom it was assigned or restored, and the quantity of acres to be held by each. The precise appropriation of the century among those who drew it in common (consortes) was not entered in the register, but left to be commemorated by the sharers by erecting suitable boundary marks (pp. 211, 271, and cf. p. 162). The chart and register ought to be made on bronze and in duplicate, and one chart erected in the centre of the colony, the other kept in the imperial record office (Caesaris tabularium). These were the most authentic evidence in all disputes about the boundaries, the quantity and title of the lands of the colony or other assignment (pp. 154, 202). They were not always made with the same particularity. One surveyor (Balbus?), in forming the plan and register for assignments by Trajan in Pannonia, not only noted the quantity of each allotment, but drew a line round it, and inscribed the length and breadth, thus defining the extent and situation of the individual properties, and of the oddments; others contented themselves with stating the quantity assigned, only when the centuries were not full (p. 121). The chart and register contained, besides the original assignment, any subsequent assignments which might be made of the same measured district. Thus some of Julius Caesar's colonists returned to military service; and when they went back to their own lands, the plots of those who had died were assigned to others. Hence, the register often contained, under the same century, the names of the original holders, and of their successors, no note apparently being made of the cause (p. 162). Subsequent assignments of unallotted plots and grants of reserved oddments, &c., would appear in the register; but changes in their holdings, made by the settlers themselves, by purchase, or lease, or exchange, or occupation of vacant parts, were not entered. New acquisitions being often made according to mutual convenience, and without regard to the bounds set out in the survey, the register ceased to be decisive of disputes between them or their successors, and recourse had to be made to the private deeds (cautiones) which might describe the quantities and the marks (p. 131). In the case of a veteran dividing his allotment among his sons, there appears to have been a practice of erecting boundary stones in lines, having a common and definite relation to the boundary stone on the principal balk. Such stones were called comportionales.

9. The account given above relates to what was considered the normal survey and arrangements. The decumanus maximus running east and west ; the kardo maximus running north and south ; the surveyor facing to the west, so that the west was ultra kardinem and the north dextra decumanum; the centuries rectangular and equilateral, containing each 200 jugera, and each side being 20 actus long; the balks duly proportioned in width to their importance as the main, secondary or smaller patent lines of division, strictly parallel to the chief decuman or kardo, and marked by significant boundary posts of proper character, height, and thickness, and duly inscribed with their number and description--such was the standard of a Roman surveyor for the distribution of land among Roman colonists. But in almost all these points differences were found. Sometimes the decuman was drawn, not due east and west, but in accordance with the visible quarter of the rising of the sun on the foundation of the colony (p. 170); sometimes it was made to run along a great road, as the Via Appia (p. 179); or in the direction of the greatest length of the territory; or even, as in the neighbourhood of Capua, north and south, with the kardo east and west (pp. 29, 168). Where a colony was close to the sea, the balks were made parallel to the shore, and called maritimi, and the cross balks parallel to the mountains, and called montani (p. 168).

The most perfect arrangement was for the centre of the limitation to be at the centre of the town, and the decuman and kardo to pass out by gates at each end of the four sides; but often this was not possible or convenient, mountains or the municipium whose territory was now taken for distribution being in the close neighbourhood, and then the centre crossing was put in some suitable place outside (pp. 178 sq., 194). A praefecture was usually divided on a system independent of that of the pertica to which it belonged (pp. 160, 171). The Nolan district was, from the first, surveyed with two different sets of balks, the one set adjoining the other, but meeting them somewhat obliquely. The two sets of decumans were distinguished as dextra and dexterior, sinistra and sinisterior (p. 162). If a second colony was established, with land adjoining a former one, it was (at least sometimes) so arranged as to have its balks placed (without regard to the points of the compass), not parallel, but at an angle to the old one (pp. 31, 170).

