The Army of the Roman Republic: The Second Century BC, Polybius and the Camps at Numantia, Spain

By Mike Dobson

The main source of archaeological evidence for Late Roman Republican camps is a complex of installations around the Iberian city of Numantia in Spain, excavated by Adolf Schulten in the early 1900s. This book reassesses Schulten and concludes that much of his interpretation is questionable. Radically different alternative reconstructions making use of recent fieldwork are presented for several of the sites. A discussion of dating evidence leads to alternative dates being offered for some of the camps. To aid interpreting the sites, army organisation and art of encampment for the period of the Numantine Wars is discussed. This study gives added importance to the sites at Numantia, for they not only form the main source of archaeological evidence for Late Republican camps, but provide evidence for the form of camp for both the late manipular army and the early cohort one.

• Language: English
• Publisher: Oxbow Books
• Release date: Feb 8, 2007
• ISBN: 9781782975588

Book preview

Introduction

Between 1902 and 1933 the German ancient historian and archaeologist Adolf Schulten (1870–1960; figs 1–2) carried out an extensive programme of excavation and fieldwork in Spain. The most important part of his work was carried out between 1905 and 1912 at Numantia,¹ in north-eastern Spain (figs 5 and 6). It was to make him one of the most notable German archaeologists of the early twentieth century, with much having been written not only about his archaeological work, but also about the man himself (Blech 1995, especially 38, n. 1; Wulff 2004a, ccxliii–cclvi presents a full bibliography of Schulten’s prolific quantity of publications).

Schulten uncovered a complex of Roman camps at Numantia, which he dated mostly to the second century BC. This made the camps the earliest known archaeologically. Schulten’s work indeed still provides the earliest archaeological evidence for Roman camps.

Of particular significance was the fact that the sites overlapped in date with the discourse on the Roman army and camp written by Polybius (History² 6.19.1–6.42.6). This text had long been recognized as fundamental to the understanding of the Roman Republican army, and formed one of the important elements of military studies from the Renaissance period, with analysis of the text going back to at least the fifteenth century (Nissen 1869, 23 n. 1; Fabricius 1932). It also featured heavily in the work of the late nineteenth and early twentieth century German ‘school’ of research into the Roman Republican army, e.g. Marquardt (1881), Mommsen (1893 and 1900), Stolle (1912), Fischer (1914), Veith (in Kromayer and Veith 1928), Fabricius (1932), Oxé (1909 and 1939) and Schulten himself (1905–1933b). Schulten’s discoveries consequently had added importance at the time, as they gave the first and vital insight on the ground into what Polybius described.

Fig. 1 Adolf Schulten, photographed in 1905, aged 35, just before his visit to Spain that year – the year his excavations at Numantia began (Pericot García 1940; with thanks to Michael Blech for information about this photograph).

Fig. 2 Adolf Schulten in his eighties, the ‘usual’ photograph published of him. The Spanish beret well reflects his life-long fondness for Spain; he even partially attributed his long life to good Spanish red wine (DAI, Madrid).

Numantia is also significant in the history of research into the Roman army because of the colleagues that Schulten invited to work with him. The use of such help was mainly because Schulten had only limited excavation experience and so he needed to bring in expert help to actually supervise the excavations, as he himself admitted (1931, 14). Schulten also enlisted others to process the finds (Luik 2002b, ix). Those involved reads like a ‘Who’s who" of early twentieth century Roman-German archaeology, as it included:

Walter Barthel (1880–1915) Director of the Römisch-Germanische Kommission (1914–15).

Hans Dragendorff (1870–1941) Produced the first classification system for samian pottery (1896); first director of the Römisch-Germanische Kommission (1902–11); secretary general of the Deutsches Archäologisches Institut (1911–22).

Ernst Fabricius (1857–1942) (fig. 3) Professor of Ancient History at Freiburg; excavator of several Roman military sites along the German Limes; Director of the Reichslimeskommission (1902–10); probably his most important achievement was to establish publication on the Roman frontier in southern Germany and Austria through the series Der Obergermanisch-Raetische Limes des Römerreiches, which he spent many years editing and also contributing to.

Ernst Haeberlin (1847–1925) Primarily a lawyer in Frankfurt, but more famous as a numismatist.

Heinrich Jacobi (1866–1946) Director of the Saalburg Museum (1910–36).

Ulrich Kahrstedt (1888–1962) Worked on several Roman military sites in Germany; Professor of Ancient History at Göttingen (1920–29).

Constantin Koenen (1854–1929) (fig. 4) Archaeologist at Bonn Museum; discoverer and excavator of several Roman military sites in Germany, most notably the legionary fortress at Neuss (1886–1900).

Maximilian Groller von Mildensee (1838–1920) Technical Director of the Limeskommission der Akademie der Wissenschaften, Vienna; director of excavations at the Roman fortress of Carnuntum (1897–1906) and at Enns-Lorch, Lauriacum (1904–19).

The most important of these, as far as the actual excavations were concerned, was Koenen. Schulten seems particularly pleased to have gained him as a colleague at Numantia, as more than once he praises his expertise and valuable excavation experience gained from having completed important excavations at Neuss and elsewhere (Schulten e.g. 1905b, 162; 1931, 14).

Koenen worked with Schulten for the first season in 1905 and seems to have played a very large part in supervising and recording the excavations (Schulten 1905b, 162; 1931, 15 ff.). Schulten obviously greatly appreciated Koenen’s help as he expected him to work with him the following year, but he was unable to (Schulten 1931, 22). From Schulten’s published comments on who was present at Numantia during the excavations, it seems that Koenen was not in fact present for the next few annual seasons. He did not return until 1910, but Schulten clearly respected Koenen so much, as he had been of the greatest service on the first season (der sich um die erste Kampagne die größten Verdienste erworben hatte – Schulten 1931, 31), that Koenen was invited to direct excavations at Renieblas in Schulten’s absence that year between May and July, as Schulten could not leave Germany before the beginning of August when the university summer semester ended (Schulten 1911, 3; 1931, 31). Koenen was again present the following year, supervising excavations at Renieblas (Schulten 1931, 31 f.).

This means that Koenen had a very significant role in excavating the complex of camps at Renieblas. He had also been very important in the success of actually revealing the city of Numantia itself, in the very first season. It is consequently appropriate that Schulten dedicated the second of his volumes on Numantia to the memory of his friend and co-worker (dem Andenken meines Freundes und Mitarbeiters – Schulten 1931, v).

