Re-Construction of the Small Xanten-Wardt Dart Launcher ^†

Abstract

Based on the dimensions of the small Xanten catapult, this study reconstructs a full-scale model to validate its manufacturing techniques and evaluate its effectiveness. The process underscores the role of experimental archaeology: the activity facilitates a dynamic sequence of queries, guides the interpretation of signs—not merely physical ones—refines the perception of the cognitive model, and relies on an interdisciplinary approach and strategy. The reconstruction fosters social engagement and scientific dialogue, supporting the adoption of new strategies for knowledge transmission and cultural valorization. The conclusions of this study contribute to the debate on the causes of damage inflicted by the Roman legions on the perimeter walls of Pompeii.

1. Introduction

There are several reconstructions of the Xanten-Wardt, a small dart launcher from the Imperial era (1st century AD) discovered in 1999 at the bottom of a river, now Lake Südsee, in the administrative district of Düsseldorf (Germany). The minor tributary of the Rhine passed through Castra Vetera [1], a legionary camp of the Roman province of Germania not far from the town, colonized by the emperor Trajan and consecrated to Saints Vittore and Malloso, from which the toponym ad san(c)tos, later became Xanten [2]. The find, imprisoned in the solidified sand on the bottom of the river, has become famous for the use of sophisticated identification techniques (such as computed tomography and X-rays). It is one of the rare examples of a catapult preserved with an almost intact wooden frame. The “capitulum anatonum,” in the definition by Vitruvius—who dedicates Book X of De Architectura [3,4] to the construction of Roman artillery—in fact is missing only the lateral upright; however, this is an opportunity, rather than a defect, allowing one to view the interior where the pair of elastic torsion bundles was housed. After its restoration [5] the engine has come to be exhibited at the RömerMuseum and Archäologischer Park Xanten (APX) in North Rhine-Westphalia (LVR) (Figure 1).

Tatters of filaments stuck on the bars that hold the bundles to the metal flanges aligned with the upper and lower holes of the frame, observed under an electron microscope, reveal the nature of the connective tissue typical of bovine tendons, which is suitable for transmitting the force accumulated by torsion when the bowstring is released. At first glance, the small size of the wooden frame is striking. Compared with the average sizes of powerful missile launchers that existed in the ancient western world, the diameter of the holes drilled in the frame through which the twisted bundles pass, blocked by a bar to the metal flanges, is almost halved (45 mm). According to Vitruvius, the minimum diameter of the hole drilled in the frame is 79 mm (De Arch., X, 9, 401–405). The flanges found at Emporion (now Ampurias, Spain) have similar dimensions (Figure 2a,b). Based on these measurements, Erwin Adelbert Schramm (1856–1935) reconstructed the first full-scale replica in 1912 (Saalburg Museum), and several functional reconstructions exist today (Figure 3).

This measurement [6,7] does not differ significantly from the other diameters found during the excavations conducted between 1984 and 1995 in Teruel (near Caminreal), and subsequent excavations in Cremona (Figure 3). It is logical to deduce that the optimal calibers of ancient light artillery were unified, so that worn or broken components could be easily replaced [8]. This necessity was well understood by the most famous scientists and mathematicians of the time. All, directly, reflect the proportional relationship linking the diameter of the hole, taken as a module, and the size of the individual components measured against the whole. The calculation of symmetries described by Vitruvius is exemplary. For the military engineer following Caesar, the typical size of the scorpion, approximately 33 modules in length, 13 modules width, 14 modules at the base (Figure 4), was linked to the weight of the projectiles. Vitruvius reports the correspondences expressed in Roman pounds and fingers (De Arch., X, 10–13, 19 AD).

Figure 3. (a) At the top: (left) Schramm’s reconstruction of the Ampurias frame, Saalburg Museum, and (right) the Caminreal frame, reconstructed by Baatz, Aalen Museum, photo by Len Morgan, from [9] (Figure 3). (b) In the center: (left) Full-scale reconstruction of the Ampurias catapult by Flavio Russo, Archeotecnica.com and (right) a sculpture of a catapult on the tombstone of Caius Vedennius Moderatus (81–96 AD), “architect of war machines,” found along the Via Nomentana. (c) The full-scale reconstruction of the Xanten catapult by Flavio Russo, Archeotecnica.com.

