Roman surveyors used a groma to set out straight lines and right angles. For levelling they had other instruments, including the chorobates described by Vitruvius in the late first century BCE. The two tools answered different surveying questions. The groma helped determine where a boundary or alignment ran across the ground; a level established whether one point was higher or lower than another. Roads, fields and aqueducts each demanded some combination of these measurements.
A new allotment of farmland needed boundaries meeting at consistent angles. A road could follow straight stretches only where the terrain allowed, with rivers, hills and difficult ground altering the route. An aqueduct posed another problem: its water had to move downhill, sometimes for tens of kilometres. Engineers needed to keep a gradual descent across the entire course, because a section laid uphill could interrupt the flow.
Surveyors also relied on less conspicuous equipment. Measuring rods, markers, assistants and calculations turned observations into distances and plans. Roman roads and land grids still leave traces in the landscape, and written cadastral records provide another view of their work. At the same time, evidence for the precise construction of some famous tools remains incomplete. A modern diagram may be useful without being a securely documented ancient design.
Where Roman surveying came from
Roman surveying drew on earlier Mediterranean practices. Egyptian communities measured agricultural land, while Greek towns and colonies used geometric planning long before Roman armies dominated Italy. People could establish right angles and straight lines without the later Roman groma. Roman expansion greatly increased the administrative uses of surveying. Surveyors laid out public works and helped divide the land Rome conquered and distributed.
The word groma has sometimes been connected with the Greek gnomon through an Etruscan intermediary. Its etymology remains too uncertain to locate the instrument’s invention. Etruscan religious views about orientation influenced Roman approaches to space, and Greek mathematics supplied relevant geometric knowledge. As M. J. T. Lewis explains in his study of the groma, the history of Roman surveying involved more than Romans acquiring one tool from their neighbours.
The people carrying out this work included agrimensores, literally measurers of land, and gromatici, a term derived from the groma. They could serve public authorities, the military or colonial land-allocation projects. Surveying also had legal consequences: a boundary might establish a former soldier’s allotment, affect taxes or resolve a dispute between two property holders. The lines surveyors marked were therefore tied to political authority and ownership.
The groma: a cross that established straight lines and right angles
In the standard reconstruction, a pole supports a horizontal cross with two arms at right angles. Four cords hang from its ends, each with a weight to keep it vertical. Looking through opposite hanging cords lets a surveyor establish a straight direction. The other pair gives a line perpendicular to it. The groma established directions and right angles. It cannot by itself tell the surveyor the distance between two points or the height of the terrain.
To extend a line, the surveyor chose a reference point while an assistant placed a ranging pole further away. The assistant moved until the pole lined up with the relevant hanging cords, and that position could be marked. The next point could then be established from another station. By sighting along the other pair of cords, the surveyor marked a crossing alignment at right angles. Measured intervals between successive lines could produce a rectangular field layout, a camp or a planned street grid.
Reliable measurements required practice and repeated checks. The hanging cords had to settle, the cross had to be set properly, and surveyor and assistant needed clear views of each other’s markers. A small error in angle could become a substantial error far down the line. Stakes had to be placed, distances recorded and stations repositioned. A rugged landscape usually prevented any one sight from spanning the whole route, so surveyors worked through shorter stretches and checked their alignment against fixed points.
Modern diagrams label components with names such as ferramentum for the support, rostrum for the arm projecting from it, and stelletta for the cross. These names and reconstructed dimensions do not establish a standard Roman blueprint. In particular, no surviving evidence establishes that each rostrum had to measure exactly one Roman foot. Roman instruments could differ even when they applied the same geometric principle.
What the Pompeii discovery actually tells us
Pompeii produced the most famous physical evidence: metal parts discovered in 1912 beneath the deposits from Vesuvius’s eruption in AD 79. They are sometimes described as a complete groma, although archaeologists did not find the instrument fully assembled. Matteo Della Corte presented a reconstruction in 1922. Adolf Schulten had already proposed a model with an offset arm, and such modern attempts to fit the components together strongly influenced the familiar illustration of a Roman surveyor’s tool.
An offset arm could solve a practical sighting problem. With the cross directly above its supporting pole, the pole might obstruct the sightline; an offset could put the hanging cords over a marker while the support stood aside. But Thorkild Schiøler disputed the accepted reconstruction in 1994. He noted the recovered cross’s mass, around 4.8 kilograms, and questioned the loads imposed on the proposed support. Not all of the relevant components had been found together in a single intact position either.
In his assessment of the surviving evidence, Lewis also treats the assembly of the Pompeian remains as an open question. A slightly inclined support staff has been suggested as one alternative to the conventional offset design, but it is still a reconstruction. The groma’s geometric principle is better established than the precise assembly of the Pompeian remains. A working replica demonstrates mechanical possibility without proving that every Roman instrument used the same arrangement.
Excavations in Pompeii have offered a possible connection to surveying beyond the metal instrument. Geometric floors in the House of Orion, in Regio V, have been interpreted by Luisa Ferro, Giulio Magli and Massimo Osanna as referring to surveying diagrams. Their 2020 study of the pavements explains the proposed identification. It remains an interpretation of domestic decoration, rather than an inscription naming the owner’s occupation, but it raises the possibility that images associated with technical land measurement appeared in Roman houses.
Measuring distance, checking angles and dealing with wind
Establishing a direction was only the first task; the surveyor also needed distances. A ten-foot rod known as a decempeda offered one means of measuring, alongside cords and repeated markers. A Roman foot was about 29.6 centimetres, so such a rod measured roughly 2.96 metres. By repeating standard lengths, assistants could set intervals between boundaries; more irregular plots required area calculations instead of merely measuring the edges of a rectangle.
A measurement taken along a slope exceeds the horizontal distance between its endpoints. For a survey defining an allotment on a plan, this difference could matter considerably. Shorter measurements and calculations helped surveyors work across uneven ground, although the exact steps used on every Roman project cannot be reconstructed. Technical treatises preserve methods and mathematical principles, not a detailed record of how each team surveyed each individual field.
Wind could displace suspended sighting cords and spoil an alignment. A sensible surveyor could wait until they settled, seek shelter or repeat the observation in calmer conditions. We cannot tell how frequently work had to stop for this reason, because surviving evidence gives no regular account of daily interruptions. The simple shape of the instrument should not be mistaken for simplicity of operation; setting out reliable lines over a large area required experienced people.
Other ancient tools could measure angles or levels. The Greek dioptra used sights for angular observations, while the Roman libra belonged to techniques for establishing height relationships. Vitruvius describes these devices as serving different purposes. In Lewis’s analysis of the libra, Lewis examines the libra and the distinction between horizontal alignment and vertical levelling. A surveyor needed to select an instrument according to the question being measured, rather than expect one device to perform every task.
The chorobates and Vitruvius’s instructions
Vitruvius gives the most detailed surviving ancient account of the chorobates in Vitruvius, On Architecture 8.5. He describes a long straightedge of about twenty Roman feet, or nearly 5.9 metres, resting on supports. Plumb lines hung near vertical reference marks on the braces. If the cords coincided with the marks, the beam was level. The engineer then had a horizontal reference against which to compare the rise or fall of points along a planned route.
Vitruvius also supplied a test for conditions in which wind disturbed the plumb lines. The instrument’s top contained a groove five Roman feet long. Filling that groove with water showed whether the beam was level: water reaching both ends equally indicated an even horizontal reference. It was a check on the instrument, not a device that laid out an aqueduct automatically. Engineers still needed measurements at successive stations and calculations connecting those heights to the fall required for flowing water.
Vitruvius knew other levelling tools, including the dioptra and libra aquaria, but preferred the chorobates. This was Vitruvius’s preference, not a universal practice. It does not show that every Roman engineering crew routinely transported a beam nearly six metres long through difficult country. Such a device would be awkward to move and install. Archaeologists have not recovered a complete original chorobates capable of resolving every question about the design Vitruvius had in mind.
Why the chorobates is still debated
Vitruvius’s description permits several possible reconstructions. Vitruvius explains the length, the plumb lines and the water test, yet scholars differ over the arrangement of the supports. His Latin term ancones may indicate projecting arms or brackets. Familiar modern pictures usually show a table-like structure, but the illustrations are attempts to give shape to a description written almost two thousand years before them.
Engineer Isaac Moreno Gallo proposed and tested a more vertical reconstruction in his work on Roman topography. His published studies on Roman surveying demonstrate an alternative to the widely reproduced table-shaped design. The surviving description does not allow either arrangement to be established as the only ancient version. Experiments can help determine whether a proposed instrument works, while the missing evidence about its original structure remains missing.
These reconstructions imply different ways of conducting a survey. Levelling the instrument itself is only the beginning; the engineer must compare it with a point some distance away and transfer the reference from station to station. A vertical arrangement and a broad table-like frame may permit different procedures for doing that. A successful demonstration with a replica establishes technical feasibility, not that a Roman craftsman built precisely the replica we see today.
Road building required more than a straight line
Roman roads often contain long straight sections, but their routes also accommodated difficult terrain. They faced streams, hills, settlements, marshy ground and places where construction costs would be prohibitive. In his study of Roman road surveying, Lewis describes how routes through hills could consist of successive straight alignments, adjusted to maintain practical gradients. Choosing the course was an engineering decision made in the real landscape, not only a geometrical exercise.
After choosing a practical corridor, surveyors could use the groma to keep the route on a chosen bearing and set out perpendicular distances. Builders still had to settle its width, drainage and embankments, and find workable levels for bridge approaches. A groma did not identify the best mountain pass or determine how a road foundation would behave in prolonged rain. The initial alignment had to be followed by many decisions about building and upkeep.
Surveyors could align their instruments with visible landmarks. Some popular descriptions imagine surveyors relying routinely on columns of smoke to sight their way through thick forests, but surviving sources do not establish such a standard method. A smoke plume could not make an otherwise hidden target visible through intervening trees. Where woodland or uneven ground prevented long sightlines, the surveyors had to carry shorter alignments forward while the builders opened a corridor.
Roads were measured for administration and travel as well as construction. The Roman mille passus comprised a thousand double paces, approximately 1.48 kilometres. Surviving Roman milestones may give the distance from a named community, and many record an emperor’s official titles. They help historians reconstruct routes and understand Roman administration. A milestone preserves the measured distance, not the instrument originally used to establish it.
Centuriation: turning land into a surveyed grid
With centuriation, Roman authorities divided territory into a framework of intersecting lines, especially when establishing colonies and distributing conquered or public land. Two principal axes were the decumanus and cardo, with parallel boundaries forming a larger grid. A common centuria measured twenty actus along each side, or about 710 metres, and comprised 200 iugera, roughly 50 hectares. These standard dimensions describe an important practice, not an unvarying size for every Roman parcel of land.
Land division also recorded the exercise of Roman political power. Roman colonies could allocate measured plots to settlers, often including discharged soldiers. Land once used or occupied by conquered communities might be reassigned, displacing earlier residents. Boundaries, drainage and road construction changed how the surrounding territory was used. The regular lines that archaeology sometimes reveals were administrative decisions about people and property, given durable physical form by skilled measurement.
Roman surveying manuals acknowledge the need to adapt geometric plans to difficult ground. A work attributed to Hyginus Gromaticus includes the discussion of colonial boundaries of rocky land unsuited to the usual measured plots. It also mentions a colony whose main decumanus followed the Via Appia. Surveyors could use existing roads and account for difficult ground rather than impose the same mathematically regular pattern across every hill and ravine.
Some administrative records preserve the results of surveys more directly than surviving tools. At Orange, the Roman city of Arausio in southern France, fragments of Roman cadastral plans carved in marble survive from administrative land maps. They cannot be matched with every modern field or reconstruct the whole original survey, but they demonstrate measurement being incorporated into public records. Land surveyors also needed knowledge of boundary stones, ownership rights and the evidence relevant to disputes between neighbours.
Archaeologists now examine suspected Roman land grids with maps, excavations and spatial analysis. A 2023 study of 67 Italian centuriations weighed terrain alongside possible astronomical orientations; it does not prove that a single astronomical rule determined every grid. A 2024 research project in southern Italy uses historical mapping, archaeology and computational methods to search for surviving boundaries. A straight line in a modern landscape must still be dated before it can securely be called Roman.
Aqueducts: measuring a fall over kilometres
Aqueduct builders had to control elevation along an entire route. Most Roman aqueduct channels moved water by gravity, so the source needed to be above the destination and the conduit had to continue descending despite obstacles. Engineers had to compare elevations at successive locations, then decide whether valleys, slopes or bridges required changes to the course. A chorobates or another levelling instrument could establish individual readings. Keeping the whole watercourse within the required elevation range demanded repeated measurement and careful building.
Vitruvius’s instructions for the chorobates explain how the instrument provides a level reference. In the following chapter, he recommends that an aqueduct channel descend at least half a Roman foot for every hundred feet, equivalent to roughly 0.5 percent. That recommendation should be read as his stated design advice, rather than as the measured gradient of all Roman aqueducts. Actual channels varied according to terrain, water supply and the choices engineers made along their routes.
The water system of Nîmes in Roman Gaul illustrates how shallow a fall engineers could manage. The aqueduct ran approximately 50 kilometres from a spring near Uzès, crossing the Gardon on the three-tiered Pont du Gard. The Pont du Gard site’s technical description reports an average gradient of about 25 centimetres per kilometre, around 0.025 percent. That is twenty times shallower than the general fall recommended by Vitruvius. Across the winding route, the source stood only a little more than twelve metres higher than the destination.
Those small differences in elevation had to be controlled throughout the route. A rise in the channel could interrupt gravity-fed water, while excessive descent elsewhere might require a different line or expensive structures. The Pont du Gard was the conspicuous bridge on a system that otherwise ran largely underground. Once built, an aqueduct also needed supervision. The writings and career of Sextus Julius Frontinus demonstrate the continuing administrative work of inspecting, allocating and maintaining Rome’s water supplies.
What survives of Roman surveying
Reconstructions of Roman surveying draw on several kinds of evidence. The Pompeian metal fragments supply evidence for physical parts of a groma, though not one certain assembly. Vitruvius explains a proposed levelling instrument in writing. Later manuscript copies preserve Roman land surveyors’ discussions of grids, legal problems and measurements. Ancient cadastral inscriptions, excavated roads, boundary markers and traces of fields show how surveying was applied across the landscape.
Each form of evidence has its limits. Illustrations in medieval manuscripts may transmit Roman ideas without reproducing every original drawing accurately. A replica proves that a particular mechanical arrangement is possible, not that it was made in the same way in antiquity. A grid visible on an aerial photograph may include later alterations. The strongest explanations compare tools, texts and excavated terrain instead of relying on any one idealized reconstruction.
The groma allowed surveyors to establish directions, rods and cords helped them measure distance, and levelling devices gave them a way to compare heights. Teams of workers joined those observations together over land that rarely matched the lines in a diagram. Roman roads, aqueducts and cadastral plans testify to that repeated, coordinated work. Rome used these techniques across extensive construction works and land divisions, adapting geometric plans to terrain and competing claims of ownership.