The century of 200 jugera was, at Beneventum, arranged so as to have 25 actus along the decumans, and 16 along the kardines (p. 59); in other parts the so-called century had only 50 jugera; at Cremona, 210 (p. 30); in other parts 240 jugera, viz. 24 actus by 20; at Emerita, 400 jugera, the decumans having 40 actus and the kardines 20; in a praefecture at Turgalium, the figures were just reversed (p. 171). The distances of the boundary stones or posts often varied in different parts of the same district. If land which had once been divided and assigned was a second time made the seat of a colony, or of a new distribution of land, the old set of balks would be cut across by the new set (p. 178), and the old boundary stones remaining were liable to be confounded by a careless observer with the new ones. Some of the distributions of the Gracchi and Sulla were thus superseded by later arrangements (p. 165).

10. Another mode of dividing and assigning colonial land is expressly contrasted with the above. Ager divisus adsignatus coloniarum habet condiciones duas, unam qua plerumque limitibus continetur, alteram qua per proximos possessionum [p. 1.89]rigores adsignatum est, sicut in Campania Suessae Aruncae. If divided into plots of greater length than breadth, the division was per strigas; if of greater breadth than length, per scamna: length was reckoned in the direction of north and south, breadth in that of east and west. The terms are used also, in camp measurement, for oblong plots [see CASTRA]. The origin of the terms is obscure: striga is probably a strip (from stringere); scamnum was a farmer's word for unbroken ground (Col. 2.2.25, 3.13.2). Ager strigatus or scamnatus seems then to be land divided into oblong plots, the sides of which were not parallel to decumans and kardines, but to the boundaries of the neighbouring occupiers (Front. l.c. is strangely mistranslated by Rudorff, p. 290). According as the length of plot was north and south or east and west, it was called strigatus or scamnatus. (Nipsus describes land divided into centuries of 24 actus in length and 20 in breadth as ager scamnatus, p. 293, and possibly all the oblong centuries named above should be called scamna or strigae.) This method of division is expressly recommended by Hyginus (205 sqq.) for land in the provinces, subject to a tax, as distinguished from the tax-free land of a colony. A difference of legal condition of the land ought to be indicated by a difference of measurement; and the art of the land-surveyor is capable of dealing with the varieties of the several provinces. Hence he blames those who divided Pannonia per centurias; but, recognising the importance of an accurate survey with measured plots, referable to definite lines made visible by balks and marked by boundary stones, he recommends decumans and kardines of the usual width, but dividing the land into strigae and scamna, the strigae to be twice as long as broad, the scamna to be twice as broad as long (dimidio longior sive latior).

11. In an account of the different districts of Italy from a surveyor's point of view, which is found among the Gromatici, and was probably taken from official records (libri coloniarum), other modes of distribution of land are spoken of. Thus in the account of Atina: Ager pro parte in laciniis et per strigas est adsignatus; of Aricia, Ager ejus in praecisuris est adsignatus (230); of Spoletum, Ager in jugeribus et limitibus intercisivis est adsignatus ubi cultura est; ceterum in soluto est relictum in montibus vel subsecivis, quae reipublicae alii cesserunt, nam et multa loca hereditaria accepit ejus populus (225). Laciniae were probably irregular patches of land; praecisurae pieces bounded by straight lines but not rectangular, and cut larger or smaller according to the goodness of the land (cf. p. 211). The land of Spoletum had, in its cultivated parts, been regularly mapped out in centuries, but was allotted in portions of a certain number of acres divided from each other, not by the regular balks, but by cross or intersecting balks. The rest of the land had been left undivided (in soluto or absoluto). In some parts of Italy the land was registered as having been assigned in nominibus, or more fully in nominibus villarum et possessorum (so at Volturnum, pp. 238, 239), i. e. the register did not contain a list of centuries with a technical allotment of them to individuals, but a list of assigned estates described by their natural boundaries, or the names of their former possessors.

