Ancient circular enclosures in northern Spain

Dido and Hengist are remembered as early heroes of isoperimetry for having solved the challenge of maximising the area of a land grant made to them by stringing together strips of oxhide and using the resulting closed superthong to trace, respectively, a semi-circle at Carthage and a full circle at Kaercorrei.

What was news to Pappus of Alexandria (“that, of all equilateral and equiangular plane figures having an equal perimeter, that which has the greater number of angles is always greater, and the greatest of them all is the circle having its perimeter equal to them”) is today common knowledge, and any mountain shepherd will tell you that, with a given length of fencing material, the best way of maximising space available within a compound and minimising risk of unobserved attacks or break-outs is to trace a circle.

(Recent small stone folds in the Mediterranean are almost all rectangular, but that is rather different.)

Omitted from the Dido and Hengist stories, however, is the question of how they knew that their enclosure was efficient, which it to say circular, which is to say, that it had a more-or-less fixed radius. Although the concept underlying the myth seems feasible to me, this seems to me a serious practical weakness. (For that matter, neither have I ever heard explained how the Anglo-Saxon hide was measured.)

The following piece by Xabier Cabezón, originally published here, and translated here with his permission, suggests a solution to the problem, as well as throwing light on many other aspects of these most interesting enclosures.

A note on the translation of monte. The DRAE says:

monte. (Del lat. mons, montis).
1. m. Gran elevación natural del terreno. [ie, mountain]
2. m. Tierra inculta cubierta de árboles, arbustos, matas o hierba. [Uncultivated ground covered with trees, shrubs, bushes or grass.]

I have generally used the expression “mountain land”, but readers should be aware of the word’s broader semantic range.

Further, this translation has been crammed in between other activities and has not been checked. Caveat emptor.

Sels, by Xabier Cabezón

This article aims to give an overview of sels and their surroundings, and does not pretend to be exhaustive. Much of it is based on the publications of Díez de Salazar and of Zaldua (see the bibliography).

These curious circular pieces of ground are examined in the following sections of the article:


The word sel is of pre-Roman origin (Real Academia Española, RAE). In western Basque it is called korta, while in the Guipuzcoa area the name used is sarobe, saroi, or saroe. In eighteenth-century Biscay the summer or minor sel was called kortaerdia and the winter or major one kortaosoa. Sel may originally have been a synonym of bustaliza or busto (from the Latin bustum < combustum, burnt), which corresponds to one of the meanings of the Basque ol(h)a (Roncesvalles 1284).


Authors differ as to the definition of sel, but the majority agree that sels were bounded portions of land to be used in accordance with particular conditions. They are almost always related to common (mountain) land.

The RAE’s Diccionario de la lengua española defines them simply as “Meadows in which cattle generally spend the summer.”

Although their origin is unknown, formally speaking it is generally accepted that they emerged in the context of a pastoral economy organised around common (mountain) land and pastures. A sel was ground which was differentiated in terms of its demarcation from other land, and which seems to have been subject to pastoral usufruct. Some authors see sels as synonymous with majadas, livestock folds. The historical permanence of a division into summer and winter sels even suggests the practice of transhumance.

Sels are to be found along the entire Cantabrian slope, and as far away as Extremadura and Castilla-La Mancha.

The first written records referring beyond doubt to them in “the land of the Basques” [Euskal Herria] are from the twelth century, while in Cantabria they are documented from the ninth. It is possible that the origin of the first sels is more ancient: in excavations conducted by Zaldua adjacent to the central stone of sels at Urnieta in Guipuzkoa, on at least two occasions fragments of charcoal were found which C14-dated to the 2nd century AD. While these remains have not been linked directly to the sel itself, the tradition of burying ash below the central, or ash, stone would suggest that this was the case. This charcoal could also have arisen from ploughing the ground to make the sel. It is also feasible that these are the remains of natural fires, or of charcoal made in the area before construction of the sel. [Serendipity from translator: “An ash midden was always established at the entrance to [southern African kraals] and the cattle were driven over the ash thus minimizing the infection of ticks and other pests.”]

Since at least the 14th century, the Guipuzkoa sels have, apart from being places related to the care of livestock, played an important role in forestry. The use of common mountain land, whether or not by means of sels, was regulated by a series of ordinances of the councils or settlements that owned them, as well as by agreements between these bodies and their principal private users, smithies.