The third of Schulten’s Numantia volumes was dedicated to another of his collaborators, Fabricius – to the discoverer of the Limes (dem Erforscher des Limes – Schulten 1927b, v). The work of Fabricius was to prove highly influential in the development of ideas about the form of the Roman Republican/Polybian camp. Fabricius noticed similarities between what he had seen in Spain and the description of the camp by Polybius. This led him to produce a reconstruction of the layout of the Polybian camp. It was published in the Journal of Roman Studies in 1932. It has become the generally accepted interpretation of the Polybian camp.

Consequently Fabricius’ plan is still reproduced directly or in a copied form in many works about the Roman army. Fabricius’ whole article has even been reprinted in a relatively recent translation of Polybius’ description of the Roman camp (Miller and DeVoto 1994; regrettably this translation contains many inaccuracies and so is not referred to here). The reevaluation of the Polybian camp below in Chapter Three questions Fabricius’ interpretation and suggests it may in fact display a basic misunderstanding of what Polybius is actually describing.

Fig. 3 Ernst Fabricius (Römisch-Germanische Kommission, Frankfurt).

Fig. 4 Constantin Koenen (Seeling 1979, 46).

Schulten was also indebted to the work and companionship of General Adolf Lammerer (1864–1946), of the Königlich-Bayerischen Topographischen Bureau. His substantial contribution was to carry out very thorough and detailed topographical surveys of the area, and to produce many large-scale maps and plans. Schulten dedicated his fourth volume on Numantia to Lammerer as a token of thanks, expressed very simply – dem General Dr H.C. Adolf Lammerer (Schulten 1929, v).

The importance of Schulten’s work has not diminished since its main publication between 1905 and 1933. Any discussion about the Polybian camp and the general form of Roman Republican camps is usually still based upon Schulten, with little questioning of his interpretation. Such ready use of Schulten should, however, be discouraged. This is partly because his reconstructions are arguably too closely influenced by his interpretation of Polybius. This interpretation, influenced by the views of Nissen (1869; et al. 1904), Oxé (1909), Stolle (1912) and Fischer (1914), is discussed at some length by Schulten during his discussion of Lager III (Schulten 1929, 119 ff.; fig. 53).

Many of Schulten’s ideas are not in contention. His strong interest in Polybius, however, caused him to tend to fit the archaeology to Polybius, rather than having a more objective approach of using Polybius merely as an aid in the interpretation of the remains; this will be seen particularly in his reconstruction of Lager III. Indeed, adopting such a ‘bias’ is precisely how Schulten is able to present such a confident interpretation of the features on the ground, as closely reflecting elements of the theoretical form of camps occupied by the Polybian army.

This is not to denigrate Schulten’s use of Polybius. He is to be agreed with in recognising that the close coincidence in date of the Numantine and Polybian material was an ideal situation, with the archaeological and literary evidence offering the potential of aiding the interpretation of each other. For this very reason, almost as much attention is given in this book to a re-evaluation of Polybius as to the archaeological evidence from Numantia.

The close connection between the archaeological and literary evidence can be seen when ascertaining the nature of the armies that occupied the sites. If reliance had to be placed solely on the archaeological evidence, few details of troop organisation would be apparent. Here Polybius can help. As will be shown below in Chapters Two and Three, the help is nevertheless not without its problems. Despite these it will be suggested that armies based on the Polybian manipular system occupied camps at Numantia and the internal details of these armies’ organisation can be proposed.

However, the picture becomes more complicated, since even as Polybius was writing, the army was starting to change its tactical organisation, from a manipular-based structure to one of cohorts. As discussed below in Chapter Two, the exact process and chronology of this transition to cohorts are uncertain. By using incidental references in literary sources besides Polybius, it can nevertheless be suggested that the transformation was probably complete by the beginning of the first century BC.

The occurrence of the development during the second century means that potentially some camps at Numantia could have been occupied by manipular-organised armies and others by cohort-based ones. This raises the question that two distinct patterns of camp layout may be apparent at Numantia and attributable to the different requirements of a manipular and cohort-organised army (below, Chapter Four); without the literary material, direct awareness of such a question would not be possible.

The literary evidence, including Polybius and the main source for the Numantine Wars, written by Appian within his Spanish Wars (below, Chapter One), also makes it clear that foreign troops comprised part of the armies that occupied the camps at Numantia. The archaeological evidence on its own would be unlikely to reflect this aspect of the armies so directly, if at all. As discussed below in Chapter Two, the general use of foreign troops by Rome is referred to by Polybius and other authors. Their presence specifically at Numantia is attested by Appian. He is imprecise about the structure of these forces and their accommodation. In this case the literary evidence is frustrating – it is known that foreign troops were present, they clearly needed accommodation, but with the details of their organisation unknown and the limited nature of the archaeological record, it will be shown that relating such troops to particular buildings or potentially to a whole installation is impossible, with the exception of a building just outside Lager III.

As with Polybius’ description of the army, his account of the camp is vital in leading to an interpretation of the Numantine sites, but its value has to be qualified. Fundamentally, there is dispute amongst scholars about the general form of the camp described – is it a two-legion camp or half of a four-legion one, and if the latter, what is the form of the twolegion camp? A solution to these questions will be proposed in Chapter Three, but Polybius’ text still leaves some details of layout unclear.

Even with an interpretation of the theoretical form established, there is the problem of knowing how many features of these theoretical camps were used at Numantia. In part this problem is caused by the period of the army and hence of the camp described by Polybius being probably slightly earlier than those at Numantia. It is likely that at least some aspects of army organisation and so also of camp layout would continue into closely succeeding periods. As a result, it is arguably legitimate to see if features on the ground comply with the Polybian manipular camp.

Aspects of particularly the early phases at Numantia will be shown below to exhibit clear signs of features referred to by Polybius, though a complete and straightforward theoretically formed Polybian camp does not seem present. More importantly, without Polybius’ text, the interpretation of many features at Numantia would be less obvious and it would certainly not be possible to suggest the type of manipular-organised troops that occupied particular buildings. Hence the value of Polybius in reconstructing Numantia.

However, Polybius alone cannot provide all the answers. For example, the detailed arrangements of the tents of each unit are not clear from his text. Here the archaeology of Numantia can help, plus in this case, literary evidence in the form of the later tract De Metatione Castrorum.³

The date of this text is uncertain and has often been discussed (e.g. Grillone 1977; Dobson 1979; Frere 1980; Birley 1982). It was certainly not written before the Flavian period. A possibility is Domitian, but as Birley (1982) has pointed out, De Metatione Castrorum probably represents a compilation of material from various periods. It is conceivable, therefore, that the tract describes some practices which were no longer in use at the time of writing, and some which are a suitably adapted form of those originating in an earlier period. Clearly it is hazardous identifying such early material within the tract and there is the uncertainty of establishing the original period and form of such features. Potentially, De Metatione Castrorum could in places reflect Republican practices and so can be used legitimately, with caution, as a source for castrametation in the midsecond century BC.

