Online Review: Book

The Catapult: A History

Duncan B. Campbell


By Tracey Rihll. Pp. xxiii + 381, figs. 59, maps 6. Westholme Publishing, Yardley 2007. $29.95. ISBN 1-59416-035-X (cloth).

The catapult, we are told, was invented in 399 B.C.E., and was increasingly used across the Hellenistic and Roman worlds in different shapes and forms. This is the story that Rihll has chosen to tell, in an upbeat text where reckless slingers brave “friendly fire,” besieged towns suffer “shock and awe,” and Roman bodyguards “take a bullet for their bosses.” Her lively style incorporates quirky explanations, comparing the action of spanning a gastraphetês (belly bow) to that of “someone using a manual lawnmower in an unkempt garden” (37) or describing the composite bow as “the shape of a beer belly, viewed in profile” (47), and we learn (in one of many digressions) that the emperor Caracalla wore a “full-length hoody” (194). Rihll’s enthusiasm for the subject is clear, although phrases like “if it isn’t, I don’t know what is” will not instill confidence in the reader. So, has she sacrificed scholarship on the altar of popularity?

Early in her book, Rihll observes that “Greco-Roman civilization lasted about a thousand years and covered about a million square miles, and it would be absurd to think that we know about everything significant that happened in it” (21). This does not, of course, give us carte blanche to speculate wildly. Theories must be firmly based on fact; we must be careful when moving from the known to the unknown. At one point, she speculates that “a barrel of some sort” might be added to a catapult in order to facilitate the discharge of multiple lead shot (185). Unfortunately, Rihll is all too fond of this kind of blue-sky thinking, unencumbered by detailed explanation and unsupported by even the tiniest scrap of evidence (sometimes even contradicting the evidence).

The book is broadly chronological in treatment. After an introductory chapter on slingers and archers as the precursors of the catapult (ch. 1), we begin with the reign of Dionysius of Syracuse and the invention of the gastraphetês (ch. 2); any notion that this device might have existed prior to 399 B.C.E. is brusquely dismissed as “nonsense.” In chapter 3, Rihll spins an engaging tale about the fourth-century diffusion of the bow machine (here called the “tension catapult”—but surely all catapults employ tension of some sort?) before covering the early history of torsion catapults under Philip II and Alexander the Great (ch. 4). In chapter 6, we read about the Hellenistic period, with a separate chapter on the technical writers (ch. 7). The Roman experience of catapults is divided into chapters on the Republic (ch. 8), the Principate (ch. 9), and late antiquity (ch. 10). Rihll is rather less sure-footed when negotiating the Roman material, and, although a useful appendix lists (most of) the known catapult remains (the Zeugma material has been omitted), the archaeological evidence is never exploited to the full.

Throughout the book, Rihll flirts with the idea of variation in catapult design and finally commits herself to three basic machines: “the scorpion came in two standard sizes, larger and smaller, as did the ballista,” she writes, taking her cue from Livy’s well-known catalogue of captured Carthaginian matériel at Carthago Nova, “but there appears to have been a third kind of catapult too” (181). She cannot quite put her finger on this third sort but is keen to explain scorpions as “little catapults, personal hand weapons” (182). And in chapter 5, she departs from her timeline to explore these as the main theme of the book.

Of course, it has long been recognized that small, hand-held machines existed alongside larger, stand-mounted ones (e.g., BJb 186 [1986] 126–32), despite the jacket blurb that disingenuously celebrates “the author’s intriguing discovery that there were little personal catapults that were used like rifles.” One of the catapult washers found at Ephyra (Greece) in a pre-167 B.C.E. context belonged to a small weapon designed to shoot foot-long arrows, and the technical writer Philon, writing in the late third century B.C.E., even mentions the possibility of a “half-span” arrow shooter (designed for 4½ in. darts), a machine so tiny that (if it really existed) it must have been handheld. And that is just the Hellenistic evidence. Nevertheless, rather than developing these leads, Rihll embarks on a flight of fancy to suggest that missiles resembling the distinctive ballista bolts discovered in late Roman Dura Europus were specifically designed for handheld catapults. The example she selects, Dura catalogue number 823, has lost the metal tip and the wooden stabilizing vanes common to all of these missiles, but its original dimensions can be restored with confidence. “A catapult built according to Philon’s formula to shoot this bolt,” she writes, “would have had a spring diameter of about 1½ inches, or 36 mm, a stock length of about 585 mm, or 24 inches, and arms about 256 mm, or 10 inches, long” (92). And, she might have added, a dioptra (the aperture through which the arrow exits the weapon) of 36 mm, which is important because the ballista bolt in question, although one of the smallest in the Dura collection, is too thick to have squeezed through such a small aperture. This is not the first of Rihll’s theories to be spiked by an evident unfamiliarity with the material.

While discussing the Hatra ballista, she makes the astonishing claim that it was a euthytone (226). This is perhaps an attempt to prove beyond doubt that it was an arrow shooter because, as the early Roman technical writer Heron explains, “euthytones shoot arrows only, but palintones are called stone-projectors by some because they discharge stones, but they also shoot arrows (or stones) or both” (Heron, Belopoeica 3=W 74.5–9). However, the defining characteristic of the euthytone was possession of a grooved diôstra (slider) to take the arrow, and (as Heron explains) the two torsion springs were fixed “the width of the slider apart” (Heron, Belopoeica 26=104.5–6), creating rather a narrow spring frame. The most striking attribute of the Hatra spring frame is the enormous gap, about 1.5 m, between the torsion springs. We can debate many aspects of the Hatra machine, but one thing is certain: it is definitely not one of Heron’s euthytones.

Euthytone” is not the only technical term to trip up Rihll. Earlier in the book, she takes the artillery scholar Eric Marsden to task for his interpretation of plaisia katapaltôn cited in a well-known fourth-century military inventory from Piraeus (IG2 1627 B). Marsden had argued, quite reasonably, that the plaision should be identified as the spring frame of a torsion catapult, which is normally called the plinthion. But Rihll, apparently unaware of the Suda’s entry for plaision, which gives plinthion as an alternative, would rather translate it as “slider” (65). (The irony of her admonishment that “giving unique meanings to words is, in principle, a dodgy procedure” [65] is sublime.) Her critical gaze falls next on Marsden’s suggestion (and it was only a suggestion) that the sôlênes katapaltôn in the same inscription might be stocks for catapults, for she prefers to interpret these as “the fixed part of the stock which Marsden calls the case” (66). But she is mischievously selective in her citation here, for Marsden was well aware that sôlên is strictly equivalent to syrinx (pipe), the fixed part of the stock; it was perhaps the omission of the other part (the diôstra) from the inventory that prompted his combining the two. Finally, concerning the same inscription, she disparages Marsden’s equation of epistulion (literally, a “lintel”) with the peritrêton, the top part of a spring frame for which he coined the term “hole carrier,” encapsulating its function, which was to carry the bore holes through which the torsion springs are threaded. Far from “straining one’s credulity, as well as the Greek” (65), it seems very reasonable to describe the peritrêton as the lintel of the spring frame. Rihll’s preferred translation as “belly bar” really does strain credulity, but her aim is to remove any signs of torsion artillery from the inventory, so that it can bolster the meager haul of evidence for nontorsion bow machines.

The literary sources are notoriously silent on the subject of bow machines. So Rihll exploits the so-called artillery towers of Attica and Boeotia, diagnosed as such from the presence of windows (as opposed to loopholes) on the upper floors; but surely windows are primarily provided for surveillance and cannot be prima facie evidence of bow machines. She announces the arrival of bow machines at the court of Archidamus III “about 380 at the earliest” (48), but what Archidamus actually saw was a katapeltikon belos (catapult arrow), probably presented to him by Sicilian troops, who were arriving on Spartan soil in the early 360s. Rihll presents a third piece of evidence: Polyaenus’ brief report of a battle somewhere in Thessaly, in which Philip II’s Macedonian phalanx got “a real pasting” from Onomarchus’ stone throwers (60). Although there is no suggestion in the source that these were anything other than men throwing rocks from their vantage point, Rihll believes that “Polyainos’ evidence for catapults in Phokis [sic] complements the archaeological evidence of the towers” (61). But her startling conclusion, that “the Phokians or the Thessalians independently invented the torsion one-armed stone-thrower” (61), may as well have been drawn from a conjuror’s hat.

It is important for Rihll to quickly introduce the monagkon (one arm), later known as the onager, so that she can reveal “an apparently little-read paper” (E.P. Barker, “Palintonon and Euthutonon,” CQ 14 [1920] 82–86, cited by both Marsden and Garlan in their standard texts) to prove that the torsion catapult “resulted from the union of monagkon and gastraphetês” (78). (Artillery scholars will know that this was earlier suggested by Rudolf Schneider in his article on “Geschütze,” RE.) It is worth noting that, if Rihll insists on having three sorts of catapults at Carthago Nova (mentioned above), the most obvious third sort is this one-armed machine. But, having been suddenly revealed in chapter 3, the onager disappears again, only to resurface briefly in chapter 10, where the only description to survive, that of Ammianus Marcellinus, is dismissed as “nonsense [that] should not distract us” (246).

