The Brochtorff Stone circle, at Xaghra, Malta

The great Neolithic temples on Malta are among the oldest temples in the world, most of them erected before even the pyramids were built. Yet what were they and how did they work? The most important and illuminating excavations of this period were those that took place at the Brochtorff’s stone circle, at Xaghra, from 1987-1994. 


To understand the background to these excavations, it is worthwhile to look briefly at the modern history of Malta. From 1971 to 1984, the fiery and controversial Dom Mintoff was Prime Minister: he did not like the professional classes, and the professional classes did not like him. And the archaeologists simply concentrated on surviving. Following a conference in 1985, an invitation was made to Cambridge scholars to initiate a new joint phase of fieldwork. By the time Mintoff’s party had lost power in 1987, the archaeologists had breathed a sigh of relief and decided to celebrate by carrying out a big, spectacular excavation. Where was this to be? 

On Malta there are two basic types of monument of the Temple Period, which is now dated from 4,000 to 2,500 BC. There were the temples themselves, and the burial sites, or rather, there are many temples, and one major burial site. Were there any more burial sites? 

The great temple at Ggantija was built around 3500 BC making it one of the oldest temples in the world. The Brochtorff Stone Circle was erected 300 metres away.

The most important Temple Site is Tarxien (pronounced Tarsheen), today situated somewhat incongruously in the suburbs of Valletta. However, 500 yards away is the site known as the Hypogeum (Greek for ‘under earth’), or in Maltese, Hal Saflieni which has recently been splendidly restored for visitors. It was discovered and ransacked first by builders and then uncovered by Father Magri in 1901, but subsequently what was left was competently excavated in the 1930s by Sir  Thermisticles Zammit, who, among other things, was the real founder of Maltese archaeology.  He estimated the original deposits as containing some 6,000 skeletons, set on three intercommunicating levels, accompanied by a number of superbly carved idols. It seemed clear that the temple at Tarxien and the Hypogeum must be related to each other, as shrine and burial place. 

The Maltese directors, Professor Anthony Bonanno and the late Dr Tancred Gouder, on the site.

But were there any other such ‘Hypogea’? The obvious parallel lay on the neighbouring island of Gozo, near the small village of Xaghra (pronounced Shara). This is the site of one of the other great temples of Malta, known as Ggantija (pronounced Jigant-ee-ya – it means giantess). Here there are two temple structures, side-by-side, which form the highlight to any visit to Gozo. Some 300 yards away – somewhat further by modern roads – is a mysterious circular field, marked today by the remains of only two remaining megaliths in the field walls. This was excavated in 1826, by the British Lieutenant Governor, Otto Bayer, and, fortunately, the excavations were painted by Charles de Brochtorff – hence it was named Brochtorff’s Stone Circle. Subsequently, the site was lost again, but in the 1950s, it was rediscovered through a study of Brochtorff’s drawings. Here was surely the parallel to the Hypogeum they were all looking for and a prime site for re-excavation.

Aerial view of site at the height of the excavation

The site consists basically of two elements: round the outside was a circle of upright stones, with an entrance marked by two huge monoliths facing the Ggantija temple. However, at the centre of the stone circle, was a natural series of underground caves: these began as simple burial places and rock cut tombs, but the site was then ritualised, and the caves were enlarged to form a huge burial and ritual complex. 