Figure 3. (a) At the top: (left) Schramm’s reconstruction of the Ampurias frame, Saalburg Museum, and (right) the Caminreal frame, reconstructed by Baatz, Aalen Museum, photo by Len Morgan, from [9] (Figure 3). (b) In the center: (left) Full-scale reconstruction of the Ampurias catapult by Flavio Russo, Archeotecnica.com and (right) a sculpture of a catapult on the tombstone of Caius Vedennius Moderatus (81–96 AD), “architect of war machines,” found along the Via Nomentana. (c) The full-scale reconstruction of the Xanten catapult by Flavio Russo, Archeotecnica.com.

The smallest or largest elastic torsion engines are measured based on these data; the measurements of other components of the system, such as the trigger device, did not vary in proportion to the calibers, being dictated by ergonomic considerations.

It is thus evident that the small Xanten scorpion was not an exception. Flanges of the same size have been found both inside a Tunisian wreck of the Republican era (no. 3, 45 mm, Mahdia, Tunisia) and in the Volubilis catapult of the Imperial era (no. 467, 44 mm, Mauritania, Morocco). The modiolus of the wooden engine found in 1998–2001 during excavations inside the military laboratory and depot near the hill fortresses of Carlisle Castle is also of similar diameter. Alan Wilkins has visually compared the proportions between the reconstruction of the small Xanten scorpion directed by him and the find now on display at the Tullie House Museum [9]. The time span that covers the findings mentioned above confirms the standardized use of catapults, even if of small caliber, as they are easy to handle, dismantle and transport. The authors were also fascinated by these features. They therefore chose to build a full-scale demonstrator, calibrated via the documented measurements, to encourage an interactive dialogue with scholars and students on the wise and sophisticated engineering of Imperial–Republican era artillery. This is a theme around which comparative analyses and certified reconstructions of ballistae and scorpions revolves. The research was funded by the MUR because it is considered original and innovative. The construction phases, almost as if they were moments of an archaeological investigation, follow the steps necessary to create an experimental product. The question that the article aims to answer, focusing on the time gap, is the technical characteristics of the weapon, those on which one can base the reconfiguration of terminal ballistics parameters or, in other words, those characteristics that produced the damage that can be seen in the present day on the extrados of Pompeii’s city walls [10,11].

2. Materials and Methods

In his treatise on siege machines for demolitions at a distance (βελοποιικά) Philo of Byzantium, active between 280 and 220 BC, proportions the ballistae based on the modiolus: “The diameter of the opening that passes the turnbuckle [the elastic bundle, ed.] is the starting point” [8] (p. 193).

It is therefore necessary to define the diameter of the hole and the module to derive the components of dart and stone throwers. None of the treatise writers, however, “indicates the convenient way nor the procedures for the construction of the war machines, nor the way to employ them in a satisfactory manner, it being the habit of authors to write for readers whom they supposed to be already informed about all the details” [8] (p. 173). The contribution of Vitruvius’s treatise (De Arch., X, 9, 1–4, 401–405) is therefore important and should be complemented with the later treatise on machines attributed to the genius of Heron of Alexandria the elder (1st century AD) and transcribed from ancient Greek into Latin [12,13]. Systematically ignored, the volume, translated from ancient Greek into Latin, was rediscovered by H. Kochly and W. Rustow, who reintroduced the elastically propelled artillery treatises by printing the Griechische Kriegsschriftsteller, published in Leipzig in 1853 [7] (p. 30). Real studies began only after the discovery in Lyon, between 1855 and 1987, of a metal box which some assigned as belonging to the driving body of the Heron’s cheiroballistra [14].