There were various types of common land, including montes francos, dehesas and sels:

  • Montes francos were lands available in common use to all inhabitants of the propietary settlement. Working was limited: firewood (branches, general debris, fallen trees) could be supplied to dwellings, but felling was forbidden. Fruit could be collected, and hunting and the maintenance of certain types of livestock was permitted.
  • Another type of common mountain land is the dehesa, ejido, ondazillegi, goiburu, and so on. This land was enclosed, fenced, or simply indicated by boundary stones. Its use was strictly regulated, in good part to avoid deforestation. For this purpose, periods of cutting and repopulation were established, during which charcoal making might be authorised. A certain proportion of trees was trained to grow tall, being destined for large timbers, ship masts, and the like. Montes francos tended to give way to enclosed montes.
  • Sels were also areas limited by stones and enjoyed special protection, in particular with reference to charcoal-making and tree-felling. Protection did not always mean prohibition, since in certain circumstances these activities might be permitted. They continued also to be actively linked with pastoralism.

Díez de Salazar describes the use of common forests in the 14th to 16th centuries thus:

  • Woods were divided as a function of three factors. First, to respond to the needs of smiths, whose influence in contemporary society was very important. Second, to provide inhabitants with a range of essential services (firewood for their homes, wood to build with, a place for their livestock). Finally, to ensure that the enormous wealth represented by the forest benefit the incipient towns and their weak economic resources at a time when various social forces fought to dominate it; in this fashion came into being a series of publicly administered goods (“propios”) whose impact on the local economy was proportional to the forest resources available to the municipality.

  • In order to meet these three needs, forest was divided into various parts:

    1. dehesas, ejidos or parts reserved for smithies collectively or, more commonly, individually;

    2. a common area for the regulated use of all inhabitants, and with a tendency to tree conservation (for example, by forbidding felling);

    3. a part reserved for inhabitants of the municipality, allotted periodically to the highest bidder, solely for the sale of timber.

  • The part allocated to smithies was usually free, which is to say that an area of forest was indicated for their use without payment. However, various conditions were attached: in some places, this provision was linked to an obligation to work a minimum number of quintales [A quintal is a weight of 100 libras, pounds] or to have been inactive for a maximum number of years (generally no more than four); return of the plot to the council if unused; prohibition on the sale of firewood to places in other jurisdictions; payment of tributes to the municipality as payment and recognition of the plot and this provision, and so forth.

In 1415 an agreement was signed between various owners of smithies in the Leizarán and the councils of Berastegi and Elduain. Some aspects relating to sels have been grouped together on a page dealing with sels in the Leitzaran. The agreement mentions that the sels contained huts for cowherds. The smiths also complained of the creation of new sels, depriving them of mountain land for charcoal making. It was agreed to declare sels all those that were demonstrably older than 70 years, which is to day, prior to 1345.

In the 16th century, also in the valley of Leizarán, the smithies were assigned a certain number of sels in which they were permitted to make charcoal. Was the remaining mountain land insufficient, and did the smiths also manage to make charcoal in the sels?

One’s attention is drawn to one aspect related to sels: although sel is a term related to common mountain land, it is evident that there were also private sels, held in particular by old families [familias de parientes mayores, held to be descendants of the “first” inhabitants] and ecclesiastical communities.

Plan of a Berastegi sel
Demarcation plan of a sel (from Garmendia, De etnografía vasca…, p. 148)

Physical constitution

Our knowledge of what sels were like is relatively modern, so that we don’t know what they may have been like in an initial period related solely to herding. It is possible that there was some means of fencing them, and that they were furnished with some kind of hut or accommodation.

Sels have one peculiarity which seems pictoresque to our eyes: they were round plots. Villarreal de Bérriz (Biscay, 18th century) described a sel as, “a [plot of] mountain land in a perfect circle with just a stone in the centre, which they call the ash stone.”

The stone marker at the centre of the sel was given various names: ash stone, piedra cenizal, austerritza, austarr or auts-arri (Villarreal de Bérriz), artamugarri, kortarri. In the 18th century it is also described as a central marker, mojón centrical. The group of names, austarri, austerritza and piedra cenizal refer to the tradition of burying ash under the marker, as described by Villarreal de Bérriz: “The Piedra Cenizal is called in Basque ‘Auts-arria‘, and in both languages it appears that it means Stone of the Ash, doubtless because our first Ancestors made fire on it while they grazed their flocks, and it was law, or custom, that others should come no closer than a certain distance.”

Apart from its possible symbolic significance, the austarri had a primordial function: it was used as the starting point for the measurements used to demarcate the sel. The central marker was the “light, guide, governor” of sels (Hernani, 1784). It might consist of a marker made of stone or of various blocks, also of stone, with, for example, the largest blocks forming a cross.

Monolithic central markers generally had on their upper face an engraving consisting of four, eight (this is most common) or sixteen radial lines, which in theory pointed to a corresponding number of peripherical stones or baztermugarri. These, simpler than the central one, signalled the sel’s periphery, and were placed in accordance with the measurements explained below. The situation of these secondary stones with respect to the central one was frequently based on the cardinal points, without doubt in order to facilitate subsequent location of the stones amid vegetation or forest.

Circular sels are in the majority, but there also exist square sels, which probably constitute an evolution of the circular variety. Specifically, in the Artikutza area, in Goizueta in Navarre, there were more than 100. This estate formed part of the extensive holdings of the Monastery of Roncesvalles.

Zaldua indicates that in 1996 some 30 austarris had been found in all Euskal Herria; the same author calculates that the province of Guipuzkoa alone may contain up to 500 sels.


The size of sels depended on the laws and ordnances of each area, which could vary over time.

In general two sizes can be distinguished. Large ones were known as winter, major or common sels, or korta-osoak. Small ones were called summer or minor sells, or korta-erdiak. They were also known according to their dimensions: 12- or 6- goravilla or codera sels, and so forth.

A unit of measurement was used in Guipuzkoa which until this day has only been encountered in relation to sels: the gorabilla or goravilla, codera or pitipia. It measured seven estados (or brazadas, brazas, toesas), each of which consisted of seven feet. As one foot = 1/3 of a vara = 27.9 cm, the goravilla in these terms was 13.67 metres in length. In western Guipuzkoa the amalauoñ (hamalau oin = 14 [non-Imperial] feet) was also used, equivalent to two estados.

The commonest size of common or major sels was 12 goravillas in radius, equal to 84 brazas, equivalent to 164.05 present day metres. The surface area of a circle with this radius is 8.455 hectares. Minor sels had half the radius: 6 goravillas or 42 brazas (82.03 metres) and 2.11 hectares surface area. These measurements were used ya in 1514. But many there were many sels of other sizes:

sel type notes zones estados (radius) amalauoñ (radius) goravillas (radius) metres (radius) hectares
6 goravillas (summer) 1 Aralar, Ataun 42 21 6 82 2.11
9 goravillas Zeanuri, Gorbeia, Dima 63 9 123 4.76
12 goravillas (winter) 84 estados (summer) 12 Elduain, Oiartzun, Lekeitio 84 42 12 164 8.45
63 amalauoñ 126 estados (winter) 32 Mutriku, Berriatua 126 63 18 246 19.02
168 toesas 4 Orreaga, Oiartzun 168 84 24 328 33.82

  • (1) Juntas de Guipuzkoa
  • (2) Villarreal de Bérriz, 1736
  • (3) Barandiaran, 1928
  • (4) Roncesvalles, 1800

The square sels mentioned in Artikutza were 400m square (16 ha surface area).

Some authors (for example,Carrion 1996)have indicated special features of Basque units of measurement, in particular those based on multiples of 7 and 12, applied in particular to the radius of many sels (12 goravillas, each of 7 brazas, each of which is of 7 feet).


Sels were by definition circular, but their delimitation was achieved by laying eight equidistant boundary stones (generally) on the periphery.

This was always undertaken commencing from the austarri or central stone, and the distance to the peripheral stones was determined with a cord of an appropriate length (for example, 12 goravillas). Once the first bazter-mugarri had been placed the rest were situated using the radial cord (12 goravillas in the example) and another which determined the distance between two consecutive peripheral stones, until eight stones completed the circumference.

With reference to the latter distance (between contiguous stones) it should be noted that it was not measured exactly. Specifically, for a 12 goravilla radial sel, the cord in question should measure 9.184 goravillas, but the use was always indicated of one of 9 goravillas (committing an error of the order of 2%). Presumably the resulting distance between the first and eighth stones was somewhat greater than between the rest (some 14%), but the system of stones complied nevertheless with its mission of indicating the path of the limiting circumference of the sel.

If the placement of the stones had been realised on the basis of (even approximately) the incisions in the upper face of the austarri, and additionally oriented according to the cardinal points, their subsequent location would be less complicated.

In this sense doubts have been expressed if in practice they were delimited as indicated. The biggest problem adduced is that of existing obstacles on the ground (woodland, scrub, etc.), together with the sometimes problematic visibility between the central and peripheral stones. What is known is that all the documents that mention the subject agree in describing the same method.

Central stone of the Urdalekun sel
Austarri of the Urdalekun sel in the valley of Leizarán (Xabier Cabezón, 1984)


Although sels as entities are undeniably ancient, this does not mean that a particular sel necessarily is. For example, in the 15th century new sels were built in the valley of Leizarán. The same may be said of austarris or mojons: the austarri of a sel may be as old as it, but it may perfectly well be of subsequent construction.



Carrión Arregui, Ignacio María.- Revista VASCONIA, nº 24 .- 1996.- pp. 59-79.
Díez de Salazar Fernández, Luis Miguel.- Ferrerías en Guipúzcoa (Siglos XIV-XVI) (2 vol.).- Haranburu editor.- San Sebastián, 1983.
Díez de Salazar Fernández, Luis Miguel.- Ferrerías Guipuzcoanas. Aspectos socio-económicos, laborales y fiscales (Siglos XIV-XVI).- Fundación Social y Cultural KUTXA.- San Sebastián, 1997.
Garmendia Larrañaga, Juan.- De Etnografía vasca (4 ensayos) – El caserío, Ritos fúnebres, Galera del boyero, Las ferrerías.- Colección Documento.- Caja de Ahorros Provincial de Guipúzcoa.- San Sebastián, 1976.
Villarreal de Bérriz, Pedro Bernardo.- Maquinas hidráulicas de molinos y herrerías y govierno de los arboles y montes de Vizcaya.- Sociedad Guipuzcoana de Ediciones y Publicaciones de la RSVAP y Caja de Ahorros Municipal de San Sebastián.- San Sebastián, 1973.
Zaldua Etxabe, Luis Mari.- Saroeak Urnietan – Seles en Urnieta – Stone Octagons in Urnieta.- Kulturnieta, S.A.- Urnieta, 1996.

See also the general bibliography on the Leitzaran page.

Similar posts

  • el carter del rei
    Si has enviat una targeta a algú a Anglaterra aquest Nadal, hauràs hagut d’utilitzar un codi alfanumèric de set caràcters com
  • Interchangeability of nominative and genitive forms of Spanish patronymics?
    I’m thinking of examples like Álvarez/Álvaro, Alves/ Alves, Benítez/Benito, Díaz/Diego, Domínguez/Domingo, Fernández/Fernando, Giménez/Ximeno, Gómez/Guillermo, González/Gonzalo, Gutiérrez/Gutierre, Henríquez/Henrique, Ibáñez/Juan, Juánez/Juan, López/Lope, Márquez/Marco,
  • (The) United States (of whatever)
    Re a post by Amando de Miguel in his interesting, if fairly Pleistocene, language column for Libertad Digital, I’ve compiled a
  • Spring is here (again)
    I have been up the coast a couple of times this week (off again tomorrow) and I don’t think I’ve ever
  • Communal herding arrangements in the Pyrenees
    The sheep and goats above have just arrived back in Plan from low pastures to spend the summer in the mountains,


  1. Well Trevor, quite an interesting read once again. The breadth of your mind never ceases to amaze me. I aspire to thy wisdom.

  2. Forgive my ignorance, but is this completely different from the way the Romans did things? Were the two schools of land surveying, the Romans etc. versus the Basques etc.?