In the course of the book, Rihll tosses around several theories whose speculative nature is not always emphasized. For example, there is the theory that lead glandes (sling bullets) are actually small catapult shot, despite the fact that such missiles predate the catapult’s invention and are oddly shaped if they were intended to travel along a slider. The large quantity of glandes from Burnswark (Scotland), which are most probably the debris from a Roman siege (Britannia 34 [2003] 19–33), nevertheless leads her to “hazard a guess that the site was specifically for catapult training,” and furthermore “that some of the people wielding such catapults were riding horses” (222). It is true that Arrian describes Hadrian’s cavalry using “light spears and missiles, shot not from a bow but from a machine” (first noted in BJb 186 [1986] contra 309 n. 36), so the force besieging Burnswark may very well have included horsemen, but there is no particular reason to suggest it and no evidence to prove it. Like Rihll, I have often wondered precisely what Appian meant when he wrote that Sulla, besieging Piraeus in 87 B.C.E., killed many “with catapults shooting twenty of the heaviest molubdainas (lead lumps) together” (Mith. 34). She seeks the answer with glandes shot from a crossbow, and thus misses an opportunity to explore Philon’s incidental remark about deploying 20-mina stone projectors around harbors to throw “lead jars” (Pol. 3.56=95.14 Th.)

Such stone projectors provide another subject for speculation. Rihll claims that “their missiles are ruthlessly destructive to siege engines” when shooting along a steep trajectory, and that “from ground level they can lob stones over the wall from inside to out” (138). No ancient author ever describes this, and it would seem to be a particularly hazardous technique. Not only does it require absolute faith that the machine will not malfunction and damage the defenses over which it should be shooting; but, in addition, it depends upon a knowledge of indirect aiming. It is far more sensible to deploy stone projectors, as Philon suggests, in ground-level external batteries (Pol. 1.32=82.6–14 Th.), where their flat-trajectory shooting can wreak maximum damage. Philon does, in fact, mention deploying catapults inside a besieged town (Pol. 3.26=93.1–2 Th.), but these are intended for use if the enemy breaks in.

Rihll is also fond of threading a string of perhapses and maybes together to come up with a surprising new theory (which tends to evaporate when subjected to critical scrutiny). For example, Tacitus’ innocent conjunction of wagons and artillery blocking the road at Cremona in 69 C.E. raises the possibility that “the wagons were, presumably, for carrying the artillery.” But in Rihll’s hands, this becomes proof that the wagons were “actually carrying it during the engagement,” and suddenly “we appear to have catapults on wagons, in the field” (201). This would be an important development, presaging the carroballista, a mobile arrow shooter about which we know very little. But in actuality, all we have are army wagons probably transporting catapults, the larger ones in a disassembled state.

Following a similar thread of conjecture, Rihll deems it “quite plausible” that Apollodorus of Damascus invented the arrow shooter known as the cheiroballistra, because (a) machines very like it are depicted on Trajan’s Column, (b) Apollodorus was “superintendent of the column,” and (c) “architect and catapult builder were often roles performed by one and the same man” (212). This theory sits awkwardly with the statement, made only a few pages earlier while still discussing the events of 69 C.E., that “the all-metal-frame kheiroballistra was invented sometime around now” (202). It is surely somewhat inconvenient that Heron, whom Rihll accepts as the author of the treatise that describes the machine, flourished fully a generation before Apollodorus, her preferred inventor.

I have mentioned only a few areas of contention, and perceptive readers will find much to quibble about. The index of ancient sources cited in the text is welcome, but the general index gives uneven coverage (e.g., omitting key ancient authors and devices discussed in the text), and several modern sources cited in the notes have been omitted from the bibliography. Nor is the text entirely free of errors. Minor slips can be forgiven, such as referring to the followers of Vespasian as “the Vespasians” (200) or miscalculating the reigns of the Attalids (165), or even repeating the Loeb’s mistranslation of trispithamos (three-span) as “three-palm” (118). But errors, such as stating that “the Hatra frame was roughly 8 feet long” (226) when no frame survived, are of an altogether different order. All in all, it seems that Rihll’s seductive “discoveries” have been given priority over a simple study of the evidence. The reader is advised to tread warily.

Duncan B. Campbell
2 Oak Avenue
Glasgow G61 3HB
United Kingdom

Book Review of The Catapult: A History, by Tracey Rihll

Reviewed by Duncan B. Campbell

American Journal of Archaeology Vol. 112, No. 1 (January 2008)

Published online at

DOI: 10.3764/ajaonline1121.Campbell

Add new comment

Plain text

  • Use [fn]...[/fn] (or <fn>...</fn>) to insert automatically numbered footnotes.
  • No HTML tags allowed.
  • Typographic refinements will be added.
  • Web page addresses and e-mail addresses turn into links automatically.
  • Lines and paragraphs break automatically.
Click "Save" to submit your comment. Please allow some time for your post to be moderated.