The find stimulated studies by enthusiastic scholars [15], who indirectly made understandable the meaning of a kind of axonometric exploded view contained in some folios of the Codex Parisinus inter supplementa Greca 607, foll. 56r-58v, Bibliothèque Nationale, Paris [16] (pp. 201–209). A young army officer, Erwin Adelbert Schramm (1856–1935), settled the controversy surrounding the functioning of the Roman artillery [17]. Based on the interpretation of Philo’s Greek text (βελοποιικά) Schramm, then an army major, built a demonstrator that was tested in front of the Kaiser (1912). The tables attached to his German translation describe to scale the mechanisms [17] (cf. Tab. 7) were studied in depth in the following period, translated into English [6] and repeatedly commented upon [18,19]. The original fragments of Philo’s text are annotated by Flavio Russo, historical advisor to the army staff. Among the working demonstrators reconstructed by the scholar are the scorpion and ballista exhibited at the Archaeological Area of Saepinum, Altilia (CB), (Figure 3b) and the small Xanten scorpion (Figure 3c and Figure 5). Flavio Russo’s theoretical and practical work is the basis for the life-size reconstruction carried out for SCORPiò-NIDI by Dr. Michele Fratino, archaeologist and member of the cultural and creative enterprise JustMO’.

3. Results: Reconstruction of Xanten

It is difficult for us to ascertain to which inventor to attribute the appearance of the propeller sculpted on the tombstone of Caius Vedennius Moderatus (81–96 AD), “architect of war machines” as explained in the slab found along the Via Nomentana in 1816. On the other hand, the formal analogy between the machine carved on the tombstone and the view of the small engine exhibited at the museum of the Xanten Archaeological Park is quite evident (Figure 3b).

The restoration offers scholars the wooden frame, and offers it in an excellent state of conservation for further study. According to the indications provided by Philo of Byzantium, the machine is of the euthyton type: the arms make an “effort in the right direction” (eu-tònos in Greek) to twist the pair of filaments of the bundles. The reverse movement of the arms (palin-tònos in Greek) allows an excursion greater than 90°, giving, with the same module (tònos), greater power to the machine. Even if the dimensions are proportionally decreased, the following objective is achieved: “to launch a missile at a long distance on a given target to deliver a powerful blow” [7]. Although it is a euthyton, the dart launcher found in Xanten draws greater power from the use of iron arms that replace wooden ones. The greater effort to which the frame is subjected is supported by the armor plating that reinforces it (Figure 5).

The module is, as usual (Vitruvius, De Arch., X, 10–13, 19 AD), derived from the diameters of the holes drilled in the wooden frame, a detail precisely recorded in the technical survey drawings of the artefact.

An accurate reconstruction of the small scorpion was carried out by engineers Len Morgan and Tom Feeley under the guidance of Alan Wilkins, an expert in ancient history and archaeology. Their reconstruction is a faithful replica: millimetric precision reproduces construction flaws, while highlighting the functional defects (Figure 6a) [20].

Our demonstrator, by contrast, while based, as is appropriate [21], on the survey of the torsion elastic engine, is primarily aimed at verifying the construction procedures, the use of techniques, and the materials compatible with the period. The analysis is therefore focused on the functioning of the components, such as the shooting and loading/unloading devices of the bowstring, nowadays entrusted to the ingenuity of individuals, as the ancient treatise writers took it for granted that their readers already knew all of the necessary details [8] (p. 173). In the case of the small scorpion, a question to be resolved is the absence of winches, essential for the use of stationary catapults.

Our demonstrator, however, although based, as is necessary for experimental archaeology [22], on material, literary, iconographic sources, intends to verify the functioning of the components and especially those of which there is no mention or description in the ancient treatises. It is implicit, for example, that the “handheld” catapults do not provide winches, which is essential instead for the ground ones. However, despite the small caliber, they had to be loaded with the aid of a mechanism, human strength not being sufficient for the purpose, something which in turn caused the abandonment of the previous gastraphete [8]. The issue was resolved using a ratchet with the aid of a support. The balance point of the weapon in the hand is immediately behind the frame [23]. The endless harpoon and the release device were integrated through comparative study, ensuring that the choices were consistent with the contextual manufacturing criteria and methods of the reference culture (Figure 6a,b).