12. In this official account, the idea (though imperfectly carried out, at least in the documents extant) was to give the origin and plan of the first survey and subsequent changes, the persons to whom the assignments were made (e.g. militibus, veteranis, familiae Augusteae, militi metyco), the direction of the decumans and kardines, the character of the balks (limitibus Graccanis, or triumviralibus, or Augusteis, or Gallicis, or montanis, or maritimis, or intercisivis), the existence or not, and sometimes the width, of a public right of way (e.g. iter populo debetur ped. x. or ped. cxx., or iter populo non debetur), the mode of marking the boundaries, whether by artificial stones or posts (termini lapidei, silicei, molares, lignei, roborei, oleaginei; or again, Tiburtini, of stone from Tibur, or enchorii from the spot, or Graccani (like those of Gracchus), Augustei, Tiberiani, Neroniani, &c.), erected at various intervals. (e. g. 60 feet, 80, 100, 120, 150, 160, &c., 600, 720, 840, 1020, 1200, 1440, 1500--all divisible either by 20 or 50, i. e. 2 or 5 rods), or by natural or at least ruder marks.

13. This latter class of marks is found indeed in regularly divided and assigned land, especially in marking the boundaries of private properties, after sales and inheritances have disturbed the original allocation, and the technical limits have not been duly observed. But their principal use is (with or without artificial boundary stones) to distinguish the territory of one colony or borough from another, and the lands occupied by one man from those occupied by another in a district which has not been divided and assigned. Such lands were called occupatorii ( “squatters' lands” ) as regards their legal character, but, not having official surveyors' straight boundaries (cf. Liv. 5.55 fin.), come under the general class of ager arcifinius. Non mensuris actis quisque modum accepit, sed quod aut excoluit aut in spem colendi occupavit (p. 138). These occupiers sometimes made plans of their holdings (possessionum suarum formas), but the plans were private, not public, documents, and consequently did not bind their neighbours or others. The bounds adopted by the occupiers, whether of their own motive or by agreement with their neighbours, were of very various kinds (pp. 138, 361), and different kinds were usual, some in one district, some in others. Unaquaeque regio suam habet consuetudinem is a common maxim of the surveyors (p. 349, cf. 139, 140, &c.).

The boundaries between the territories of adjoining colonies or towns were usually taken from the natural features of the country. The ridges of the mountains (juga montium), watersheds (divergia aquarum), and rivers (flumina) or streams (rivi) were the principal; but the line of division was often marked in their absence, or accentuated even when they were present, by boundary stones (termini). But trenches (fossae), springs (fontes), public roads (vice), chapels (sacella), and tombs (sepulcra, monumenta) were also not uncommonly made use of for this purpose (p. 114). All these were also found as boundaries between occupiers. Roads were not a sure indication of a boundary, for they often crossed an estate instead of bordering it; and this was so whether they were properly public highways made and kept up at the expense of the state, and under a special commissioner [p. 1.90]curator), or only country roads maintained by the neighbouring occupiers under the charge of the village wardens (magistri pagorum, p. 146). Chapels were often erected where three or more estates adjoined (in trifinio, in quadrifinio). They had an altar for each estate, and an entrance on the side facing it. The altars were at 15 feet from the temple (p. 303). The areas thus enclosed were the sacred precincts, and were surrendered by the adjoining occupiers, who, by this devotion of a definite area, were able to keep their crops from being trampled by the people on the days of celebration of the compitalia (p. 57).

Tombs, monuments, gravestones (cippi) were found in the interior of estates, especially where the land was not well adapted for cultivation (p. 140), but the boundary or a road was the usual place for burial (pp. 27, 140, 349).