  3. Something a bit like that may have been going on – square vs circle people; squares are more efficient than circles (circular plots were forbidden in Cuba in 1816 (Bosquejo económico político de la isla de Cuba)), and there may have been some agricultural vs pastoral correlation. I know bugger all about this, so you might want to apply elsewhere. It would be interesting to know how the Greeks and Romans laid out stadia

    Here follows a bit of general theory of Roman-type land surveying (the fact that you often find parallelogram Roman forts shows they didn’t have it completely sorted) as well as some other crap, copy-pasted indiscriminately:

    The assignment of lands among the Tuscans was conducted with religious ceremonies, and private property was secured by the sanction of the gods. The first step was to draw two main lines with scientific accuracy ; one from south to north, which, as corresponding to the axis of the world, was termed cardo; the other at right angles to it, which therefore ran due east and west, and was named decumanus. Other lines, as many as were required, were drawn parallel to these two and divided the district, which was to be parcelled out, into allotments of a quadrangular form and of an equal area ; the extremities of all the lines were marked by a row of stones inscribed with numerals. The sale or transfer of landed property by individuals, when the whole estate was not alienated, was always in parts according to the duodecimal scale. See Niebuhr’s sections on this subject.

    When Sesostris made an assignment of land in Egypt, he parcelled out the district into allotments of equal extent and of a quadrangular form, (Herodot. 2. 109 ; ) and as the estates granted to the members of the military caste are said to have consisted of twelve aruree each (2. 141), it is probable that every allotment under a public assignment had its particular subdivisions on the duodecimal scale; Herodotus adds that each arura was the square of 100 Egyptian cubits, (2.168.) According to Niebuhr, the common landmeasure of Etruria was the rorsus, or square of 100 feet. Thus the Tuscan agreed with the Egyptian ; but the Roman differed from both. The jiyerum, as the very name implies, was a double measure ; and the real unit in the Roman landmeasure was the act us, or a square of which each side was 120 feet. Vol. ii. p. 626.

    Muller says it would be extremely arbitrary to derive the limitatio of Etruria from Egypt, merely because the allotments of Sesostris were quadrangular; since similar customs may arise in various countries from very different principles. (Etrusker, vol. ii. p. 152.) But when we remember that in addition to their quadrangular form, the allotments were also of an equal area which was subdivided upon a duodecimal scale, and that the common landmeasure among the Tuscans and Egyptians was the square of 100 feet, whilst that of the Romans was the square of 120 feet; when we see that Tirhakah, king of Cush, can be traced to Italy, and that the Tuscan language may be illustrated from the Coptic, of course the imputation of arbitrariness at once falls to the ground. Neither, as Muller states with respect to the cardo or meridian line, needed the Tuscans to wait for the gnomon till the Greeks had previously received it from the Babylonians in the age of Pherecydes and Anaximenes: Tirhakah’s colony certainly brought with them sufficient science to construct a gnomon and determine the cardinal points.

    The decempeda (percha in Spanish) was a ten-foot pole used by the agrimensores, imperial land surveyors. Similar to the standard measures – lengths of iron – to be found at the gates of medieval towns for purpose of ref by merchants and consumers in that horrible things happened to people who cheated

    [Churches often don’t align east-west, and for a variety of reasons. One:

    Then to her Patron Saint a previous rite
    Resounded with deep swell and solemn close,
    Through unremitting vigils of the night,
    Till from his couch the wished-for Sun uprose.

    He rose, and straight–as by divine command–
    They who had waited for that sign to trace
    Their work’s foundation, gave with careful hand
    To the altar its determined place.

    The sun of course rises at a great variety of different points

  4. The whole region begins to reek of Cervantean memories. Ten miles from the station of Argamasilla is the village where he imagined, and the inhabitants believe, Don Quixote to have been born. Somewhere among these little towns Cervantes himself was thrown into prison for presuming to attempt collecting their rents when the people did not want to pay them. This is what I seem to remember having read, but heaven knows where, or if. What is certain is that almost before I was aware we were leaving the neighborhood of Valdepenas, where we saw men with donkeys gathering grapes and letting the donkeys browse on the vine leaves. Then we were mounting among the foothills of the Sierra Morena, not without much besetting trouble of mind because of those certain circles and squares of stone on the nearer and farther slopes which we have since somehow determined were sheep-folds. They abounded almost to the very scene of those capers which Don Quixote cut on the mountainside to testify his love for Dulcinea del Toboso, to the great scandal of Sancho Panza riding away to give his letter to the lady, but unable to bear the sight of the knight skipping on the rocks in a single garment.

    William Dean Howells
    Familiar Spanish Travels

Your email address will not be published. Required fields are marked *