Our demonstrator replicates the loading device applied by Russo [21]: the mechanical ratchet consists of a toothed wheel, a beak or coplanar tooth that allows movement in only one direction, except for a certain mechanical game related to the distance between the teeth. The teeth of the wheel are asymmetrical and angled so that there is little friction in the sliding direction, and vice versa in the locking direction, the tooth mates with the beak to create a large surface to oppose the movement of the wheel. The orientation of the parts is fixed a priori, a rigid and hinged beak can exploit the force of gravity to move into the locking position (Figure 7).

Particular attention was paid to the construction of the projectile. For catapults, Vitruvius specifies, drawing on Philo, that «All the proportional relationships of the organs of these weapons are calculated on the basis of the length of the dart they have to hurl, the ninth part of which is made to correspond to the hole in the frame, through which passes the bundle of twisted fibers that support the arms» (Vitruvius, De Arch., X, 9, 1–4, 401–405). Knowing the diameter of the modiolus (Figure 8a,b), we can hypothesize the length of the dart according to Heron’s calculation [8], d = L/9. The tip, made of iron and rarely bronze, must be pointed and sufficiently long to penetrate adequately. Forged on the anvil, it was vaguely pyramidal. However, there is no lack of conical tips. Among those preserved in museums, some are inserted into the wooden shaft with cannon-shaped or tang-shaped finishes [24]. Compared with the whole, the metal tip represented a significant fraction of the total weight of the dart. To keep the projectile in the right balance, experience teaches that the center of mass must be positioned at about 1/5 of the total length. To stabilize the trajectory in the stretched shot part, they were equipped with feathers. It is undoubtedly emblematic that Ammianus Marcellinus, a Roman historian and soldier of the late imperial age, mentions in their reports the horizontal position of two feathers, with a third placed vertically upwards to avoid their rubbing along the launch channel [8].

Test

The idea of recording the trajectory ranges to inductively calculate the speed of the dart in relation to the distance covered is supported by elementary knowledge that current digital techniques (e.g., Sabre Sky Screen) make reliable. Accurate measurements are obtained by connecting a photoelectric device to the PC; through the light decrease of the 3D digital timer sensor one can record the passage of the projectile on the screen. Modern equipment for ballistic testing also allows for testing parameters that are useful for calculating terminal ballistics [25,26]. However, the use of a refined technology is not decisive for our interphase objective. The shooting carried out with a mobile device allowed for the capture of 10 frames in a total of one minute. Taking the fixed ground supports as reference, and using the known length of the dart, the displacement in each frame is calculated (Figure 9).

The approximations generated by the perspective distortion, the blurring generated by the scale enlargement, the bending of the arrow and the considerably unstable shot—a result of its placement without support—are evident. Nonetheless, the experiment demonstrates the validity of the approach that allows us to trace the causes from the effects. From the first position of the arrow visible in its length (as recorded in the third frame) the dart would move about 67 cm (

Table 1. Correspondence between the 10 images (60 fps each) and the distance covered by the dart in centimeters.

FRAME		Distance Covered in cm
1	2
2	3
3	4	33.6
4	5	35.8
5	6	35.3
6	7	34.3
7	8	36.1
8	9	33.8
9	10	35.4
10	11	35.1
1	11	344.2

). Converting the distance into meters and the time into seconds, a speed of about 20.65 m/s (i.e., 3.442 m/0.1667 s) is obtained. A negligible quantity compared with that derived and the laws of terminal ballistics formalized in the 1st century AD through the “patient repetition of the result” [8] and theoretically applied to the ballistic effects detected along the urban circle of Pompeii (cf. Russo and Rossi article in these proceedings).

To perform “safe” shots, considering the group of students present, the tension of the bundles was reduced. But the real reason for the low speed compared with the real power of the catapults lies in the elasticity coefficient of the hemp rope, suitable for the construction of a demonstrator for exhibition but certainly not suitable for shooting tests. The shreds of connective tissue that remained stuck and examined under the microscope for Xanten, suggest the use of animal nerve fibers. Their elasticity is today approximated with the production of synthetic fibers intended for the reconstruction of intra or extra-articular ligaments. It would therefore be necessary to repeat the experiment with suitable ropes, perhaps those used and mentioned above. Our construction is therefore neither the first nor the best, but was useful to retrace the historical and archaeological data. It was important to approach construction phases, verify the functioning, and demonstrate the usefulness of a “cognitive gaze” that is common to the tactile and optical experience in order to read the concrete and invisible signs of what was collected and integrated.