14. But many other marks of boundaries were used between estates. Such are smaller heights than mountains or hills. These, if not more than 30 feet high, were called brows (supercilia), and with these it was the rule, in the absence of any special agreement, that the whole of the slope belonged to the occupier of the higher ground as a natural support to his land (pp. 42, 109, 128, 143). Or there might be for some considerable distance a continuous bank (ripa decisa, p. 361), which would then be a physical embodiment of the ideal straight boundary-line (pp. 108, 109, 151, 241), or a continuous hollow (cava, p. 281), or, by agreement, sometimes an unploughed edge of land (solidus margo, pp. 129, 152). Trenches were sometimes of a military character, and accompanied by mounds (aggeres, p. 370; Varr. R. R. 1.14). Such mounds were sometimes without trenches, and then in the Reatine district called muri (Varr. l.c.), or there might be walls (macheriae) of stone as near Tusculum, or of brick as in Gaul, or of unbaked bricks as in the Sabine district, or of earth and pebbles as near Tarentum and in Spain (Varr. l.c.); or merely pebbles thrown up into a ridge (congeriae, pp. 221, 150; attinea, p. 139), or a series of pebble heaps called scorpions (scorofiones, p. 142), or earth heaps (grumi, p. 241). Varro speaks of the bounds of estates being marked by wooden fences made of posts near each other, and either intertwined with branches, or pierced so as to admit of two or three horizontal poles, or by close-set trunks of trees; but these are not mentioned by the surveyors, though posts of holm-oak, or olive, or juniper, are mentioned as used for boundary stones (p. 138). Another class of boundary marks consisted of quickset thorns (vepres, p. 147: viva saepes, Varr. l.c.; Col. 11.3.3), often with trees standing in them, like our own hedgerows. Trees are indeed one of the most frequently mentioned boundaries, and in this case, probably from their standing ranged at considerable intervals on the outskirts, like advanced guards or pickets, were called arbores antemissae. These were distinguished from other trees on the estate either by being left to grow untouched, or by being of a different kind, either planted for the purpose, or left when others were cut down (pp. 41, 143). Pines, cypresses, ashes, elms, poplars (cf. Hor. Ep. 2.2, 173), elders, oleasters, quinces, almonds, figs, date-palms, are all mentioned (pp. 143, 354, &c.). Foreign trees were often chosen for this purpose. If trees were planted by common consent, they were clipped or felled by each neighbour in common or in turn. It was very usual, if they belonged to one neighbour only, to mark them by scoring the bark on the side looking to the non-owner. If they were within the five feet which was regarded according to law (i.e. Lex Mamilia) as common to both neighbours, they were marked on each side. A tree at a corner of the boundary was marked with a gamma, T, to point to the new line; if at the meeting of several estates, with a cross (decus or decussis, pp. 127, 140, 144). Sometimes the boundary could be sufficiently inferred from the rows of vines or olives (cf. Cic. Caecin. 8.22) of one neighbour meeting those of the others at an angle, or from some other plain difference in the mode of cultivation (pp. 129, 282).

Boundary stones of every kind were used, both by themselves and as assistants to other marks. They were sometimes of the live rock shaped or marked for the purpose (p. 302), sometimes of separate stones, sometimes of foreign material, triangular, square, rounded, bevelled, polished, notched, hollowed, numbered, worked in a variety of suggestive or symbolical patterns, and placed either along the boundary itself or in the neighbourhood, with indications where the boundary ran. Earthenware jars, wells, and other enclosures3 are frequently mentioned; and the better to preserve proof that a stone was intended to mark a boundary, it was very usual to bury under the stone things which were different from the components of the soil at the place, and not easily perishable. Such were ashes, coal (cf. Aug. Civ. D. 21.4), potsherds, broken glass, chalk, gypsum, or coins (pp. 401 sqq.). Mounds of earth with such enclosures were called botontini (p. 308). The practice seems to have originated in the rites of consecration. The neighbours whose estates met at the place each put their stone on the ground, anointed it, and crowned it with fillets and garlands; then in the pit, dug to receive the foot of the stone, they made a sacrifice, and, having set fire to the victim, dropped the blood into the pit and threw in frankincense and corn (fruges), honeycombs and wine, and, having consumed them all, placed the stone on the glowing embers, and rammed it tight with fragments of rock (p. 140). Wooden posts so used were often from this called pali sacrificales (pp. 43, 208).