To reconstruct the detailed arrangements of some areas of the camp and particularly the central administrative range, it will be seen in Chapter Three that neither Polybius nor the archaeology from Numantia can be of assistance. To reconstruct these areas recourse has to be made to other literary sources and retrospectively applying features from later sites.

The analysis of the sites below in Chapter Four will show that the direct value of Polybius diminishes for the later phases at Numantia, for these differ noticeably from the Polybian form. Schulten too was aware that the later sites did not comply so readily with the Polybian scheme. As will be seen, these phases need not be regarded as questioning the validity of Polybius, but, as Schulten also proposes, might instead reflect the changes in the structure of the army from maniple to cohort.

This organisational development presumably meant that the ‘Polybian’ camp had to be adapted accordingly, causing features convenient specifically to maniples to be replaced by arrangements more suitable for cohorts.

It will be claimed that the later phases of the Numantine camps consequently provide evidence for the nature of the changes involved. Schulten believed that some of the later sites (e.g. Lager V and Peña Redonda) displayed evidence for changes in layout caused by moving to cohorts. Schulten also believed that only a form of proto-cohort was present at Numantia, whereas it is argued below that fully fledged cohorts were in use. As a result, Schulten’s reconstruction of the later sites presents a far less radical departure from the Polybian scheme than is proposed here.

In the case of these cohort-based camps at Numantia, the roles of the literary and archaeological evidence change compared to the earlier phases at Numantia; the archaeology helps to overcome the absence of literary evidence for the form of the post-Polybian, cohort camp. The overall reconstruction must still rely on the literature to provide information on the organisation of a cohort-based army, which the layout of the camp would obviously reflect. As a result, from this evidence and by retrospectively projecting features of sites of the early Empire, theoretical forms of the various types of early cohort camps can be reconstructed (below, Chapter Three).

In conclusion, the evidence from Numantia will be seen as reflecting not only the ‘pure’ Polybian manipular camp, but also its departure as the army moved away from its manipular form to one based around cohorts. The camps at Numantia consequently will be claimed to stretch beyond Polybius and look towards the military organisation of the closing century of the Republic. As a result, there is little need to emphasise their high importance for the study of the Roman Republican military system.

With the importance of the sites at Numantia, and particularly their relationship to Polybius, it is surprising that Schulten’s work was not continued. Instead, his work was readily accepted by scholars, many of whom simply used his plans of the reconstructed layout of the sites without examining the accompanying text and detailed plans of what was actually found. Consequently, whenever Numantia or the Roman Republican army were discussed, Schulten’s plans were reproduced and synopses of his text included without query.

Testimony to the ready acceptance of Schulten’s interpretations, even today, is that the book Schulten published in 1933 (1933b), summarising his work, is still in print (published by Arno, 1975). The Spanish translation of this book (Schulten 1945), has even recently been republished (Wulff 2004a), with a long introduction discussing Schulten, archaeology and nationalism (Wulff 2004b).

Such open acceptance is unfortunate, as the review of Schulten’s excavation and postexcavation methodology below in Chapter One reveals the archaeological record to be partially unreliable. Indeed much of Schulten’s interpretation of the camp layouts is questionable and in some cases radically different reconstructions can be offered. Even Schulten’s dating of some of the camps can be disputed and as a result, for example, a different set of camps for the Scipionic circumvallation can be proposed.

One of the main reasons for the lack of work after Schulten is due to research into the Roman army and camp described by Polybius, and the Republican military system in general receiving relatively little interest. This is in stark contrast to the amount of continuous effort that has been given to the study of the Roman army of the Empire and its camps, fortresses and forts. There are only a few notable exceptions to the lack of ‘post-Schulten’ work on the Republican army, namely the studies of Harmand (1967), Sumner (1970), Connolly (1978 and 1981) and Keppie (1984). Republican equipment is now at least receiving greater attention, with, for example, articles often being published in the Journal of Roman Military Equipment Studies (e.g. Bohec 1997; Connolly 1997–2002; Feugère 1994; Horvat, 1997; Iriarte et al. 1997; Liberati 1997; Quesada Sanz 1997a–b).

It can only be hoped that the Republic will receive more focus in the future, particularly since it is in the Republic that so many of the features of the Imperial army have their origins.

Another reason for the lack of continuity from the work of Schulten is that on the Iberian peninsula, in contrast to in Germany and Britain, there has been little interest in Roman military archaeology. For several decades after Schulten, very little fieldwork or excavations took place on Spanish or Portuguese Roman military sites and when work did occur, the results were rarely published (Morillo Cerdán and García-Marcos 2002, 779; Morillo Cerdán 2006, 7).

Fig. 5 The location on the Iberian peninsula of sites mentioned in the text.

An exception is at Antanhol, Portugal. Fieldwork here in 1940 led to the site being proposed as a Roman camp, which, on historical rather than archaeological grounds, possibly dated to as early as the 130s BC, but evidence to support this, including observations during the construction of an airport there in the 1950s, is not convincing (Bairrâo Oleiro 1959; Luik 1997b, 234 f.).

Indeed, there were no new Roman forts or camps discovered on the Iberian peninsula until the 1960s, coinciding with developments in scientific archaeology and the start of aerial photography there (Morillo Cerdán and García-Marcos 2002, 779; Morillo Cerdán 2006a, 7 f.; for resumés of work see Luik 1997b; Mora 2006; and Quesada Sanz and Kavanagh de Prado 2006).

The 1960s also saw the start of publications about the Roman army on the Iberian peninsula. A large influence was the work of García y Bellido, who published on the Roman army in Spain (1961). This, and excavations at well-known military sites, such as at the legionary fortress of Legio VII Gemina at Léon, also by García y Bellido (1970), helped increase interest in Roman military activity in Spain (Morillo Cerdán and García-Marcos 2002, 779). Further publications appeared as a result, such as by Roldán (1974), García y Bellido (1976) and Le Roux (1982).

Research during the 1960s and 1970s was largely centred on the army of the Empire. Excavations did take place at a Republican camp at Almazán in 1968, which had been excavated by Schulten (Gamer and Frías 1969; below), and fieldwork by Tarradell resulted in him proposing an early Roman Republican stone-built fort at Tentellatge (Castell del Vilar) (Tarradell 1978). Keay (1988, 31) accepted Tarradell’s proposal and suggested dating the site to Cato’s activity of 195 BC. The interpretation is nevertheless questionable as no Roman material was recovered from the site, but only Iberian. Also, parallels in construction technique with the late third century BC stronghold at Tarragona suggest that Tentellatge should instead be regarded as a pre-Roman civil settlement (Dobson, in preparation). Work on the Republic in these decades also included a very important reworking of the coin evidence from Numantia by Hildebrandt (1979; below Chapter One).