Careful geophysics – the first use of this technique in Malta – uncovered the major caves and large pit dug by Bayer on the site. Shortly afterwards the excavations unusually struck first the earliest intact burial deposits on the site – a burial pit of the preceding Neolithic period, known locally as the Zebbug (pronounced Zee-bouge) phase, after a well-known pottery type. It was discovered in the second season, and was promptly excavated, and has already been fully published in the Proceedings of the Prehistoric Society of 1995. The final monograph of the remaining excavations will shortly be published as a monograph of the McDonald Institute of the University of Cambridge, edited by Caroline Malone, Simon Stoddart, David Trump, Anthony Bonanno and Anthony Pace. The Zebbug tomb consisted of a 70 cm deep vertical shaft, which at the bottom, opened-up into two rock-cut tombs, each large enough to hold a body laid out full length and a little bit more (maximum length 2.6 metres and 1.4 metres high). In fact, they were used for collective burials typical of the Copper Age, so every time a new burial was to be made, the previous burial had to be shoved aside and the bigger bones removed for use elsewhere. This meant that generally the only bones to remain in the chambers were the small bones, which the archaeologists carefully collected and recorded. The greatest number were of left knee-caps, which totalled 65 in all – 54 adults and 11 children. The bones were all carefully analysed and proved to be exceptionally healthy for prehistoric times. The burials mostly had pots with them, some filled with red ochre for the funerary rites, but these too tended to get smashed when the burials were put aside, and only survived as fragments. Nevertheless, they have produced the finest collection so far of Zebbug style pottery. At the entrance to the tomb stood the guardian, an enigmatic pillar, crudely carved with a human head, remarkably like the one found at Zebbug itself.

This small pillar with a crude carving that is possibly meant to represent a human head was found in the early rock-cut tomb that preceded the stone circle and its cemetery. 

There were quantities of red ochre found, all of which was probably imported from Sicily and there were many personal ornaments, including over 400 beads of shell, and 27 pendants – dubbed “jelly babies” by the excavators  because of their resemblance to human figures. There were also 6 V-perforated buttons, 16 miniature stone axes, and 52 sea shells, including one large shell that had possibly been used as a musical instrument. Maltese art was under way. 

The only two complete skeletons were the two latest ones – one on either side. Each of them was buried with a complete pot –  the only complete pots found; one of them was clearly the last because by the time he was interred, the pottery style had changed and he was buried with a Ggantija Phase bowl.

The village of Xaghra showing the relative positions of the Brocktorff circle and the Ggantija temple. 

Around 3,700 BC, this Zebbug phase came to an end – not only the burials, but also the settlement of which there are faint traces attested by a few scattered sherds. The site was abandoned and attention was turned to a new site 300 yards away, where the first of the 2 great temples at Ggantija was erected. This, if not the first, was certainly one of the first of the classic Maltese temples. Then , perhaps 400 years later, a second, even larger temple was added adjacent to the first, making it one of the largest and most impressive of all the Maltese temples. 

And 300 yards away and at least 300 years later, the great burial site at the Brochtorff circle was established, linked probably by a ceremonial route marked by standing stones from temple to burial site.

Round the outside, a stone circle of tightly packed stones was erected, with two large uprights forming the entrance, significantly facing the temple site, only 300 metres away. The significance and indeed the dating of the stone circle is not at all clear – if there was one at the Hypogeum site, it has not been discovered. But it clearly marked the limits of the major burial site and stone was the ready building material. 

Plan of excavations. Note the position of the rock-cut tomb that preceded the circle, and also the row of flagstones marking the entrance to the underground cave system coming in from the east. Two surviving uprights of the stone circle are marked to the north, and a third at the north-west corner of the farmhouse.

Originally the site consisted of a natural series of caves in the ground, into which burials were inserted. Soon these were extended and improved with areas marked off by large stones slabs; the soft limestone was quarried away to form new space for additional burial cells, until it became fully ‘ritualised’ with all the paraphernalia of an impressive site, where the dead were, no doubt, not only buried, but also venerated in a sequence of which we can perhaps begin to reconstruct some of the details. 

The entrance would have been down a pathway from the entrance. Half way along the path, just before the steps down into the underworld, was the holy of holies – the original ancestor – a 40 year-old male, buried fully extended. Several further bodies had been placed above him, but radiocarbon dating suggested that these were earlier, apparently his ancestors, who must have been previously stored elsewhere. This is a very male zone, accompanied by feasted animal bones, and an axe.

Cutaway drawing to show how the tomb worked. The entrance was through the two uprights to the top right. A paved entrance way then led to the first ritual station. A flight of steps then led down into the underground chamber. Facing the visitor was a trilithon. Was this where the most recent burial was exposed? Then behind the visitor in a recess by the stairway was another low table on which the figure of the twin goddesses was placed. The cache of nine small figurines was also found here. Note the large ‘offering bowl’ set in the floor.

Then down some steps – presumably, for just a few survived – into the underworld. Facing you, as you descend was the great trilithon; it has not survived but is seen clearly in the Brochtorff painting of 1829. On the left hand was a display of many skulls, the earlier ancestors. Do we imagine that here the most recent ancestor would have been exposed, awaiting his transition to the afterworld of his predecessors? Beside it was apparently a low table with three supports which may have held the principal idol: this has only survived in fragments, but the fragments suggest that there must have been a half-life-sized statue. Why or when it was destroyed is not clear. Was it destroyed by the 1829 excavators carelessly? Or was it perhaps destroyed in anger in antiquity when the shrine finally went out of use?  Adjacent to it was an offering bowl, a stone jar set into the floor of the cave. 

But what happened to the bodies when they had finally become just bones?  Some of them found their way into the ‘discard zone’ a huge pit just behind the low table,  filled with a huge collection of bones, many covered with red ochre. Some of the skulls had filled with sediment which shows that they had been moved from elsewhere.

There was also a display area, a place where bodies were displayed for a certain period of time, before being moved to different locations. A large number of bones were recovered from at least 162 individuals, roughly equally male/female. There were a few animal bones but these were not meat bones, but display bones: heads and feet, perhaps symbolic of the identity of the buried person (sheep, pig or dog), together with the bones of some scavenging animals. 

The skeletons buried outside the caves at the threshold to the staircase down into the cave. At the bottom of the pit was the articulated skeleton top. Was he the founder of the site? In the filling of the grave were the disarticulated bones , which radiocarbon dating suggested were earlier. Were these his ancestors, originally stored elsewhere?

Just to the left of the entrance was another shrine: again there was another low table set on three upright slabs, and it was here that the main idols were found, buried in the ground in front of it, where they had fallen.  At the centre was the finest piece of all, two corpulent figures set on a wicker bed, and described in the accompanying box.  Adjacent to them was an even more remarkable find, a cache of nine small figurines, each big enough to have been grasped in the hand of a worshipper. They were found altogether in a group as if they had been together in a bag: do we imagine the leaders of perhaps nine different tribes, each clasping their tribal emblem, at the ceremony taking place?  Or do we imagine that the images represent the stages of the ceremony, itself a re-enactment of the life cycle: birth, maturity and death?

Recording in progress in the caves that were hollowed out at the sides of the central burial area.

Around 2,500 BC, the Temple Culture collapsed. The reasons for the collapse is as mysterious as the reasons for its origins, but it appears that it ended abruptly. There was a sharp break, as if the entire population of the island were destroyed and replaced by newcomers. Or did the local inhabitants simply rebel against their Gods?  Now everything was different; suddenly we are in the Bronze Age with the advent of flat axes and triangular daggers.  The pottery traditions too are entirely different, while the burial rite changes dramatically to cremation.

Ironically, the type site for this new period is once again the great temple at Tarxien. Here the temple was overthrown, and was covered by debris. But in the ruins of the main courtyard, an extensive cemetery was excavated by Sir Themistocles Xammit in 1916-17, and this phase is known as the Tarxien cemetery or TC – which has nothing to do with the Tarxien Temple Culture.

Numerous clay figurines were also found among the burials. Most have lost their head and come from the display area in the mortuary complex.

There was a fairly extensive TC settlement at the Xaghra Stone Circle, and this confirms the idea of a sharp break.  It became a dwelling site. There were several spreads of grey material which appears to have been broken down mud brick, the remains of mud brick or mud plastered dwellings, set inside the stone circle principally within the sunken areas of cave collapse.  There was also a collection of spindle whorls and loom weights, again attesting to domestic activity. There was a spread of knives made of obsidian, which were principally imported from Pantelleria, south-west of Sicily, showing that the islands were part of a wide trading network.  However, the most interesting aspect was a piece of negative evidence: not a single sherd of  TC pottery was found in the fillings of the caves: the burial site had already come to an end, and the caves had already been filled up and closed or collapsed, before the TC people came along and made their settlement. There is just one oddity: on the edge of the burial chamber there are some deliberate deposits of vessels or figurines marking the limits of the preceding burial ground. Is this coincidence, or is it a deliberate sense of memory about what they knew had happened in this particular location? Or a piece of intelligent archaeology by the newcomers? 

The screen that stood to one side of the central shrine area.

This TC settlement marks the end of activity on the site. From there on it is abandoned and it is not until 1826 that it was eventually rediscovered and excavated. It was fortunate that the excavations were painted by Brochtorff, for without them even this excavation would have remained unknown; it is perhaps fitting that Brochtorff has given his name to the stone circle which has now so dramatically returned to human consciousness.

The question now emerging is whether each of the temple complexes across the islands also had associated mortuary sites. The search is on to find out!

The excavation was directed by Anthony Bonanno, the late Tancred Gouder, Caroline Malone, Simon Stoddart and David Trump. Steven Ashley was the principal artist, particularly of the cut out.

The final reoprt of the remaining excavations will shortly be published as a monograph of the McDonald Institute of the University of Cambridge, edited by Caroline Malone, Simon Stoddart, David Trump, Anthony Bonanno and Anthony Pace.

Fat ladies? or mother goddesses?

These twin figures sitting on a couch were found lying face downwards in front of the table in the ‘shrine’ area from which they had presumably fallen. Both were headless when found but one head found nearby fits one of the figures.

The style is well known in Malta; they are often called ‘fat ladies’ and resemble similar figurines in other parts of the world, although many are without clear gender. However the Maltese examples are noticeable for their voluminous skirts beneath which miniature feet peep out. Note that the feet are coloured with red ochre. They are sitting on an elaborate couch with an arm to one side and apparently legs.

The figure to the left is apparently holding a baby, or perhaps a miniature version of itself, while the one on the right appears to be holding a bowl. Simon Stoddart has wondered whether the bowl originally contained ochre and that the figures represent the life cycle, the baby representing birth and the ochre death.

The style is fairly well known in the Maltese temple. There was a full sized one, half of which still survives in the Tarxien temple, while the best known example is the ‘Sleeping Goddess’ also found at Tarxien. To suggest that they are ‘mother goddesses’ is to open up a can of worms which is best left unopened. 

Small figurines? or tribal emblems?

The most remarkable find of all are the nine figurines above, which are totally unlike any other art found in Malta – or indeed anywhere else. At first sight they may seem to resemble the Cycladic figurines, but at second glance they are very different – and at least a thousand years older.

There are nine in all, found all together in the shrine area as if they had been tied together in  a bundle or a bag, and had fallen off the shrine table. Six of them are figurines, and three are oddities. The six figurines are all in different states of completion: two of them, numbers 1 and 6, are fully finished with their base decorated by leaf shaped ornamentation. Three of them, that is 2, 3 and 4 are more or less finished, but number five is only a rough-out. There is a standardised hairstyle, that of a bob, but number 2 appears to be wearing a cowl or some form of necklace, while number 1 has some sort of diadem at the front.

In addition there are the three oddities. Number 8 has a human head set on a thick tubular base; it is the only one that will stand alone. No, 7 is a human face set on two stubby legs, while No. 9 is an animal head, possibly a pig.

 All the stick figures are very small, only 6 to 8 inches (15-18 cms) high and it has been suggested that they were designed to be held in the hand. Were they perhaps tribal emblems, to be held by the leaders of the tribes as they participated in ceremonies or mourning? Or are they are further representation of the cycle of life?

Time Line

Three mysteries of Maltese chronology

Why does occupation begin so late?

Malta does not appear to have been occupied until around 5000 BC. Why is this so late?  Perhaps there is an earlier occupation that has  not yet been discovered. There is one feature that could  point towards an earlier occupation, that of mass extinction. In the Late Pleistocene Malta, like many Mediterranean islands, had an exotic animal population, including baby elephants and hippos and giant swans and dormice.  Yet these all died out very suddenly at the end of the Ice Age.  Why?  Such extinctions are often attributed to man, but it seems that man had not yet arrived in Malta. Did such extinctions happen naturally?

Why did the temples begin so early?

Temple building began at around 4000 BC, which makes the temples  the oldest upstanding buildings in the world. This means that the oldest Maltese temples cannot be derived from Egypt or anywhere in the Near East – they are far too early. The oldest parallels appear to lie in some of the chambered tombs of Europe.

Why did the temple culture end so early?

Again we do not know, but it certainly ends very abruptly, around 2500 BC, that is, it ends about the time the great pyramids were being built. Suggestions as to why it ended or where the new Bronze Age peoples come from remain mere guesses.

Maltese Chronology ( From Malta: Prehistory and Temples, by David Trump)


Ghar Dalam                  5,000 – 4,300 BC
Grey Skorba                  4,500 – 4,300 BC
Red Skorba                   4,400 – 4,100 BC

Temple Period

Zebbug                   4,100 – 3,700 BC
Mgarr                           3,800 – 3,600 BC
Ggantija                       3,600 – 3,200 BC
Saflieni                         3,300 – 3,000 BC
Tarxien                         3,150 – 2,500 BC

Bronze Age

Tarxien Cemetery         2,400 – 1,500 BC
Borg in-Nadur              1,500 –   700 BC
Bahrija                            900 –    700 BC


Phoenician                    700 – 550 BC
Punic                             550  – 218 BC
Roman             218 BC – AD 330

Visiting Malta 

Visiting Malta is easy. Thanks to the long English presence, English is spoken everywhere – tea is available everywhere, and pubs are ubiquitous. At times it can resemble any popular English seaside resort. It is nevertheless different. 

There are two main aspects to visiting Malta. There are the temples (of course!) but there is also Malta from the time of the Knights onwards – Valletta is a fine Italianate city, and there are some superb fortifications to be seen and visited. 

And don’t forget the adjacent island of Gozo, which can be visited on a day trip – though some people prefer to make it their base – it is quieter, with many pleasures of its own. 

The only disappointment are the beaches – the geology of Malta is essentially limestone – and beaches depend on sandstone for their sand, so the few beaches tend to be small. So don’t bother with the beaches  – concentrate on the archaeology instead. 

The Blue Guide is disappointing (though nevertheless useful) while the Rough Guide also has quite a lot in the history and archaeology. 

Further Reading

The outstanding guide to the Maltese Temples is Malta: Prehistory and Temples, by David Trump, with fine colour photography, and design and layout, by Daniel Cilia. However this is published by Midsea Books, in Malta,  and is not available in this country except through, Oxbow Books, Park End Place, Oxford, OX1 1HN,  or web, who have it available price £17.50. This combines both a guide book, and a text book account, and largely replaces David Trump’s earlier guidebooks – though the previous edition, now published by the Progress Press, is still useful. Strongly recommended. 

A second, glossily illustrated, larger format book has also just appeared from the hands of Daniel Cilia, entitled Malta Before History (Miranda publishers of Malta,, weighing in at nearly five kilos and $US132.50.  The content reveals to the insider some intriguing elements of the local archaeological politics, as well as original contributions to the landscape and construction of the temples. 

Three other books, both well illustrated and with articles on the Brochtorff Circle, but only available in Malta are:
Maltese Prehistoric Art, ed Anthony Pace, published by the Fondazzjoni Partimonju Malti
Malta, an Archaeological Paradise, by Anthony Bonanno, Professor of Archaeology at the University of Malta, published by M.J. Publications. 
Limestone Isles in a Crystal Sea. The geology of the Maltese islands by Martyn Pedley (of Hull University), Michael Hughes Clarke and Pauline Galea is a brilliant, lavishly illustrated, synthesis of the geology of the islands. It is also an academically informed way of finding the best sandy beaches!

The most accessible publication in England of the context of the Brochtorff Circle is to be found in the pages of the Cambridge Archaeological Journal of 1993, pp 3-19.