4. Discussion

Experimental archaeology is a discipline that has been successfully established in the European panorama in order to study, disseminate and enhance the inherited heritage. Re-enactment and living history follow experiences oriented towards the verification, reproduction, communication and simulation that benefit archaeological sites, especially open-air museums [27]. The idea of relaunching the image of the city of Pompeii, disseminating the experimental activity and the historical–military study carried out within the MUR-funded research, applies the experimental method and provides an interpretation of the archaeological data based on hypotheses considered valid because they are refutable [28].

The reconstruction of the small handheld scorpion contributes to knowledge through a dynamic process of questioning that began with the speculations of Schramm (1856–1935). The demonstrator tested in front of the Kaiser proved to put an end to the disputes raised through the interpretation of the cryptic Greek text (βελοποιικά). About thirty modioli, half a dozen ratchet stops, six supports for coils of various size; a kamarion and several iron and bronze armor plates for catapults and ballistae, a frontal shield and some winch fragments are the finds on which Flavio Russo’s study draws [8,21]. The procedural phases of the 1:1 scale reconstruction, carried out by Michele Fratino for SCORPiò-NIDI, refer to his theoretical and practical work. The undocumented parts were integrated through comparative study. Constant references are the texts by Alan Wilkins and the reconstructions by Len Morgan and Tom Feeley. The practice of building physical models [29] makes the heuristic aspect tangible by proposing multidirectional uses in which the “phenomenized” drawing dispels doubts within a unitary dialectical, cyclical and interactive path. Brunelleschi used these to convince patrons fascinated by the realistic views of the painters [30] (p. 117); Filarete considered them an erudite homage to their clients [31] (40; 207); Alberti used them to verify the calculation of the “symmetries” [32] (I: 860–862); Michelangelo built them to provide construction sites with a safe guide [33]; and Vasari used them to study precision [34] (V: 18; 467) and the very small and the very large.

All objectives are not lost despite the digital constructs, the multimedia and multimodal fruition that accompanies the advanced visualization, the printing of digital copies [35] and the sophisticated calculations on the inverse models.

5. Conclusions

The experience gained through Flavio Russo’s Archeotecnica.com has allowed us to document, almost as if they were moments of an archaeological investigation, the steps necessary to create a material product calibrated to the measurements of the find discovered in the vicinity of Xanten. A dynamic process of investigation supports the experiment. Drawing on documentary and iconographic sources and ancient construction techniques, particular care was taken in choosing the appropriate tools to cut the solid beech wood, beat the iron and forge the bronze parts. The weak point of the full-size demonstrator remains the nature of the filaments twisted by the arms to accumulate elastic energy to be released on command. Despite the evident approximation, the test carried out in the presence of scholars and students demonstrates the usefulness of adopting methods of experimental archaeology. Without forgetting the involvement of the senses and emotions, visitors were transformed into actors, confirming that the best strategy to extend the protection and enhancement of the inherited heritage is their involvement. In situ and online platforms help to amplify people’s ability to socialize around objects and physical environments to make “Science” dialogue with the perception of the cognitive model.

These conclusions fuel the debate on the determination of the causes that produced the damage (1st century BC) detected on the extrados of the city walls of Pompeii and attributed to the siege of Sulla, the topic of the current comparison. Merging new forms of knowledge and promoting a concert of science, humanity and humanism is a duty of our era. The connections fall into the so-called Third Mission. Research and teaching are primary institutional activities of the University, which promotes basic research, applied research and the transfer of technological innovation to the economic–social system, contributing to meeting the development needs of society through its scientific and professional skills. Enhancing research results through the creation of brands, patents and industrial spin-offs leads to the development of innovations suitable to satisfy market demand and needs and/or to generate an improvement in environmental and social impact.