15. Both religion and law combined to give sanctity to boundary stones. The rites of the annual festival of Terminus on the 23rd of February, are described in Ov. Fast. 2.639-684, and closely resemble those just described from Sicculus Flaccus. The sacrifice of an animal was not part of the rite according to Dionysius (2.74) and Plutarch (Qu. Rom. 15), so that the custom probably varied in different parts (cf. Hor. Epod. 2.59; Juv. 16.39). The first institution of bounds is referred to Numa: the stones were to be sacred to Juppiter Terminalis [p. 1.91]Ζεὺς ὅριος), and any one destroying or moving them was to be held accursed, with full allowance before God and man to any one to slay them (Dionys. l.c.). The Gromatic collection contains (p. 250) an old prophetic utterance evidently of some Etruscan priest. “Juppiter had claimed Etruria for himself, and bade them measure the plains and mark the fields (signari agros). Knowing man's avarice and land-hunger, he willed that all should be fixed by bounds. Now in that eighth, all but the last age” (the eighth age would begin after 761 years from the beginning of Etruria, Censorin. 17.6), “men would in guile violate the rules. Whosoever shall move a bound to the increase of his own and decrease of another's possession would be condemned by the gods. If a slave, he would suffer from his master. If his master was a party to the wickedness, his house would be rooted up, and his clan (gens) perish. They who moved the stones would get sickness, wounds, and weakness of limb. The land would be stirred with tempest, whirlwind, and earthquake (labes). The fruits would be hurt and dashed down with rain and hail, would perish from the dog-days, would be killed by blight. And many dissensions would arise among the people.”

The earliest law which we know of in which penalties are directed in matters of this kind is the LEX MAMILIA (which see), the substance of which was passed by Julius Caesar before the year 703 U.C. It imposed the duty upon the occupier of restoring a boundary stone placed in any colony, borough, or other district, which had fallen; prohibited any one from building on or ploughing up the balks or decumans, or obstructing the trenches under a penalty of 4000 sesterces for each transgression (in res singulas), to be recovered by any of the colonists or burghers who prosecuted. Moving or removing boundary stones (terminos). maliciously was subjected to a penalty of 5000 sesterces for each stone, half to be given to the principal agent in procuring his conviction, and half to the public chest (p. 263; Dig. 47, 21, 3; cf. Lex Urson. 104 in Bruns, p. 121). Nerva directed any slave, who moved a stone to be capitally punished, unless his master would pay the fine. And Hadrian (A.D. 119) directed persons moving boundary stones, with the object of taking others' land, if of a superior class (splendidiores), to be banished for a longer or shorter time, according as they were young or old; their agents to be chastised and put to public works for two or three years. Stealing such a stone ignorantly to be punished with flogging (Coll. 14.12; Dig. 47, 21). Paulus adds that the banishment was accompanied by a confiscation of one-third of the offender's property; and that slaves offending in this respect of their own will were condemned to the mines. Trees (arbores terminales or finales) were protected as well as the regular termini (Paul. Sent. 5.22.2). Stealing boundary stones and felling boundary trees were also subjects for condemnation in a suit finium regundorum (Inst. 4.17.6).