A proposed early Roman Republican camp at Alpiarça in Portugal was re-examined by Kalb and Höck in 1981/82. Their survey and small excavation were unable to come to any firm conclusion about the origin or date of the site (Kalb and Höck 1983 and 1986). Luik and Fabião include the site as a possible early camp, but both are cautious about this (Luik 1997b, 233; Fabião 2006b). The evidence seems more likely that of a pre-Roman settlement than a camp (Dobson, in preparation).

At Empúries, on the north-eastern coast of Spain, excavations in 1982 and 1983 yielded very partial remains of a stone internal building and a length of stone rampart with two interval towers, dating to the first half of the early second century BC (Aquilué Abadías et al. 1984, 36 ff.; Morillo Cerdán 1991, 148 f.; Mar and Ruiz de Arbulo 1993, 188 ff.). This could be from military occupation, which literary sources show occurred in this area in 218 and 195 BC (Luik 1997b, 215 ff.; Keay 2003, 156). Unfortunately too little of the structure was recovered to be informative about the layout and nature of the associated installation (Aquilué 2006).

The first really sizeable work on the Republican army on the Iberian peninsula did not appear until 1984, with Ulbert’s publication on Cáceres el Viejo. This reassessed Schulten’s excavations at the camp and the smallfinds, and included some comments on those from Numantia (Ulbert 1984).

Ulbert was the start of increased attention paid to the Republican army on the Iberian peninsula, but it is still very limited compared to that of work on the Empire (a comment frequently encountered, e.g. Keppie 1984, 51; Morillo Cerdán and García-Marcos 2002, 779; Pamment Salvatore 1997, 53).

Only a few works of general discussion about Roman Republican military activity in the Iberian peninsula have appeared (e.g. Blázquez Martínez 1999, 2000; Luik 2005a), but include good summaries of confirmed and proposed sites by Luik (1997b) and Morillo Cerdán and Aurrecoechea (2006, 211 ff.).

Over very recent years, interest in Rome’s early military activity in the peninsula thankfully seems to be increasing. Fieldwork and some excavation have taken place, and a few new possible military sites discovered (Morillo Cerdán and Aurrecoechea 2006, 211 ff.).

A notable example of recent work is near Zaragoza, at Los Planos de Mara. This is 4 km from the Celtiberian town of Segeda, which was allied to Numantia. Survey work and small-scale excavation since 1999 strongly suggest the presence of a camp dating to the midsecond century BC, possibly even contemporary with Lager III near Numantia, though significant structural features have yet to be found (Burillo Mozota 2006).

Not surprisingly, since the initial impetus created by Ulbert in the 1980s, most research interest has focused on Numantia.

Despite the amount of interest, it is lamentable that no significant excavation has taken place or been allowed at Numantia since Schulten. The only exceptions are recently at the city itself and a few trenches across the siegeworks (below, p. 392). This is despite increasing damage by present agricultural, building and illegal metal detection activities.

Some aerial photographs of Numantia were published by Blázquez Martínez (1986), but their value is unfortunately limited, due to their poor quality. In 1995/6 aerial photographs were also taken by Braasch, known for his excellent aerial photographic work in Germany. Four of these were published by Luik (1997b, colour plates 1 and 2, with sketch interpretation p. 250 f.); the remainder are unpublished and are stored at the Römisch-Germanische Kommission, Frankfurt and the German Archaeological Institute (DAI), Madrid (pers. comm. Braasch 1998 and Luik 2006).

Since the 1980s, research has been carried out on smallfinds from Numantia, both reassessing those found by Schulten and from subsequent fieldwork. This has included discussions of amphorae (Sanmartí 1985a, 1985b, 1992), pottery lamps (Romero Carnicero 1990) and imported pottery (Sanmartí and Principal 1997). One of the most significant recent publications on the finds from Numantia is by Luik (2002b). This gives a very full discussion of Schulten’s finds that are in the Römisch-Germanisches Zentralmuseum (RGZM), Mainz (below).

Fig. 6 The location of Numantia, between modern Soria and Garray, the camps around it nearby and those to the east at Renieblas (after Mapa Topográfico Nacional de España. Soria 350. 1:50,000).

A regular programme of fieldwalking in the area of Numantia has been carried out since 1982 by Morales Hernández, a locally based archaeologist. The results of this have confirmed aspects of Schulten’s results, but also questioned some, such as the course of the Scipionic circumvallation and Schulten’s attribution of one of the camps to Scipio (1995, 2000, 2001; Morales Hernández and Dobson 2005). Morales Hernández has also published a good summary of research and finds made in the area since Schulten (1995). The discussion of a lead slingshot that Morales Hernández came across during the mid-1980s at Renieblas has added important material to the dating discussion of Lager V (Gómez-Pantoja and Morales Hernández 2002; below, p. 41). The question of dating at Numantia has also been addressed by Sanmartí (1992) and Jimeno Martínez and Martín Bravo (1995).

During the 1990s, by coincidence, two British PhD theses related to the work of Schulten were presented, by Pamment Salvatore (1993) and by this author (Dobson 1996). Both compared the classical literary sources with Schulten’s archaeological evidence from Spain, but reached different conclusions about the interpretation of the Polybian camp, the acceptability of Schulten’s interpretations and the physical development and chronology of the cohort-based camp out of the manipular one.

As regards discussing available archaeological evidence, Pamment Salvatore’s thesis was primarily interested in the layout and planning of Republican camps in general and so discussed only those sites at Numantia where significant portions of the internal layout were recovered (Lager III and V, the Black Phase of Castillejo and Peña Redonda) and included some discussion of the camp at Cáceres el Viejo. In contrast, this author’s thesis looked at all the camps at Numantia and in more detail, but since the study was limited to pre-first century BC sites, it did not consider Cáceres el Viejo. Pamment Salvatore published his thesis in 1996 and it has become a frequently cited work on the Roman Republican army and its camps. He has also published further on the subject (1993–9).

German activity at Numantia was taken up again in the 1990s by Luik (German Archaeological Institute (DAI), Madrid and Munich University). He carried out an archaeologicaltopographical survey at Renieblas, concentrating on Lager V and the area to the southeast of Lager III, between 1997 and 2002 (Breuer et al. 1999; Luik 2002a, 2005b; Luik and Müller 2006; fig. 41). This work has both confirmed aspects of Schulten and created questions about his interpretations.

Luik’s work included working on the smallfinds from Numantia (above, p. 8) and looking at current knowledge of Roman Republican and early Imperial military activity in the Iberian peninsula. This resulted in the publication of a very useful summary discussion of sites, some of which are very conjectural, and included Schulten’s interpretation of the Numantia complex, with four aerial photographs taken by Braasch (Luik 1997b).

