In a recent lengthy article on what he calls ‘the Fantastic Three Thousand’, Moller proposed inter alia that an expeditionary Roman group during the reign of Nero (54–68 c.e.) received information about the African city of Meroë, located in Upper Nubia between the fifth and sixth cataracts of the Nile or approximately 200 km northeast of modern-day Khartoum, Sudan, having a military force of 3,000 elephants.Footnote ¹ Moller adds that this information, or rather disinformation, was designed by local inhabitants to deter Rome from further southwards expansion, even though a peace between Rome and the Meroitic kingdom had been in place since the late first century b.c.e. after a brief period of significant and mutually damaging conflict (Strab. 17.1.54).Footnote ² Indeed, Moller even imagines that local guides might have shown the two Roman centurions leading the expedition a set of fabricated elephant tracks, meaning that, in 61 c.e., ‘it is just possible that the kingdom of Meroë was able to ward off a Roman invasion using a substantial bit of bog and a few footprints in the sand’.Footnote ³ In short, Moller contends that the Romans of Nero’s day decided not to conquer Upper Nubia because they were afraid of the vast number of war elephants that they were told were in the possession of the area’s inhabitants.

This idea is based largely on a minority reading of 6.186 in Pliny the Elder’s Naturalis Historia; but, as will be argued herein, Moller’s speculation takes the literary evidence rather too far, more so given uncertainty about the actual text in question, and whether it even refers to elephants at all. What follows privileges the brief arguments previously brought to bear by Desanges on this topic,Footnote ⁴ but it is necessary to reiterate these arguments—and indeed extend and reinforce them—in the context of Moller’s conjecture, which puts forward an arguably simplistic ‘solution’ to what is likely a far more complex issue relating to defining Roman limites in the Principate. As a secondary consideration, the passage requires further treatment given that at least two influential editions of Pliny’s Naturalis Historia have chosen to follow what should be regarded as a reading that has no firm grounding in either the manuscript tradition or in context.

In Pliny’s discussion of Meroë, he introduces the Isle of Tados (insulam Tadu), located near Meroë, in the context of a report about the area made by a Roman scouting expedition during the reign of Nero (HN 6.184).Footnote ⁵ We read the following about the island at HN 6.186 in Mayhoff’s venerable Teubner edition: cetero cum potirentur rerum Aethiopes, insula ea magnae claritatis fuit. tradunt armatorum $\overline{CCL}$ dare solitam, artificum $\overline{III}$ .Footnote ⁶ In the Loeb translation, which is based on Mayhoff’s text, Rackham states: ‘Moreover at the time of the Ethiopic dominion this island was extremely celebrated. It is reported that it used to furnish 250,000 armed men and 3,000 artisans.’Footnote ⁷ Where precisely this island was located, and whether it could indeed house so many men is not of any great material concern to this discussion, even though Desanges thinks it actually refers to ‘l’île de Méroë’ and that 3,000 or even 4,000 artisans being located there is plausible.Footnote ⁸ In a note, however, Rackham wonders: ‘Perhaps the text should be altered to “elephants”’, citing Detlefsen’s conjecture in his critical edition of Book 6 that elephantum, the genitive plural of elephas/elephans, should be read for artificum or artificium at this locus.Footnote ⁹ Brodersen goes further and includes elephantum instead of artificum or artificium in his edition of the text.Footnote ¹⁰ Precisely why this reading should be considered is not stated, given that critical editions of the text have mostly elected to print artificum or artificium.Footnote ¹¹ In short, the manuscript tradition, which goes back to at least the ninth century, is adamant that artificum or a variation thereof should be provided.Footnote ¹² In their recent translation of the locus, Turner and Talbot elected to stick with the traditional interpretation: ‘3,000 craftsmen’.Footnote ¹³ Still, in Conte et al.’s edition of the text, the authors state, without any clear rationale, that ‘Il testo probabilmente è corrotto’ and that ‘Ci si aspetterebbe un termine come «elefanti»’, even though the reading of artificum is preserved in their main Latin text.Footnote ¹⁴ While there is no acknowledgement of Detlefsen’s views here, one certainly suspects the influence of an earlier commentator, which prompts us to consider why, exactly, such a reading would be preferred over the more traditional artificum or artificium, which Conte et al. translate as ‘tecnici militari’.Footnote ¹⁵

In short, the whole idea of HN 6.186 referring to elephants, as is embraced by Moller, can be traced back to Detlefsen. But where did he get that idea? Desanges proposed that ‘le savant allemand a succombé à la tentation de l’exotisme’, and this may well be so, but there are some clues to be found elsewhere in Pliny.Footnote ¹⁶ Perhaps his reading was prompted by HN 6.185, where Pliny states that the Neronian expedition reported seeing the tracks of rhinoceroses and elephants (rhinocerotum elephantorumque uestigia) near Meroë, where greener herbage was seen, together with some forest. Further association of what ancient writers called Aethiopia and elephants can be found in Book 8 of the Naturalis Historia (see HN 8.26 and 8.32).

But immediately after we read of ‘the Fantastic Three Thousand’, Pliny notes that the uniuersa gens of the region was once called Aetheria, and then Atlantia, but eventually came to be called Aethiops, after the son of Vulcan (Hephaestus) (HN 6.187). What Pliny refers to, here, is the dark skin colour of the local population, their dark countenances being equated to those of blacksmiths, with the very word Aethiops (Αἰθίοψ) seemingly being a compound of the Greek for ‘burning’ and ‘face’, thus ‘burnt-face’.Footnote ¹⁷ There is thus some kind of a connection here to the notion of the aforementioned island being replete with metal-working artisans, instead of elephants. In refuting Detlefsen’s conjecture, Desanges points to archaeological sources suggesting that iron-working was an important industry around Meroë.Footnote ¹⁸ Large slag heaps, both above and below the surface, are the principal evidence of Meroitic metal-working activity, with these important sites having been investigated by numerous archaeologists over the past few decades.Footnote ¹⁹ Of interest, too, is that a recent archaeological study has found a ‘lack of evidence for iron production from c. 300 bc–ad 300’, that is, the period of most relevance to this discussion, although the authors do note that ‘investigation of some of the smaller slagheaps might fill some of this chronological gap’.Footnote ²⁰

Despite the mentions of elephants in Book 6, it is not until the beginning of Book 8 of the Naturalis Historia that Pliny discusses elephants in detail, including their use in war. Yet he never connects ‘Aethiopia’ or ‘Meroë’ to the war elephant. Yes, he writes of the elephant’s presence in what he terms Aethiopia, but only as a source of food for the Trogodytae (HN 8.26). He again refers to elephants being found among the Aethiopes et Trogodytae at HN 8.32 and notes that, in Africa, these animals are captured using pitfalls (HN 8.24). For Pliny, war elephants are mainly associated with the Hellenistic powers and Carthage (see, for example, HN 8.11, 8.14, 8.16, 8.18), and eastern warfare in general, although the phrase ex parte orientis found at HN 8.27 could encompass Aethiopia.

By the period concerning us, the ancient Mediterranean world had long had an insatiable hunger for ivory, most of which had been acquired through trade rather than direct acquisition. Alexander the Great famously encountered war elephants in his campaigns, most notably at the Hydaspes in 326 b.c.e., and brought many Asian elephants back from his Indian campaign.Footnote ²¹ But Ptolemy II of Egypt (284–246 b.c.e.) was seemingly the first Hellenistic ruler to collect African elephants for military purposes, most likely because the rival Seleucid regime in Syria had cut off the supply of Asian elephants from the subcontinent. Ptolemaic expeditions to what is now modern Eritrea and surrounding areas are well documented.Footnote ²² Although Polybius (5.84–5) tells us that African war elephants performed poorly against the ‘Indian’ elephants of the Seleucid enemy at Raphia (217 b.c.e.), the Carthaginians continued to employ African elephants in warfare against Rome, right up to the battle of Zama (202 b.c.e.), where the Punic general Hannibal unsuccessfully used them against the well-prepared forces of Scipio Africanus.Footnote ²³ There has been some debate about the precise species of African elephant used in the classical world for military purposes, but this matter is largely immaterial to the present discussion.Footnote ²⁴

To return to Pliny, if he was truly aware of tales of a Meroitic army having no less than 3,000 war elephants when he wrote Book 6 in the mid first century c.e., why did he not mention such a thing in Book 8, where elephants are discussed at length? Arrian (Tact. 2.2) did puzzlingly write that the ancient Ethiopians used the elephant in warfare before the Macedonians and Carthaginians, and that, by his day (early second century c.e.), the use of war elephants had been abandoned for military purposes ‘except among the Indians or the upper Ethiopians’ (ὅτι μὴ παρ’ Ἰνδοῖς τυχὸν ἢ τοῖς ἄνω Αἰθίοψιν καὶ αὕτη ἐκλέλειπται) (Tact. 19.6), but it is impossible to tell whence he drew this information.Footnote ²⁵ Suspicion falls on the well-documented concept of a ‘confusion of Indias’, with considerable evidence existing to suggest that India and Ethiopia were often confused in antiquity, especially towards the period of the Late Empire.Footnote ²⁶

Likewise, Heliodorus, in his rather far-fetched fourth-century c.e. novel Aethiopica (9.15–16), has his Ethiopians using elephants in war. But one can probably put this down to an interest in exotica, or another case of a ‘confusion of Indias’, together with the fact that the opposing Sasanian Persians did use elephants—even if Asian rather than African—for military purposes around the time of composition.Footnote ²⁷ Moller’s view that ‘the details [from the Aethiopica] … highlight that the perception existed by the late antique Mediterranean world that elephants were used by the Meroites in war’ therefore seems to stretch the evidence, such as it is, a bit too far.Footnote ²⁸ Such literary instances could suggest that there was indeed a source claiming that earlier Upper Nubian armies had used elephants, but the notion that they ever used elephants for military purposes has never gained a great deal of traction, even if a reasonable number of depictions of elephants in Meroitic art have been found—perhaps tellingly with no clear evidence of a military application.Footnote ²⁹ That the important site of Musawwaret el Sufra, whose Meroitic name ‘Aborepi’ has been tentatively translated as ‘the resting place of the elephant’, is where several ancient depictions of elephants have been found does not necessarily indicate the presence or use of war elephants.Footnote ³⁰ Moreover, that Meroë was once able to put 3,000 elephants into the field is not only fantastic, but also nonsensical, even in the context of what is presumably outrageous hyperbole, especially when Rome’s great enemy Carthage was said, at its height, to have been able to house only 300 of the beasts and their fodder within the city walls! (see App. Pun. 95). The only ancient sources referring to (presumably exaggerated) corps of thousands rather than hundreds of war elephants refer to Indian kingdoms.Footnote ³¹ Moreover, one wonders about the logistical nightmares associated with stabling 3,000 ever-hungry elephants on a midriver island. That the local climate was becoming more arid ‘in the first centuries a.d.’, as a recent study has demonstrated, casts further doubt on the ability to feed such a force.Footnote ³² Finally, Desanges points out that Pliny, instead of using elephantum, would likely have used the genitive plural form of elephantorum (a form arguably less likely to be confused with artificum or artificium) with a number to signal the number of elephants in the possession of the ruler of Meroë, as is the case elsewhere in Book 6 with reference to Indian rulers at HN 6.67 (elephantorum $\overline{IV}$ ) and 6.68 (elephantorum $\overline{IX}$ ).Footnote ³³

In sum, it seems highly questionable to overturn the generally accepted reading of artificum or artificium in place of what some have imagined—arguably incorrectly—as a more contextually relevant elephantum. If that is the case, Moller’s view that Pliny, at HN 6.186, was referring to contemporary Meroitic stories of a vast and still extant elephant corps as a means to deter the Romans from further territorial expansion does not ring true, more so given that Pliny—and indeed his modern translators—clearly places such a force in the past with tradunt … dare solitam, and then switches to hodie immediately thereafter to indicate the current state of affairs. In short, the informants, presumably still the members of Nero’s expedition, related that the city was accustomed to provide 250,000 armed men and 3,000 artisans (or elephants!), with solitam referring to the earlier insula. From this, it might be imagined that Pliny thought that the 3,000 elephants—if elephants they were—were long gone by Nero’s day. While Moller’s view that such an elephant force is ‘fantastic’ is almost certainly correct given the lack of any genuinely convincing corroborating evidence for large-scale elephant warfare in the region in question, there is good cause to doubt that Pliny’s words even support this proposition in the first place. Let us conclude with the following: it is true that the Romans, who had substantial experience at defeating elephant-equipped armies in any case over a couple of centuries, did not make a concerted effort to attempt to annex the land supposedly protected by ‘the Fantastic Three Thousand’, but explanations aside from local tricksters telling tall elephant tales and creating spurious elephant footprints in the dirt are clearly required to understand their decision to forego any real imperialistic intent in the broader area of Upper Nubia.Footnote ³⁴