16. It will be seen that a great deal of the matter with which the land-surveyors were concerned was conventional, and required only aptitude in devising and describing suitable modes of demarcation of land, and knowledge of what was in fact usually meant by certain marks. But scientific knowledge was also a part of the craft, and the writings of the Gromatici contain a variety of mathematical problems. Two instruments are mentioned as in use by them: the gnomon and the groma. The problems to be solved required two things: a determination of what we now call the points of the compass, and a method of setting out on land a right line and a right angle. The gnomon was employed for the first purpose, and was derived by the Greeks from the Babylonians, according to Herodotus (2.109), and something of the kind, i. e. some kind of sundial, was known to the Hebrews in the days of Isaiah (38.8). The groma was employed for the second purpose, and was probably got by the Romans from the Etruscans along with the augurial discipline. Its precise shape is unknown: it was not so elaborate as the dioptra, nor identical probably with the chorobates of Vitruvius (8.5), else he would have used the name groma or said so, but consisted of a moveable instrument capable of standing steadily on the ground, and carrying either two straight bars crossing one another at right angles, or a plate, square or round, with marks in the periphery for the extremities of two straight lines crossing one another at right angles, and for their intersection. The verticality of the staff or stand was probably secured by a plumb-line corresponding to a groove in the staff or otherwise applicable to the stand. The horizontality of the cross-bars or plate was secured by observations of plumb-lines hanging from the extremities, and trying whether one covered the other in the line of sight (p. 32). The cross-bars or plate were fixed on a pivot, so that they could be moved round on the staff. The adjustment of the instrument was procured by shifting it about (percutere gromam, p. 285) until it was balanced (perpensus). The direction of the straight lines to be marked out on the land was given by bringing the end of one arm in a line with a distant object, and by erecting marks along the line so determined. This was dictare rigorem, “right line,” or metas, “poles.” A line at right angles was given by the direction of the arm which crossed the first arm. The point on the ground where the lines crossed was interversura. Sighting the marks was comprehendere signa or metas; verifying this by sighting in the opposite direction was reprehendere metas, &c. Two bars meeting or crossing one another at right angles, probably mounted on a pole, formed an instrument called tetrans (p. 186). Comp. Cantor, pp. 20-22, 72. The instrument above described (analogous to a modern surveyor's cross) is often in the Gromatici called ferramentum (Grom. pp. 32, 191, 285), possibly only in strictness applicable to the iron stand; by others stella (ἀστερίσκος, Heron. ap. Cantor, not. 45; comp. Col. 3.13.13), which perhaps was strictly applicable only to the upper part. But the old name groma is found in the Grom. pp. 170, 180, and written croma, p. 284; and in Hygin. de munit. Castr. § 12; Paul. Fest. p. 96; Non. p. 63. (There seems to be no sufficient reasons for thinking that groma had, as is often stated, any etymological connexion with γνώμων.) [p. 1.92]Where the ground sloped, the rod held horizontally, as tested by a plumb-line fastened to its end, served to give the measurement of the base level. To do this was cultellare (pp. 27, 33).

17. Two methods of finding the proper direction for the decumanus maximus and kardo maximus are given by Hyginus (pp. 188 foll.). The first is also given by Vitruvius (1.6.6). A gnomon or sciotherum, probably an upright pole, was erected on a plain, and a circle was described from that as a centre, and with a radius not so great as the longest shadow thrown by the pole. As the sun rose, the shadow gradually shortened, and, when its extremity touched the circle, the place was marked. After noon the shadow lengthened and again touched the circle. This point being marked also, a straight line joining these points was in a due east and west direction, and was taken as the principal decuman. A straight line drawn from the centre of the circle to the middle of the first line was the kardo maximus, and at right angles to the decuman.