This work also formed the background to the military discussion in Luik’s book on the Roman conquest of the peninsula (2005a). This book includes several of Schulten’s colour plans actually printed in full colour, probably published for the first time since Schulten, though they are necessarily at greatly reduced size. It also has numerous colour photographs of the sites taken by Luik, and Braasch’s previously published aerial photographs (Luik 2005a, 56 ff.).

As well as working at Numantia, Schulten investigated a number of sites elsewhere in Spain and in Portugal. One of the most important of these is the Roman Republican camp at Cáceres el Viejo. The camp is unfortunately of limited value in aiding the interpretation of the Numantine sites, partly because Ulbert’s reassessment of the finds recovered by Schulten’s excavations shows that Cáceres el Viejo was much later than all the Numantine installations (Ulbert 1984; below, Chapter One), and partly because the details of its internal layout are too uncertain to be applied retrospectively to the Numantine sites.

Schulten carried out a brief investigation of a camp at Almazán, 34 km to the south of Numantia (Schulten 1912, 96 f.; 1929, 187 ff.). He suggested that it was associated with the Numantine Wars, and specifically Nobilior’s campaigns of 153 BC (1929, 189); hence its inclusion in the fourth volume of his main publication on Numantia (1929, 187 ff.).

The camp is unfortunately of little help in aiding the interpretation of those at Numantia as nothing of its internal layout is known.

Further, its association with the Numantine Wars is questionable, as dating evidence recovered in the 1968 excavations carried out by Gamer and Frias (1969) can be interpreted to suggest that the site dates to the first half of the first century BC, i.e. it is much later than the Numantine sites.

Gamer and Frías were in fact unable to decide from the datable pottery whether the camp dates to the second half of the second century or the first half of the first century BC (Gamer and Frias 1969, 183 f.).

Since sherds of Campanian B, type Lamboglia 5, were recovered from the body of the rampart and the same type was found at Cáceres el Viejo but not at Numantia (Ulbert 1984, 168 ff.), the implication is that Almazán is later than the Numantine camps and possibly contemporary with Cáceres el Viejo.

A surface find of a sherd of a Celtiberian dolium, dating to 133–75 BC (Gamer and Frias 1969, 183), may also support Almazán being later than the Numantine camps, though clearly the sherd’s dating significance is limited by it being unstratified.

A sherd of Dressel 1A amphora found on the surface at Almazán (Gamer and Frias 1969, 182), though of a type present at Numantia but absent from Cáceres el Viejo (Ulbert 1984, 187), may also be consistent with Almazán dating to the first half of the first century BC, since although Dressel 1A dates mainly to the second half of the second century, it continued in use into the first half of the first century BC (Ulbert 1984, 184; Peacock and Williams 1986, 87).

Further uncertainty about the date of Almazán was introduced in the mid-1990s by finds made during trial excavations and building work in the area of the camp. These are unpublished, but the finds have been looked at by Luik. He is undecided about their significance, but they may support occupation in the mid second-century BC, so Schulten’s interpretation could indeed by correct (Luik 2005, 53; pers. comm. 2006; Sabugo and Rodriguez Pérez 2006a).

Schulten carried out small-scale fieldwork elsewhere in Spain and as a result proposed early Roman Republican camps at Aguilar de Anguita, Almenara, Alpanseque and Sobrado (Schulten 1913a, 13 f.; 1927b, 1929, 1933a, 1962).

In a more recent discussion, Luik includes Aguilar de Anguita, Almenara and Alpanseque as possible early camps, but is cautious since there is little firm evidence (1997b, 218 ff.; 2005a, 42 f.). Aguilar de Anguita and Alpanseque are also both included within the summary of Roman Republican military sites produced for the 2006 Roman Limes Conference in Spain (Morillo Cerdán and Aurrecoechea 2006). In this, Aguilar de Anguita is accepted unquestioningly as a camp (Sánchez-Lafuente Pérez 2006), but doubt is expressed about Alpanseque (Sabugo and Rodriguez Pérez 2006b). Almenara is not mentioned at all.

Reassessing the remains at these sites suggests that interpreting them as Roman camps is very dubious (Dobson, in preparation). Aguilar de Anguita and Alpanseque are more likely to be pre-Roman civil settlements, rather than camps associated with the campaigns against Numantia as Schulten proposed; hence their inclusion in the fourth volume of Schulten’s main publication on Numantia (1929, 191 ff.). Almenara is probably medieval. The location of what Schulten found at Sobrado and the features he saw are uncertain, and they have not been identified since then, resulting in complete uncertainty over the significance of this site.

Schulten’s fieldwork in Portugal resulted in him interpreting an octagonal enclosure at Viseu as being a Roman Republican camp. A reassessment of the site suggests instead that the site is post-Roman (Dobson 2000).

¹. Phrases such as at Numantia, around Numantia, and the Numantine sites are used here to mean the Roman camps both in the immediate vicinity of the ancient hilltop city of Numantia and nearby at Renieblas. Specific camps at Renieblas are referred to simply by using Schulten’s names of Lager I, Lager II, etc., i.e. not preceded by the name Renieblas. The sites encircling the city of Numantia will be referred to using Schulten’s practice of the name of the hill or area where they are situated, e.g. Castillejo, Travesadas etc. For the location of these sites see fig. 6.

². For convenience, hereafter this work is referred to simply as Polybius.

³. The title of De Metatione Castrorum, as used by Grillone in his edition of the text (1977), is adopted here in preference to the commonly used De Munitionibus Castrorum (e.g. in the ‘classic’ 1887 edition of von Domaszewski, in Lenoir’s edition of 1979 and in Gilliver’s English translation of 1993), since, as Grillone points out (1977, viii f.), the text is more about measuring out a camp than fortifying one. Also, the tract is not referred to here as ‘Hyginus’ since the authorship of the work is disputed.

1   The archaeological and literary evidence relating to Numantia

The history of the identification of Numantia

The destruction of Numantia by the Romans was a popular story in Spain during the medieval period; it had the appeal of the heroic local people resisting the invader.

The exact location of the city was uncertain. From the thirteenth century a commonly held location was Zamora, as Orosius and Florus suggested that Numantia lay on the river Duero, and this was the largest town on the river at that time (Schulten 1931, 5; fig. 5).

When the text of Appian’s Spanish Wars became known in 1557, the idea of Numantia being on the Duero seemed to be confirmed, as Appian referred to the city being on the river ∆óρıον (Spanish Wars 91; Blech 1995, 39; Schulten 1931, 5 ff.).

Doubts about Zamora being the location of Numantia began during the fifteenth century and opinion started to favour the area of Soria (Schulten 1931, 6; fig. 5).

In 1575 the antiquarian Ambrosio de Morales proposed that Numantia lay a little to the north of Soria on a hill – La Muela – just to the south of Garray (de Morales 1575, 105; Blech 1995, 39, n. 8; figs 7 and 9). This suggestion followed careful study of features mentioned in Appian’s account of the siege, and of the itineraria of the road from Augustobriga to Uxama (modern Muro de Ágreda to Osma; fig. 5), since a number of classical sources stated that Numantia lay where this road crossed the Duero (Schulten 1931, 17 with n. 3; 1933b, 156).

Supportive of this interpretation was the location of the hill relative to the Duero, and on the hill were visible remains of an ancient settlement, in the middle of which was a lone medieval chapel containing a Roman tombstone (Schulten 1931, 9; 1933b, 156; fig. 8).

This location for Numantia was popularised by a tragedy entitled Numancia written by Miguel de Cervantes in c.1581 and by numerous poems which followed this (Schulten 1931, 6 ff.). All of these extolled the virtues of the native Iberians valiantly fighting against the foreign invader.

During the eighteenth century there was renewed interest in the question of Numantia, and antiquarian activity centred its work on La Muela hill by Garray (Schulten 1933b, 156). This greatly strengthened the idea that the site of Numantia was here, with, for example, remains of city walls being observed.

The first major publication of these findings, with the earliest published plan of the site, was by Juan Loperraez in 1788 (Schulten 1931, 10; 1933b, 157).

The first excavations took place in 1803, by a Basque literary figure and senior royal financial official, Juan Bautista Erro y Aspiroz, and the Asociación económica de los amigos del país of Soria. Nothing of these excavations was published other than a Celtiberian inscription (Blech 1995, 40; Erro 1806).

In 1853 Edoardo Saavedra, a respected Spanish antiquarian and an engineer, carried out a programme of road-building in the area of Garray. This led him to research the Roman roads of the area, in particular to try to establish the course of the Augustobriga-Uxama road. As part of this, he surveyed La Muela hill and excavated on it (Schulten 1933b, 157).

Fig. 7 Numantia from the south, from El Cañal hill just west of La Rasa (Schulten 1931, Abb. 5; see fig. 9). La Muela hill (Numantia) is near the middle of the photograph. Garray can be seen beyond the left side of the hill The river Duero curves round the bottom left and near sides of the hill, on to the bottom left-hand corner of the photograph. The camp at Molino lies just this side of the wooded area on the middle right-hand side of the photograph. Dehesilla hill is on the middle lefthand side.

Saavedra also carried out work on La Gran Atalaya hill by Renieblas, to the east of Garray, since this area seemed to him to be a better location for Numantia and hence the line of the Augustobriga-Uxama road (Schulten 1931, 17; figs 6 and 12). Saavedra found remains of walls here, but concluded that they were animal pens (Schulten 1931, 17): Schulten’s work was later to reveal that these same walls were in fact part of a whole complex of Roman camps contemporary with the Numantine Wars. In 1861, Saavedra presented his findings to the Real Academia de Historia; they were not published until 18 years later (Blech 1995, 40, n. 13; Saavedra 1879).

Thanks to the fieldwork and efforts of Saavedra, in 1861 the Academia established the Comision de Monumentos in Soria, to further excavate La Muela hill. These excavations began in 1861 and continued until 1867. Only the 1861 season was ever published (Delgado et al. 1877), the other seasons having merely a summary report, produced for the Academia in 1867 and signed by Fernandez Guerra and Saavedra (Blech 1995, 40 f.; Schulten 1931, 17).

These poor quality excavations found the remains of a Roman city, but concluded that nothing of an earlier Iberian one was to be seen (Schulten 1931, 18; 1933b, 157). This had the negative effect of reducing interest in the site and even meant that farmers were allowed to rob the site for stone, despite the hill being declared a state-protected national monument in 1883 (Schulten 1933b, 157).

Fig. 8 View to the north of La Muela, up the plain of the river Tera, over Garray to the hills beyond (Schulten 1931, Abb. 2). The river Duero can be seen on the middle left-hand side of the photograph, being crossed by the road bridge from Soria. The medieval chapel containing the Roman tombstone is in the foreground.

As a result, at the beginning of the twentieth century it was still in question whether La Muela hill was the site of the Iberian city that resisted the Romans so famously. Schulten set out to prove it (Schulten 1933b, 157).

The archaeological evidence from Numantia – the work of Schulten

The chronology and extent of the excavations

Schulten relates, more than once, rather ‘romantically’ how his interest in Numantia began. He says that it started one night in Göttingen during the winter of 1901–2, when he was reading Appian’s description of Scipio’s siege of Numantia (Blech 1995, 41; Schulten 1914, vii). A few months later, in 1902, he visited La Muela hill for the first time. He was readily convinced it was the location of Numantia, as its topography matched so closely what he had read in Appian. He resolved to carry out a programme of research and excavation there to demonstrate that this was indeed the location of Numantia (Schulten 1933b, 157).

Between 1902 and 1905 Schulten worked on all the relevant available material (Schulten 1914, vii). This led to his first publication on Numantia, in 1905, in which he described all previous work on the subject and the topography of the locality of the hill by Garray (Schulten 1905a).

From this preliminary work, Schulten drew up a list of aims for a programme of fieldwork and excavation. Of prime importance was to excavate further in the city and show that the Iberian city of Numantia lay under the Roman one already known about. Secondly, the Roman siegeworks described by Appian were to be located and investigated. Schulten also wanted to locate the Numantine cemetery, so that he could learn more about Iberian culture (Schulten 1931, 12). From this he hoped to advance knowledge in three specific areas: the Roman art of war, Iberian antiquities, and the siege and defeat of Numantia (Schulten 1931, 13).

It was intended that the work would be completed by 20 workers in 20 days (Schulten 1931, 14). In reality, Schulten carried out substantial excavations annually between 1905 and 1912, with a team of generally 20 to 30 excavators, but several times with as many as 60, digging up to 11 hours a day, and he spent a total of 22 months in the field (Schulten 1931, 33 n. 5; 36). There was also a very minor season in 1927; Schulten refers to this as a small ‘postscript’ excavation (kleine Nachtrag-Grabung – 1929, 143; below p. 21).

Schulten obtained permission to excavate through the efforts of Saavedra, who was chair of the Comision de Monumentos in Soria (Blech 1995, 42 f.). The Spanish authorities ignored a loud protest from some of the local press against foreign excavations on such a nationally important site, but such protest was a hint of what was to come (Schulten 1931, 20; below, p. 16). Funding came from the Göttingen Gesellschaft der Wissenschaften and the Berlin Akademie der Wissenschaften (Blech 1995, 42 ff.; Luik 2002, ix; Schulten 1905b, 162; below, p. 21).

Excavations began during the afternoon of 12th August 1905 on La Muela hill under the supervision of Schulten and Koenen (Blech 1995, 43; Schulten 1905b; 1931, 16; 1933b, 157). By early evening the burnt remains of the expected Iberian city were discovered under the Roman city – The city destroyed by Scipio in 133 BC, the famous Numantia, was found! (Die im Jahre 133 v. Chr. von Scipio zerstörte Stadt, das berümte Numantia, war gefunden! – Schulten 1931, 17). The excavations continued until 10th November when snow and cold prevented their continuation that year (Schulten 1905b, 166; 1931, 19). They recovered a large quantity of Iberian material, revealed many buildings, sections of the defences and street layout. Evidence for a Roman siege was recovered in the form of ballista balls and slingstones (Schulten 1931, 20).

Fig. 9 The area around Numantia showing the Roman military features, surveyed by Lammerer and published by Schulten (reduced from Schulten 1927b, Plan I; original in colour; sheet size 555 x 600 mm).

The first excavation season was not limited to just the city. In the last few weeks Schulten investigated the surrounding area, looking for the Scipionic circumvallation. Sections of this were successfully found in the area of Valdevorrón and Peña Redonda, with both hills also revealing remains of Roman camps (Schulten 1905b, 165; 1927b, 217; 1931, 20). At Castillejo, field walking was carried out which yielded evidence suggestive of the presence of a camp, and two unsuccessful trial trenches were dug (Schulten 1927b, 167).

Fig. 10 Excavation of building 35 at Peña Redonda, viewed from the east (Schulten 1927b, Taf. 10.1). The seated figure, presumably recording in some way, is probably Fabricius – see fig. 21.

Schulten also investigated La Gran Atalaya hill, 5.5 km to the east by the modern village of Renieblas, in an attempt to find the camp occupied by Nobilior 24 stadia (4.44 km) from Numantia, referred to by Appian (Schulten 1929, 4; 1931, 20; Appian Spanish Wars 46; figs 6 and 12). The results of this visit seemed at the time to suggest Iberian and prehistoric rather than Roman occupation; it was not until 1908 that a sequence of Roman camps was found here by Schulten (Schulten 1929, 4; 1931, 20).

By the end of the first year, Schulten had consequently collected more than sufficient information to be able to claim safely that here lay the site of Numantia, and he was inspired to do more work in the area. There is indeed little cause to doubt Schulten’s identification of the site; the archaeological and literary evidence support each other entirely. Schulten’s conclusions were readily accepted by the Spanish.

The consequence was to so quickly create widespread interest in Numantia and stir Spanish national pride, that within 12 days of Schulten’s excavations commencing, the Spanish king dedicated an obelisk as a national monument on the hill.

The king did not visit Schulten’s excavations close by. The official excuse was that the king lacked time (Schulten 1931, 19), but in reality it was an indication of the growing opposition by the Spanish to a foreigner excavating their ‘sacred’ site. This was despite there being a long history of German archaeologists working in Spain (Blech 1999; 2002), but ties in with a period of nationalism being expressed through archaeology (Wulff Alonso 2004b).

This nationalist feeling took published form a few years later, for example, in a book on Numantia by Gómez Santacruz (1914). He was the abbot of Soria cathedral, was an important local figure at the time, an open nationalist, but not an archaeologist. He clearly resented Schulten, and disagreed with most of his interpretations. This is even openly stated in part of the title of his book: Refutación de las Conclusiones Históricasy Arqueológicas defendidas por Adolf Schulten (Refutation of the Historical and Archaeological Conclusions presented by Adolf Schulten). Gómez Santacruz’s disagreement with Schulten included claiming that Schulten’s Roman military sites were actually native Iberian settlements. Such claims were not based on any valid archaeological reasons, and his views can be readily dismissed (Bouchier 1915, Morales Hernandez pers. comm. 2007).

During the winter of 1905–6 the Spanish parliament approved a large sum of money to be given annually for further excavations at Numantia. Schulten was not surprised by this, given the importance of the site and the popular support for further work there (1931, 21). He did not expect to be told in April 1906 that the money was only for Spanish archaeologists, that he was not allowed to dig any more at Numantia and only the Comision de Monumentos in Soria would be allowed to work there. Thanks to the efforts of Saavedra, who actually supported the continuation of Schulten’s work, Schulten’s (initially rejected) request to excavate at both the city and the area around Numantia was granted. Schulten wisely only continued work away from the actual city, but even then, a previously favourable land owner of Peña Redonda had to be persuaded to allow non-Spanish archaeologists onto his land (Blech 1995, 44 f.; Schulten 1931, 21 f.).

The Spanish carried out excavations in the city in 1906 (Schulten 1931, 23 ff.). Schulten was horrified by this work and declares that it was a second destruction of Numantia and destroyed more than Scipio had done:

So bedeutet denn die spanische Ausgrabung eine zweite Zerstörung von Numantia und diese Zerstörung ist schlimmer als die scipionische …

(Schulten 1931, 24)

Schulten comments that the excavations basically sought pottery and objects, and were little interested in observing and recording the stratigraphical development of the site (Schulten 1931, 23 f.). As a result, much information was lost.

Schulten’s subsequent work on the Roman military sites consisted of (for locations, see figs 9 and 13):

1905   5 August–5 November

(Schulten 1907a, 3 ff.; 1931, 22 ff.)

Peña Redonda camp – main excavation season, almost all the camp investigated.

Castillejo – major excavation of camps.

Travesadas – camp confirmed by excavation.

Alto Real – camp confirmed by excavation.

Valdevorrón camp – further excavation.

Peñas Altas hills – excavations revealed remains interpreted as circumvallation, watch tower and fortlet (1907a 17 f.); Schulten later changed his mind and interpreted the same remains as parts of a villa from the Empire (1931, 25), but the earlier Scipionic association is in fact more likely (below, Chapter Four).

Saledilla – excavations found remains of a Numantine suburb.

Dehesilla – camp proposed from fieldwork.

Vega – excavations suggested a fortlet.

Features by river Duero downstream from Molino proposed as a Scipionic fortlet, river dams and bridge abutments.

Circumvallation – investigated north and south from Peña Redonda; Castillejo—Travesadas; Dehesilla—Molino; Vega.

Harald Hofmann (from Heidelberg) and Höller, a surveyor, assisted Schulten; Hofmann created a photographic panorama of the area, which he later published in collaboration with Schulten (Hofmann 1922).

Fig. 11 A visit to the excavations at Castillejo in the area of Building 18 of the Black Phase (Schulten 1933b, Abb. 11). Schulten provides no details about who the visitors are or when the visit occurred. The visit may have been in 1907, when Schulten comments that several distinguished German, Spanish and Portuguese archaeologists visited the excavations. They included Dragendorff Fabricius, de Vasconcellos and the Marquis de Cerralbo (1907b, 461). The seated figure far left may be Fabricius, and standing next to him could be the Spanish archaeologist Juan Cabré Aguiló (with thanks to Michael Blech for helping to identify those shown).

1907   19 July–12 October

(Schulten 1907b, 461 ff.; 1931, 25 ff.)

Vega – main season of excavation, but still uncertain about existence of fortlet.

Travesadas – further excavation.

Valdevorrón – further excavation.

Peñas Altas, Valdelilo and Saledilla hills – further excavation.

Peña Redonda – further excavation (figs 10 and 21), but on a much smaller scale than the previous season.

Molino – camp found and substantially excavated.

Dehesilla – camp confirmed by excavation.

Castillejo – second substantial season, western half excavated.

Circumvallation – investigated Castillejo—Travesadas—Valdevorrón; river Duero—Dehesilla—Alto Real.

Fabricius (see p. 2; figs 10 and 21) helped for part of the season and produced plans of the south-western area of Castillejo (1927b, 168 n. 1; e.g. fig. 104) and Molino (figs 132–3); fig. 10 suggests he also helped with recording at Peña Redonda.

Dragendorff (see p. 2) helped for part of the season and supervised excavations at Saledilla.

Hofmann and Höller assisted for a second year. Their work included planning the north-western area of Castillejo (1927b, 168 n. 1).

Assistance from José Ramon Melida (Director of Madrid Museum of Plaster Casts and Director of excavations at Numantia), Juan Catalina Garcia (Director of Madrid Archaeological Museum) and Mariano Granados (a lawyer) from the official Spanish excavation commission working at the same time at Numantia itself.

Schulten comments that he had several distinguished archaeological visitors that year (see fig. 11).

Fig. 12 La Gran Atalaya hill viewed from the south-west (Schulten 1929, Taf. 1.1).

1908   13 April–1 June

(Schulten 1908a, 477 ff.; 1931, 27 ff.)

Castillejo – third substantial season, excavating eastern half; most of the year’s work was devoted to this; three phases confirmed.

Peña Redonda – further excavation.

Alto Real – further excavation.

Circumvallation – excavations Alto Real—Dehesilla.

Circumvallation – further fieldwork Castillejo—Travesadas.

Renieblas – trial excavation on last day of season revealed camp buildings, much to Schulten’s surprise, since he only expected to find remains of an Iberian settlement.

Map of Numantia area completed by Oberstleutnant M. von Zglinicki, Abteilungschef im Großen Generalstab.

Schulten assisted by Fr. Schmidt (a university building inspector from Erlangen), who was himself helped by Ed. Herold (one of Schulten’s students). Schmidt’s work included planning the eastern area of Castillejo (1927b, 168 n. 1).

1909   28 July–18 September

(Schulten 1909a, 526 ff.; 1931, 29 f.)

Renieblas – most of season devoted to excavations at Renieblas, with substantial areas of Lager III and IV investigated.

Castillejo – minor excavation to uncover NW corner of camps.

La Rasa – camp found by fieldwork.

Schulten praises assistance of Major A. Lammerer, director of the Bavarian Topographical Bureau; his contribution included carrying out topographical surveys of Renieblas, Castillejo and Peña Redonda; this linked in with the previous year’s work of Zglinicki.

Fig. 13 The area around Renieblas showing the camps on La Gran Atalaya, surveyed by Lammerer and published by Schulten (reduced from Schulten 1929, Plan I; original in colour; sheet size 450 x 420 mm).

Fig. 14 Looking east from La Gran Atalaya over the village of Renieblas towards Numantia (Schulten 1931, Taf. 3.3).

1910   end May–end September

(exact dates not given)

(Schulten 1911, 3 ff.; 1931, 30 f.)

Renieblas – most of season devoted to excavations at Renieblas (Lager I–V).

La Rasa – excavation.

Saledilla – further excavation.

Koenen directed excavation until beginning of August, due to Schulten’s delay in Germany (above, p. 3); Schulten included an excavation report by Koenen in his own interim for that year in Archäologischer Anzeiger (1911, 4–10).

Fabricius – second visit to Numantia, at the end of the season; Schulten comments that he gave valued support with advice and action, he also recorded the southern half of Lager III.

Schulten assisted by E. Pfretzschner, an architect, he also planned the northern half of Lager III.

1911   1 August–20 September

(Schulten 1912, 81 ff.; 1931, 31 f.)

Renieblas – most of season devoted to excavations at Renieblas (Lager I–V).

Schulten assisted by Koenen, who also recorded excavations (below, p. 127; fig. 42).

Max Ebert, German prehistorian, assisted.

Kahrstedt (see p. 2) assisted.

Almazán camp (above, p. 10) – excavations.

Hill 1 km north of Soria – visited and remains of walls interpreted as remains of Roman camp.

1912   5 August–24 September

(Schulten 1913a, 1; 1931, 32 f.)

Excavations at Renieblas (Lager I–V).

Schulten supervised excavations at Lager V, while Barthel (see p. 2) carried out investigations at the other camps at Renieblas.

Lammerer carried out more survey work at Renieblas, the surrounding area and Almazán.

Hill 1 km north of Soria – excavations failed to find any evidence to support the walls being Roman; Schulten consequently concludes it could not have been associated with the siege of Numantia and at best it was a summer marching camp; this proposal is not referred to again by Schulten, so presumably he was far from confident about the significance of these remains.

Aguilar de Anguita (above p. 11) – site visited and the generally held interpretation as a summer marching camp of the Numanitne Wars accepted.

1927   a few days in mid-August

Schulten is imprecise about when the work took place other than to say he left for Spain on 15th August and spent a few days at Numantia (1927a, 196).

(Schulten 1927a, 196 ff.; 1929, 143 and 169 n. 2)

Purpose of visit was to clarify some things for the fourth volume on Numantia, which Schulten was then preparing.

Renieblas – small-scale excavations: unsuccessful attempt to find Lager V praetorium, uncovered southern entrance of Lager IV.

Lammerer (now a General) accompanied Schulten.

The results of Schulten’s nine seasons provided him with the information he set out to gain – proving the location of Numantia, establishing details of the Roman military campaigns at Numantia and learning about the Roman army of the period (Schulten 1931, 33). He was pleased with his achievement and rightly so.

Schulten’s funding

Schulten states that he used a total of 39,700 Marks for his work at Numantia between 1905 and 1912 (1931, 39). This funding was from several sources (Schulten 1931, 38 f.) (amounts in German Marks):