The other method proceeded by observing any three lengths of the shadow quickly following one another, and thus could be practised in the morning, without the necessity of waiting for the afternoon shadows. But it required more geometrical knowledge. Both methods, as they presume the use of the gnomon, are probably of Greek origin, though they are not to be found in any Greek writer now extant (Cantor, pp. 69-72). As the gnomon was not publicly set up at Rome till more than four centuries and a half after the foundation of the city (Plin. Nat. 7.212 sq.), it is probable that it was not known to the Romans long before. The augurs following the Etruscan discipline (Grom. pp. 27, 166) divided the heavens and earth by the indication given by the sunrise; and as they faced the west, so as to look as the sun seemed to them to be looking, the first shadow cast by their own body or rod would give the direction for the east and west line. By the aid of the groma, posita auspicaliter, they drew the decuman, and then the kardo, at right angles (p. 170). But sunrise varied at different times of the year and in different places; in Italy, indeed, within a range of 65° (Nissen, Templum, p. 164); and the proximity of mountains, concealing the real horizon, made a further difference; so that as the latitude of the place and the time of year, when the survey was inaugurated, varied, the balks set out in one place differed considerably in their direction from the balks in another--a fact which, whether theoretically defensible or not, was looked at with dissatisfaction by the later surveyors, armed with Greek astronomy and trigonometrical methods (pp. 31, 183).

18. The art of land-surveying, as we find it in the collection of gromatical tracts and fragments, comprised, amongst other things, the elements of geometry and practical instruction in the calculation of the contents of a variety of superficial figures, and in the use of the groma. Like their successors in modern times, they had to deal with such problems as how to set out a straight line across a deep valley or a high mountain, to calculate the breadth of a river which you could not readily cross, and to make complete and accurate surveys of plots of land, however irregular their boundaries (pp. 31-34; 192, 193; 285 sqq.; 354 sqq.). The methods appear to have been much the same as are now practised by surveyors with the cross; and the ten-link offset-staff is perhaps a fair representation of the ten-foot rod of the Romans.

According to Cantor (p. 85), who has set forth the result of his own and others' inquiries, there is no reason to doubt the truth of the assertions of Varro and Hyginus (Grom. pp. 27, 166), that the Romans owed their system of dividing and marking out land to the Etruscans. So long as the only process required was to set out balks at right angles to one another, no need was felt of trigonometrical science. But when the contents of areas of any given shape had to be calculated, and other like problems to be solved, the Roman or Etruscan discipline was insufficient. This knowledge, first developed in Egypt from the necessities induced by the Nile's annual overflow, is found in an early form in a papyrus as old as 1700 B.C. The Greek school at Alexandria developed it, and in the extant writings of Heron, who lived there about 100 B.C., we actually find most, though not quite all, of the methods for solving mathematical problems taught by the Roman Gromatici (Cantor, pp. 78, 86). The first Roman writer on geometry was Varro (cf. Ritschl, Opusc. 3.359, 474), and it was in his lifetime that Julius Caesar is said to have directed a survey of the whole Roman empire, which was carried out by Augustus, who employed four Greeks and Balbus (cf. Grom. 239, 402) to superintend it. But the story is from a writer of the 5th or 6th century after Christ, and was unknown to the best ancient authorities, and Balbus lived in Trajan's time. That something of the kind was done or attempted is likely enough. Agrippa had a map of the empire engraved on marble and put up in his sister's house; some survey was needed for systematic revenue purposes; and it was Alexandria that furnished Julius Caesar with the astronomical science necessary for his reform of the calendar. About this time, and perhaps in this way, the methods of Alexandrian trigonometry were brought fully into the course of Roman gromatic (Cantor, pp. 83, 84; Ritschl, Opusc. 3.743 ff.; Riese, Geogr. Lat. Min. pp. xi. xxiv.; Marquardt, Staatsverf. ii. p. 200 ff.). The history of land-surveying in the Middle Ages may be further followed in Cantor's work (Die römischen Agrimensoren, 1875).


1 A problem in modern books of land-surveying is to set out a quality line so as to give the right quantity in a field of variable value.

2 Perhaps from malleolus, a hammer-shaped vine-cutting, in Col. 3.63. Rudorff quite mistakes the phrase.

3 In the enumeration of boundary marks there frequently occurs this, “ vel canabula et noverca, ” often with the addition “ quod tegulis construitur ” (sic). What is meant is quite obscure. See pp. 227, 240, 349, 401, &c.

hide Display Preferences
Greek Display:
Arabic Display:
View by Default:
Browse Bar: