CHAPTER 17
DONALD ENGELS
THE need to maintain an army’s material welfare is an essential aspect of successful military leadership. The provision of food and supplies to men and animals is critical for an army’s morale and combat capability, whether on campaign or stationed along a frontier. It is a basic element of strategic and tactical success.
In this chapter, the basic literature on ancient logistics will be reviewed for both the Greeks and the Romans, noting opportunities for further research. Secondly, the significant limitations of ancient overland transport of food and water will be considered, as revealed by our primary sources and recent research. Next, the responses of military commanders, especially Alexander the Great, to these limitations will be examined. In this context, the value of accurate military intelligence will also be discussed. Finally, we will examine two actual cases, where terrain, climate, weather, troop numbers, and the capabilities of land, sea, and river transport had a major impact on overall strategy. The first case will be Alexander the Great’s crossing of the Gedrosian Desert in southern Pakistan and Iran in the summer of 325 B.C. The second case is the logistic considerations that influenced the establishment of the Roman frontier along the Rhine River in Germany during the first century A.D. Since the proper storage and protection of supplies once they reach a military base is also important, a brief section on this topic will end the chapter.
LOGISTIC STUDIES
There has been a welcome increase in the study of the logistic factors that influenced the tactics and strategy of Greek and Roman armies, but there still remains much that can be done. For the Greek world, while no specific work has been devoted to the topic, many recent books on military history have included logistic considerations for ancient campaigns. Arther Ferrill (Ferrill 1985: 82, 112, 122, 146; cf. Farrokh 2007) has noted that the logistics of the Persian army has been underrated and little studied. The establishment of a vast empire stretching from Macedonia in the west to the Indus River in the east could not be accomplished without considerable logistic expertise.
The provisioning of Xerxes’s army during his invasion of Greece in 480/79 B.C. was one of the great achievements of logistical organization in any era. Persian planning, preparation, organization, and distribution of supplies, as well as the march itself from Sardis in April of 480 to Athens in September, are all impressive feats. The work of F. Maurice (Maurice 1930) remains an excellent basis for analysis. The estimates of the numbers in the expedition vary widely, but whether the numbers used are conservative or expansive, the sheer magnitude of the Persian army and navy’s food requirements was staggering. Maurice estimated that there were about 180,000 infantry. In addition, a conservative estimate for the numbers of servants and followers would be at least 20,000, and perhaps 75,000 pack animals and cavalry horses. J. B. Bury and R. Meiggs (Bury and Meiggs 1975: 169), who also accept Maurice’s numbers for the infantry, suggest that the fleet may have numbered some eight hundred triremes with approximately two hundred sailors and marines per vessel, or another 160,000 personnel. So the total number of personnel may have been around 360,000. Herodotus’s numbers are far higher. B. Strauss (Strauss 2004: 269) is right in stating that Herodotus was an honest and accurate historian. But he may not have always been correct, because his sources, the Greek veterans and their families, whose stories he faithfully recorded, may have exaggerated their opponent’s numbers, as veterans of many wars have tended to do. All these individuals and animals would need to be fed for approximately six months, from April to September of 480. The total grain requirement alone (not counting fodder or water) for the six-month period for personnel and animals would be around 164,700 tons.
This is a huge amount of supplies for a preindustrial state to organize, collect, and distribute throughout the route. All this had to be done before the army advanced from Sardis and the navy from Ionian ports. Indeed, one of the reasons it took four years to launch the expedition was all the advance planning, preparation, organization, and collection necessary to distribute the provisions. Also, it may have taken the surpluses of four years’ harvests from areas of the Persian Empire accessible to sea transport to feed the expedition. Provisions from these regions would be brought to the campaigning area. One thinks especially of Egypt.
One may also suggest that a shortage of supplies contributed to the Greek decision for a quick, decisive battle at Salamis. The Greeks were cut off from resupply from the mainland which was controlled by the Persian army all the way south to the Megarid. They were also cut off from the Saronic Gulf and the Aegean by elements of the Persian fleet which blockaded both ends of the Bay of Eleusis.
Likewise, the Greeks were deprived of their water supply when the Persians choked up the spring of Gargaphia before the Battle of Plataea in August of 479. The Persians also occupied the main route of resupply over Mount Cithaeron and the Gulf of Corinth, now under Greek control. The only way the Greek army could be resupplied was over mountain passes across Cithaeron from the gulf. This would be extremely difficult, especially for water. This situation probably contributed to their decision to stand and fight. The Persians themselves were probably running low on food, now that they were cut off from resupply by sea. Both facts influenced the development of the battle.
The effects of hoplite warfare on the agricultural fields of the combatants during wartime are discussed by V. D. Hanson (Hanson 1989: 33–5, 2005: 35–64). Ancient references to the destruction of an enemy’s fields were probably exaggerated. Such tasks would have involved huge expenditure of labor and would not have been thorough or effective.
The logistical limitations of naval operations have also been elucidated by V. D. Hanson and L. Casson (Hanson 2005: 258–62, 283; Casson 1959: 102–3, 1994: 70–3). Warships were very limited in their carrying capacity, especially water, the daily ration being about two gallons, or twenty pounds. With about two hundred rowers and marines per vessel, this was a considerable weight. Food provisions needed to be provided every few days, usually a fifty-mile distance. Many naval operations were impeded or abandoned from a lack of advance preparation in obtaining supplies.
Recent research has further illuminated the logistic problems confronted by Alexander the Great during his Asian campaigns, especially in India. Of special interest is the colossal task of moving hundreds of vessels in the Macedonian fleet across the rivers of the Punjab during the monsoon season of 326 B.C. The vessels had to be dismantled, carried by ox carts at two miles per hour, first from the Indus River to Taxila, a distance of some forty miles, then to the Hydaspes (Jhelum), another 120 miles, and presumably further east to the Chenab and Beas. Each time the fleet was dismantled, transported overland by oxen, and then reassembled at the next river (Bosworth 1996: 12–13; Heckel 1992: 58–64). This problem needs further analysis and may help explain Alexander’s decision to stop his eastward progress and return south down the Indian rivers to the ocean. His men (as well as the animals) may have suffered significantly from the onerous burden of transporting the boats, and it may have made them refuse to march farther.
There were major military campaigns undertaken in the Hellenistic era. One notes the campaigns of Antiochus III to restore the borders of the Seleucid Empire in the east. However, there is little study specifically on logistics, as the original sources are poor (Bar-Kochva 1976). For the Roman army, there is a vast literature on logistics (Erdkamp 1998; Goldsworthy 1996, 2003). The most comprehensive study is by J. Roth (Roth 1999). Every aspect of army logistics is considered in the context of Roman military campaigns from 264 B.C. to A.D. 235. For Hannibal’s campaigns, we have the important works of J. Peddie (2005), and A. Goldsworthy (2000), who give the most comprehensive study of his formidable logistic problems. Admirable analyses of the logistic difficulties confronting Julius Caesar’s campaigns in Gaul have been undertaken by Goldsworthy (2006) and Labisch (1975). Logistic problems are considered in the context of intelligence gathering and diplomacy. Like Alexander, Caesar appreciated the value of good military intelligence.
The logistic problems confronted by the Romans during the Empire have received extensive study. Supplying the frontier army along the southern border of North Africa presented formidable problems which have been thoroughly analyzed by I. D. Taylor (1997). The frontier itself was some 1,600 miles long and was the responsibility of one legion, the III Augusta, some five to six thousand men stationed at Lambaesis in Numidia. Approximately thirty auxiliary units may have increased the total strength along the southern border to thirty thousand.
The auxiliaries were mainly cavalry units, cohortes equitatae and alae. A cohors equitata consisted of approximately one thousand men and an ala around five hundred cavalry. These small cavalry units were ideal for patroling such vast regions and could quickly respond to danger. In this region, the limits of Roman power were not so much defined by political boundaries, but by the agricultural limits of the area, climate, and geography. Roman forts were situated in the most fertile agricultural districts, since the supplies were acquired locally. Finally, J. P. Adams has elucidated the campaigns of Trajan, Septimius Severus, and other commanders undertaken by the Romans in the Near East during the first three centuries A.D. (Adams 1976).
LOGISTIC PRINCIPLES AND MILITARY INTELLIGENCE
In this section we will consider the limitations of overland transport of food and supplies in antiquity, and, actually, before the age of steam. The net amount of food carried by a pack animal (horses, donkeys, mules, camels, oxen, minus their packs) varied under different circumstances. The type of animal, the weight of the pack saddle, the length of time the animal was expected to travel per day, the length of the total campaign the animal would be expected to travel, speed and rest periods for the army, terrain, climate, and presence or absence of water along the way, all contributed to the army’s capacity to carry food and supplies. The type of wagons used (if any), the harnessing techniques for different animals, and the variety of animals, also contributed greatly to the ability of an ancient army to feed itself while on campaign (Engels 1978: 11–25, 119–30).
The amount of food required by the force, pack animals, cavalry horses, and the men and followers, was also an important logistic factor. The amount of food and water found along the route, and the amount that needed to be carried will also have been significant. If men and animals are not fed properly, they will be unable to perform their tasks effectively, rendering them inadequate if not useless.
Because of the limitations of overland transport in antiquity, most provisions had to be obtained from a limited region through which the army moved. Perhaps we can understand these limitations best by considering the following hypothetical examples.
First, let us calculate that a pack horse can carry two hundred pounds of food net. This excludes the weight of his pack, which often weighed around fifty pounds. Let us further assume that the horse had to walk the entire day with a few breaks for rest and adjusting the saddle. We also assume that the animal consumes twenty pounds of food per day, ten pounds of grain and ten pounds of fodder, and that this food must also be carried by the animal as it is unavailable along his route. This would often have been the case in many parts of the Near East and North Africa. However, the animal’s water ration of ten gallons per day, or eighty pounds, can be obtained along the route.
Let us assume that one thousand pounds of food needs to be moved by packhorses for a one day’s journey or about twelve to fourteen miles depending on terrain and climate. Six animals would be needed for the trip. They could carry twelve hundred pounds, but they would consume one hundred and twenty pounds on the journey. This would leave 1,080 pounds for delivery.
Now let us consider a nine days’ journey to deliver the thousand pounds. Here fifty animals would be needed to carry the load. During the trip, the fifty horses would consume 180 pounds apiece or nine thousand pounds. This would leave them with twenty pounds of food apiece or one thousand pounds total when the trip was over. The food needed for the horses, plus the one thousand pounds of food that was delivered, would weigh ten thousand pounds total. As the examples illustrate, it would be difficult to transport food under these circumstances for more than five days as the horses would consume more food than they could carry. If water needed to be carried in addition, the range would have been considerably shorter, a day or two at the most.
The supply of an army at a fixed location presented even greater limitations. Here the pack animals would have to move from the source of supply of the food to the point of distribution and back again. If the animals were abandoned or starved at the point of food distribution, a valuable resource would be wasted and future transport needs would not be met. To supply a point at a one-day radius (about thirteen miles) with one thousand pounds of food would require seven animals; each would consume forty pounds of food on its two-day round trip, and have one hundred and sixty pounds left at the point of delivery. They would deliver 1,120 pounds altogether. A four-day radius (about fifty-two miles, eight days’ round trip) would require twenty-five horses. Each horse would consume 160 pounds on the round trip and deliver forty pounds at the distribution point, or one thousand pounds altogether. The twenty-five animals would have consumed four thousand pounds in order to deliver one thousand pounds of food. A trip of five days’ radius, ten days’ round trip would have been impossible. Once again, in actual practice, the radius of resupply would be limited to around two-and-a-half-days’ journey, since on a longer trip the animal would consume more food than it would deliver. Thus, large amounts of food could not be carried for long distances overland. Supplies had to be gathered locally wherever possible
The most important factor affecting an army’s logistic capability was the total number of men, pack animals, cavalry horses, and followers. A small force of a few hundred may rapidly march through an arid region with little problem, gathering what supplies were necessary along the way. A force of fifty thousand might suffer a catastrophic loss of life crossing the same area unless advance preparations were made.
In recognition of the limitations discussed above, many capable commanders, including Philip II of Macedon, Alexander the Great, and Gaius Marius, all required their soldiers to carry the maximum amount of arms, food, and other supplies that they were able. Indeed the latter’s Roman soldiers were called “Marius’s Mules.” This, of course, would limit the amount of pack animals, wagons, and servants needed by the army and thereby reduce its food consumption. Since smaller forces are faster and more maneuverable than larger ones, these factors would significantly increase the element of surprise.
Another way to save food was to limit the use of wagons. Wagons were banned by Philip II and Alexander although they still made intermittent appearances. The harnessing techniques used in antiquity did not effectively utilize the horse’s pulling potential. But occasionally they were necessary. An army with fewer wagons would also be more maneuverable than ones that were restricted to the roads, which were often poor, before the Roman Imperial era.
Because of these limitations, commanders would try, if at all possible, to transport provisions by sea or river. Oceangoing ships and riverboats had a far greater carrying capacity than pack animals, and the amount of food necessary to feed the crews was a small fraction of their carrying capacity.
But often such transport was not available and so the provision of supplies had to be arranged in advance with local officials. Thus, diplomacy and alliances were often needed before the army could move into new territory. In recalcitrant areas, hostages would be taken and garrisons established before the army moved through to insure that the supplies would be provided. If the inhabitants of a region had not surrendered or made an alliance with the commander before he entered their territory, this would require that the army would split up into smaller units and traverse the territory. Or a light force would strike out from a base well stocked with provisions where the majority of the troops were stationed. Under such circumstances, supplies would not be provided by markets, gifts, or requisitions, but by pillage, plunder, and foraging units.
The army would also have to await the harvests before it moved. The limited productive capacity of ancient agriculture compared to today (although this should not be exaggerated) meant that adequate food would only be available in the weeks or months after the harvest dates. This remained true up to World War I.
Above all, the army itself or units of the army would have to keep moving; if they remained stationary for very long they would soon strip bare the thirty-mile radius of resupply for a stationary force. But on occasion the army would have to cease campaigning and enter its winter quarters. Often this was not so much to avoid harsh weather but to await the harvests which occurred in the spring, May or June in the Mediterranean region.
Under these circumstances, commanders would choose regions that could be supplied by river or sea transport for their winter quarters. If such facilities were not available, then the army would be split up into small units that would only make modest demands on the local supplies. Any halt of more than a few days would require the same procedures.
Ancient commanders were well aware of these facts and so the collection of military intelligence about the road ahead was of paramount importance (see further Russell, 474–92). Because of the limitations of overland transport, wherever possible the collection of supplies had to be arranged before the army moved through a region. They could not be carried long distances from place to place. Alexander the Great and others, including Hannibal and Julius Caesar, collected vast amounts of strategic intelligence before embarking on a campaign. Alexander collected strategic intelligence about the Persian Empire, the lengths of its routes, terrain, climate, attitudes toward Persian rule in various regions, before he left Macedonia (Engels 1980: 327–40). This information was essential for overall logistic planning. In Alexander’s case, all military intelligence and counterintelligence was directly under his control. As Napoleon said, “A general who has to see things through other people’s eyes will never be able to command an army as it should be commanded” (Engels 1980: 33). This practice enabled Alexander to obtain the information he wanted directly without intermediaries, except interpreters when necessary. No important tactical decision was made by any capable commander without advance intelligence. Indeed, the methods used by Alexander and Caesar in the collection and use of intelligence and conduct of logistical operations were very similar (Goldsworthy 2006: 213–18). The use of military intelligence by Germanicus in his campaigns in Germany from A.D. 14–16, is also worthy of note. Notable too is the absence of such intelligence in the disastrous expedition of Varus in Germany in A.D. 9.
If a region had surrendered in advance to Alexander, both tactical intelligence and supplies were easily available. Guides, usually of the highest rank, led his army along the route. They would also serve as hostages for their own good behavior. Often more than one guide would be used to check the other’s veracity and reliability.
If a region had not surrendered in advance, the collection of tactical intelligence was complicated. In Alexander’s case, guides might be obtained from neighboring regions that were hostile to the province he was crossing. This may have enhanced, but not proved, their veracity. Often such individuals were not available and locals had to be seized. Sometimes several locals would be used to independently verify the information of their colleagues. Failure to provide accurate information often had fatal consequences. Sometimes family members would be taken as hostages to assure accuracy. Until guides could be captured, mounted skirmishers would precede the army at a considerable distance. Even after the guides were found, mounted scouts and skirmishers were still sent forward from the main army to check the accuracy of the guides’ reports. Deserters and prisoners could also be used, although they might be counterintelligence agents.
When no guides of any sort could be found, and this rarely happened, Alexander organized reconnaissance operations. This occurred in the Gedrosian Desert in the summer of 325. Intelligence failure played a role in this disaster, a failure related to the wind direction in the region, important for the sailing ships which transported the supplies, but not the nature of the terrain.
THE GEDROSIAN EXPEDITION
The greatest enigma in Alexander’s campaign will always remain the disaster that occurred in the Gedrosian (Makran) Desert in the summer of 325. A large part of Alexander’s army did not cross the desert but marched west through India and Afghanistan. The remainder, that made the westward crossing of the Makran desert in southern Pakistan and Iran, seems to have numbered between thirty thousand and one hundred thousand infantry and cavalry, plus an indeterminate but large number of followers. His fleet may have had an additional eight thousand personnel. For the force that would cross the desert, Alexander had collected four months’ supplies before he embarked westward in mid-July 325. The land army began its march west along the coast of the desert, and later it moved inland. The fleet left its harbor in late October or early November. About the same time, Alexander had completed his march through the desert. He reached Pura, the Gedrosian capital (somewhere in the Bampur basin) around late October and the Carmanian capital, north of the Straits of Hormuz around mid-December. Even the apologetic Plutarch, never prone to exaggerate Alexander’s failings, wrote that three-quarters of his army was destroyed by starvation and dehydration in the crossing (Plut. Alex. 66.4–5). Arrian wrote (Anab. 6.25.3), “and so a few, out of many, were saved.” Like Plutarch, Arrian was an apologist and therefore would have no reason to exaggerate the losses suffered on this campaign. These losses must have been generally known or they probably would have been suppressed or underreported by these authors. What went wrong? Or did things turn out as Alexander had planned? Was he deluded about the harshness of the desert because he wanted to emulate Semiramis’s and Cyrus’s legendary crossings centuries earlier? Was he indifferent to the loss of life?
There are two major logistical questions relating to this campaign. The first relates to the collection of four months’ worth of supplies for the army, to be used primarily while it crossed the desert (Arr. Anab. 6.20.5). The term is stratia, army; not nautikon, fleet. The fleet did indeed end up eating much of the supplies during its delay of three months but Arrian says they were intended for the army. Why did the army starve when so much was collected, and how was it supposed to be carried and distributed?
The second problem is the advance of the land army in mid-July while the fleet was held up until late October or early November by the southwest monsoon winds (Engels 1978: 110–18, 135–43; Bosworth 1988: 139–46, 1996: 166–85). Why did the land army advance west some three months ahead of the fleet, which was probably supposed to help feed the land army along the route? As it was, the land army completed its march by mid-October, shortly before the fleet began to sail west in late October.
I have reconstructed these events as follows. The fleet was to accompany the land army on its march along the coast of the Gedrosia. This because the four months’ supplies for a force of some one hundred thousand (Arr. Ind. 19.5; Curt. 8.5.4) would have weighed in the vicinity of 52,000 tons. This was carried by the fleet and was to feed the land army as well as itself on the trip. Such simultaneous coordination between fleet and land army was common not only in Alexander’s campaigns in the Mediterranean, but also for the Persian army of Xerxes in his invasion of Greece in 480/79 B.C. The fleet’s cargo ships with their great carrying capacity were supposed to supply the army along the coast with provisions. Meanwhile, Alexander’s army was supposed to dig wells to supply water for itself and the fleet. There would be limited water available in the Gedrosia at this time of year since the monsoon winds would cause rains, however brief, in the region. Indeed, one sudden downpour washed away a considerable number of Alexander’s followers and baggage when they encamped in a seemingly dry riverbed which flooded. The supplies had to be carried by the ships and not the land army because of the limited range of overland transport of food by animals and men that were eating the food as they traveled along as noted above. The carrying of approximately fifty-two thousand tons of food on pack animals that may have carried two hundred pounds net on all-day marches with fifty-pound packsaddles would have required an impossible number of animals. The pack animals would eat approximately twenty pounds of food per day, so their range would only be ten days at most or 100 to 150 miles under desert conditions. If Alexander had only thirty thousand troops with him plus followers, the range would still be the same. However, the fleet was held up by the southwest monsoon winds (Engels 1978: 110–17, 135–43).
A. B. Bosworth has proposed that the four months’ supplies were only to feed the fleet. Alexander had about thirty thousand soldiers with him on land which would make fewer demands on the desolate countryside than a force of one hundred thousand. Alexander was informed by the natives before the army or the fleet moved west, that the monsoon winds would prevent the sailing of the fleet until late October (Arr. Anab. 6.21.2). So Alexander did not plan to have the fleet feed the army along the way, but, instead, the land army was supposed to have fed the fleet with whatever supplies it could muster from the Gedrosia (Arr. Anab. 6.24.3). The losses along the route were heavy, but not as catastrophic as Plutarch (Alex. 66.4–5) recounted (although Bosworth believes that only about nine thousand of the original thirty thousand survived, which is indeed a loss rate of 70 percent, or 21,000 of 30,000). Thus, Alexander blundered in failing to appreciate the difficulties of the march, because he wanted to emulate Cyrus the Great, but the death toll was not as drastic as it would have been for a larger force (Bosworth 1988: 139–46, 1996: 177–85). This is indeed the basic account of Nearchus the Cretan.
The major problem in interpreting these events lies in the nature of our basic primary source, Nearchus, the admiral of Alexander’s fleet. Nearchus was an eyewitness to events and had a prominent role in the expedition. He must be taken seriously. He was the main source used by Arrian, our best surviving source for these events, writing in the second century A.D. Yet, despite his prominent position, there are serious questions about his veracity that have been raised by many, including Ernst Badian and myself on completely independent grounds (Badian 1975). Here we will note some serious problems with his facts concerning this expedition; I have recounted others elsewhere (Engels 1978: 115, n.83, 141).
I wish to offer some further thoughts about the veracity of Nearchus and the causes for the failure of the Gedrosian expedition. The first problem is his statement that it was the responsibility of the land army, traveling through a desert, to provide food provisions (not only water) for the fleet (Arr. Anab. 6.24.3). This would have been difficult indeed to achieve when the army itself was starving to death. Nevertheless, he does inadvertently provide the only clue that indeed the expedition was to be a coordinated one between the fleet and the land army: “It was then on this account, and also that, being close to the fleet he might supply it with provisions that, according to Nearchus, he chose this route [along the coast of the Gedrosia]” (Arr. Anab. 6.24.3; cf. Strab. 15.2.4). References to this simultaneous coordination do not occur elsewhere in his account.
A second major problem in Nearchus’s account, noted by Arrian himself (Anab. 6.24.2), was that only Nearchus, of all the sources Arrian knew, wrote that Alexander had no knowledge of the difficulty of the route, but wanted to imitate Semiramis and Cyrus. All the rest wrote that Alexander well knew what the desert would do to his army if he was unprepared for the crossing. It suggests that the massive loss of life (pace Bosworth) recorded in all our sources was entirely the fault of the king, and that he, Nearchus, played no role in the debacle. If Nearchus was correct, this would be one of the very few times in Alexander’s campaigns that he failed to collect adequate intelligence about the route ahead (Engels 1980).
The statement becomes all the more improbable when seen from the context of Nearchus’s own description of the preparations for the Gedrosian expedition in Arrian’s Indica (20.7–9). According to Nearchus, he begged the king to give him a position of responsibility for the expedition. Alexander was at first unwilling to risk the life of his old friend, but Nearchus persuaded him,
till at length Alexander accepted Nearchus’ willing spirit, and appointed him admiral of the entire fleet, on which the part of the army which was detailed to sail on this voyage and the crews felt easier in mind, being sure that Alexander would never have exposed Nearchus to obvious danger unless they also were to come through safe. Then the splendor of the whole preparations and the smart equipment of the ships, and the outstanding enthusiasm of the commanders of the triremes about the different services and the crews had uplifted even those who a short while ago were hesitating, both to bravery and to higher hopes about the whole affair.
This does not sound like the efforts of a leader who has no knowledge of, and made no preparation for, the difficulty ahead. Certainly his experienced commanders were willing to trust him.
In this context we can examine the statement in Arrian, probably based on Nearchus (Anab. 6.21.1–2, cf. Ind. 20.1):
The season, however, was not suitable for sailing; for the trade winds (etesioi anemoi) were blowing continuously, which in that season blow not, as with us, from the north, but from the ocean and from the south (noton). But from the beginning of winter, right from the setting of the Pleiades to the winter solstice, it was reported that the ocean here was fit for navigation.
The statement itself occurs before the time the land army left for its march west. It is factually correct about the sailing season lasting from late October to late December. There are problems however. The statement has Alexander know ahead of time, “it was reported,” that the fleet could not move until late October. This contradicts Nearchus’s statement in Arrian (Anab. 6.24.3) cited above, that Alexander planned for the fleet to be close to the land army so the latter could supply the former (!). How could this have happened if Alexander knew ahead of time that the fleet would be delayed until late October?
Two passages from Arrian (Anab. 6.24.2, 6.21.1–2) derived from Nearchus completely absolve him of any role in the disaster. It was all Alexander’s fault for being delusional and not knowing the difficulties of the land route. Nearchus himself could not have coordinated the fleet with the land army as was planned because of the adverse winds, which Alexander knew ahead of time would delay the fleet (despite 6.24.3: Alexander’s planned simultaneous coordination). It was not Nearchus’s fault that the winds blew from the wrong direction and prevented him from sailing. His fault lies in not telling the facts about Alexander’s plans and his gathering of accurate military intelligence about the desert, in order to absolve himself. In short, Nearchus’s account here is self-serving, contradictory, and probably mendacious.
What of the four-months’ supply of food for the army (Arr. Anab. 6.20.5)? It is never mentioned again in Arrian’s (and presumably Nearchus’s) account. Again, it never could have been carried by the army whether it had thirty thousand or one hundred thousand troops as it would run out in ten days.
The intelligence failure here was Alexander’s lack of knowledge about the direction and duration of the monsoon winds which prevented the fleet from sailing until late October. But there is a final enigma here, and that is the wind direction recorded by Arrian, presumably based on Nearchus once again. He records that it blew from the south, notos, and not the southwest in mid-July. Zephuros is the word for west wind; to have been accurate, both words should have been used for the direction. It may be wrong to press the point too much. The wind direction was perhaps recorded only approximately; exactness may not have been considered important.
Nevertheless, the real distinction between a south or a southwest wind was of critical importance for the actual sailing of vessels in the Indian Ocean from July to October. If this recorded wind direction was correct, it could have been responsible for the land army marching out into the desert without the supporting fleet. If the wind was indeed blowing from the south (from 180 degrees on the compass) in mid-July of 325, then the fleet could have sailed westward with no problem. With the rigs and square sails in use at the time, the sails could be positioned at a 45-degree angle, facing the southeast, and made a good run to the west. The vessel itself would be sailing 90 degrees from the south wind direction, due west, or 270 degrees on the compass. A sailing vessel whose sail is set on a 45-degree angle to the wind is on a reach (Casson 1959: 115, 220–1).
If, when Alexander left in mid-July, the wind was indeed blowing from the south, then the fleet could have accompanied the army along the coast on its march. But, if the wind later shifted to the southwest (225 degrees), any westward movement of the ships would have been blocked. The ships would have been forced to tack to make any headway westward.
Tacking is a zigzag motion that sailing vessels make to sail upwind. Because of the square sails on ancient sailing vessels, they could not sail closer than seven points from the direction the wind was blowing (Casson 1959: 220). A point is 11¼ degrees on the compass; seven points then would be about (not exactly) 78 degrees. With a wind from the southwest, or 225 degrees on the compass, this means that the vessel could be no closer than 78 degrees from this direction on both the starboard tack and the port tack. A starboard tack means the wind would be on the right-hand side of the vessel as it proceeded and a port tack means the wind is on the left side of the vessel (Maloney 1987: 171–89, 336–8). A 78-degree angle means that on the starboard tack, the vessel would go in a southeast-by-south direction, or slightly south of southeast. On a compass, this would be 145 degrees. This means that the vessel would proceed to the southeast, away from the required direction of west. On the port tack, the vessel would move northwest by west or 305 degrees on the compass, or slightly west of northwest. So the ships would go slightly forward (northwest by west) on one tack and then go slightly backward (southeast by south) on the next, and repeat over and over. Any forward movement would be extremely difficult.
This zigzag motion would add considerable distance to the voyage. Depending on the length of the tacks, it could easily double or triple the length of the voyage. The length of time for the voyage would be greatly lengthened, as tacking would slow the vessels considerably, so the maneuvers would add far more than two or three times the time of the direct voyage. This would drastically reduce any supply of food and water on the fleet. Furthermore, on the port tack, with the ships heading northwest by west toward the shore, the vessels would only intermittently come close to the coast. The nearby coast could be rocky and dangerous or smooth and level. There were no landmarks known to the sailors as this was uncharted territory for them. The locations of landfall for the fleet would depend on the lengths of their tacks. Under such circumstances, any coordination with the land force would have been impossible. The fleet might encounter a rocky shore in heavy weather and be smashed to pieces. The best plan would be to wait until late October.
LOGISTICS OF THE ROMAN FRONTIER IN GERMANIA
Recent research by J. Connolly has improved our understanding of the establishment of this frontier (Connolly 2006; cf. Wells: 2003). The strategic and tactical reasons for the decision to establish the frontier at the Rhine rather than the Elbe have often been discussed. However, the logistic problems of supplying the army east of the Rhine, in such a remote region, have not been fully appreciated.
The first problem is the geography and terrain which has changed significantly over the last two thousand years (see also Hughes, 128–39). The Germany that we see now, “the great green garden” scarcely existed before 1750 (Blackbourn 2006a). A modern visitor would be astounded at how different the “natural” landscape looked in prior centuries:
Much less of it was cultivated, much more of it dominated by sand or scrub, and especially by water. The twentieth century visitor would not need to journey far before stumbling upon ponds and lakes long drained and forgotten, or marshes that contemporaries likened to the wetlands of the New World, even to Amazonia. Filled with snaking channels half-hidden by overhanging lianas, these dwelling places of mosquitoes, frogs, fish, wild boar, and wolves not only looked but smelled and sounded quite different from the open landscape of canals and manicured fields familiar to twentieth century Germans (Blackbourn 2006b: 2).
The great rivers have been transformed the most. The premodern rivers meandered over floodplains or wound through hundreds of separate channels broken by sandbars, gravel banks, and islands. They ran fast or slow depending on the season and were not adapted to year-round navigation. Along either side were not engineered embankments but dense wetland and forests.
Vast expanses of peat moor were largely untouched. These rivers, flowing slowly in the winter months, would be prone to freezing, as the Rhine did in the winter of A.D. 406/7.
The historian Tacitus describes the terrain as thickly forested, penetrated by little-known forest passes. There were also swamps and quagmires where roads, presumably of logs, constructed by the Romans would quickly deteriorate. In A.D. 9, Quintilius Varus’s troops frequently experienced problems with terrain and supply (Cass. Dio 56.19–22, cf. 54.33). During the invasion of the country by Germanicus in A.D. 15, his men “were being pushed into a swamp, well-known to the victorious assailants, perilous to men unacquainted with it …” (Tac. Ann. 1.63). Later
Caecina, who commanded a division of his own, was advised, although he was returning by a route which he knew to pass Long Bridges with all possible speed. This was a narrow road amid vast swamps, which had formerly been constructed by Lucius Domitius; “on every side were quagmires of thick, clinging mud, or perilous with streams … Everything alike was unfavorable to the Romans, the place with its deep swamps, insecure to the foot and slippery as one advanced, limbs burdened with coats of mail, and the impossibility of aiming their javelins amid the water.” (Tac. Ann. 1.63–64; cf. Cass. Dio 54, 56)
Tacitus’s descriptions have been confirmed by recent ecological research analyzed by Blackbourn and others.
When the Romans attempted to invade Germany from the Rhine River (Rhenus) eastwards, they generally used overland routes and did not attempt to bring supplies or troops east or westward by river. There are three major rivers that flow east to west and empty into the Rhine. These are the Lippe (Luppia or Lupia), Ruhr, and the Main (Moenus). The Lippe was the major river east of the Rhine. The legionary fortress of Xanten (Vetera) was at the confluence of the Rhine and Lippe and another legionary fortress, Haltern, was eleven miles east upriver. These rivers did not penetrate very far to the east and would be subject to the problems of the terrain discussed above and below. So the rivers were not generally used for transport, either for food or soldiers, and the forces attempted to proceed east and westward over land instead. The commanders Drusus in 11 B.C., Varus in A.D. 9, and Germanicus in A.D. 14–16, preferred to leave the fleets on the Rhine, Elbe (Albis), or the North Sea coast and have the armies march inland without naval support.
Drusus’s campaign of 11 B.C., in the region of the Lippe and Wesser rivers, ran low on supplies, something unexpected if they could be brought in by river (Cass. Dio 54.33). In Varus’s fatal expedition of A.D. 9 into the Teutoburg Forest (Kalkriese), near the Ems, the army and its supplies were marched overland far from the rivers (Cass. Dio 56.19–21; Wells 2003: 45–55). The general Caecina proceeded west overland to return to winter quarters in A.D. 15. Although he passed swamps, quagmires, and woods, a river is not mentioned, and they were short of supplies (Tac. Ann. 2.63–68). In Germanicus’s campaign of A.D. 16, the Roman fleet was left at the mouth of the Ems, while his forces marched up the river valley by themselves (Tac. Ann. 2.8). Only once in Germanicus’s campaign of A.D. 15 was the army transported by the fleet up a river, the Ems.
Another possible supply route for campaigns in Germania was the coastal land route along the North Sea. In A.D. 15, two legions were disembarked to march eastward along the coast to the Weser River delta. This would lighten the load of the vessels sailing through shallow waters. The soldiers on the coastal route were caught in a storm and many were drowned, as Tacitus explains in gruesome detail (Tac. Ann. 1.70). The coastal land route along the North Sea coast proved to be equally or more dangerous than the inland land routes:
After a while, through the force of the north wind and the equinoctial season, when the sea swells to its highest, his army was driven and tossed hither and thither. The country too was flooded; sea, shore, fields presented one aspect, nor could the treacherous quicksands be distinguished from solid ground or shallows from deep water. Men were swept away by the waves or sucked under by eddies; beasts of burden, baggage, lifeless bodies floated about and blocked their way.
After numerous difficulties and provisioning failures using the inland (land and river) and coastal land routes, in 16 Germanicus decided to invade using the North Sea. The sea was used by Roman forces earlier, but now it was going to be the major invasion route.
But by embarking on the sea, invasion would be easy for them, and a surprise to the enemy, while a campaign too would be more quickly begun, the legions and supplies would be brought up simultaneously, and the cavalry with their horses would arrive, in good condition, by the river-mouths and channels, at the heart of Germany. (Tac. Ann. 2.5)
Supplying military campaigns in Germania between the Rhine and Elbe rivers by the North Sea meant that the Romans could enter the large rivers that flowed south to north, the Ems, Weser, and Elbe. From these rivers the Romans hoped they could systematically subdue (if not conquer) the region. So, during the campaigning season of A.D. 16, Germanicus built a fleet of one thousand vessels to transport his legions and numerous auxiliaries and supplies over the North Sea to the Ems and Weser.
However, the North Sea is prone to severe storms at any season, but especially during the winter. So at the end of the campaigning season in 16, Germanicus planned that the army should begin to return to winter quarters in Gaul early, in high summer. They embarked on the fleet of one thousand vessels which was almost totally wrecked in a storm. The vessels were transporting legionaries, numerous auxiliaries, plus animals and supplies:
As this ocean is stormier than all the other seas, and as Germany is conspicuous for the terrors of its climate, so in novelty and extent did this disaster transcend every other, for all around were hostile coasts, or an expanse so vast and deep that it was thought the remotest shoreless sea. Some of his vessels were swallowed up; many were wrecked on distant islands, and the soldiers, finding there no form of human life, perished of hunger … Germanicus’s trireme alone reached the Chauci. (Tac. Ann. 2.23–24)
Like Alexander before him, Germanicus was defeated, not by opposing armies, or lack of knowledge about the terrain ahead, but by a lack of accurate intelligence about weather conditions that proved to be fatal to thousands of soldiers and sailors.
Thus transporting troops, supplies, equipment by the North Sea and up the rivers Ems, Weser, and Elbe represented formidable obstacles, even in the summer months. Transporting supplies along the land route of the North Sea coast would be equally dangerous, as would transport overland or over the small rivers of the interior. These rivers were frequently not used. In winter, the region would be essentially cut off from reinforcements, supplies, and equipment. These facts probably contributed to the establishment of the Roman frontier on the Rhine, not on the Elbe.
PROTECTION OF STORES
The final section of this chapter will treat the proper storage and protection of supplies at military bases of the Roman army. This was of paramount importance. If supplies were collected but later rotted or were spoiled by rodents in storage, the efforts of collection would be rendered useless. Storage facilities were located in the forts or in nearby communities with easy access to sea or river transport. The storage buildings (horrea) were purpose-built, raised off the ground on wooden, brick, or stone supports, and well ventilated to prevent moisture from spoiling the grain (Goldsworthy 2003: 87, 99; Taylor 1997: 248–57).
The greatest enemy of the Roman soldier, his military stores, and his food supplies, was Ratus ratus, the black rat (the brown rat [Ratus norvegicus] did not appear in western Europe until around 1600). It is therefore of considerable logistic importance to consider the potential impact of this rodent and measures taken to combat it. This animal has been endemic to the Mediterranean since the first century A.D., and it carries thirty-five dangerous diseases including bubonic plague and typhus. It will gnaw through almost any material, including wood, leather, and grain. Importantly, one black rat will potentially destroy one thousand pounds of grain per year; he will eat approximately ninety pounds and urinate and defecate on the remainder, rendering it unfit for consumption (Engels 1999: 1–17, 136).
The maintenance of the safety of the army’s food supplies and the health of the troops was of paramount importance for any competent commander. Left unchecked, black rats had the potential of inflicting serious, even fatal, damage on Roman military forces through destruction of food supplies and spreading of disease. Since the black rat is a climbing species, also called the roof rat or the tree rat, neither the dog nor the ferret can adequately control it. The raised horea would offer little deterrence to the rats.
Only one natural predator was (and is) available around human settlements, Felis sylvestris catus, the domestic cat. This animal has the capacity of killing and eating three small mammals or birds per day or 1,100 per year. If one cat can kill 500 black rats per year, well within its capacity, it will potentially save 250 tons of food per year for a military base, farm, or village. One mummified domestic cat, dating to the second century A.D., found at the Roman port of Quseir el-Qadim on the Red Sea, had no fewer than six black rats in its digestive system before it died. Little wonder it received a formal and respectful burial.
Cats indeed were familiar companions of the Roman soldier when stationed in his fort, and probably of the sailor as well. This relationship can be seen in some soldiers’ names containing the word cat (cattus, aelurio). An inscription from the Third Legion Cyrenaica has a figure of a cat engraved on it, and reads Leg[io] III Cyr[enaica] feliciter invicta, “The Third Legion Cyrenaica auspiciously undefeated.” Also, the first appearance of the Latin word cattus, which became the root for the word cat in most Western languages comes from an inscription from Rome dating to A.D. 144. The sixth century of the third cohort of the Praetorian Guards called themselves Catti, “The Cats.” Cat remains are found in Roman military installations throughout the Rhine and Danube frontier regions. They are also found at Roman naval installations such as Pevensey in southern England. The connection remained in the British navy until 1975 as cats were considered essential for a vessel’s safety and good luck. This relationship should be considered for our understanding of the Roman soldier, and the protection of his stores and food supplies, to be complete.
BIBLIOGRAPHY
Adams, J. P. 1976. The logistics of the Roman army. Diss., Yale University.
Badian, E. 1975. “Nearchus the Cretan.” YCS 24: 147–70.
Bar-Kochva, B. 1976. The Seleucid army: Organization and tactics in the great campaigns. Cambridge.
Blackbourn, D. 2006a. The conquest of nature: Water, landscape, and the making of modern Germany. New York.
———. 2006b. “The making of the modern German landscape.” Historically Speaking 8: 2–4.
Bosworth, A. B. 1988. Conquest and empire: The reign of Alexander the Great. Cambridge.
———.1996. Alexander and the East: The tragedy of triumph. Oxford.
Bury, J.B., with R. Meiggs. 1975. A history of Greece to the death of Alexander the Great. New York.
Casson, L. 1959. The ancient mariners. New York.
———. 1994. Ships and seafaring in ancient times. Austin.
Connolly, J. 2006. “The logistics of the Roman army in ancient Germany.” Honors Thesis, University of Arkansas.
Engels, D. 1978. Alexander the Great and the logistics of the Macedonian army. Berkeley.
———. 1980. “Alexander’s intelligence system.” CQ, ns 30: 327–40.
———. 1999. Classical cats: The rise and fall of the sacred cat. London.
Erdkamp, P. 1998. Hunger and the sword: Warfare and food supply in Roman republican wars, 264–30 B.C. Leiden.
Farrokh, K. 2007. Shadows in the desert: Ancient Persia at war. Oxford.
Ferrill, A. 1985. The origins of war: From the stone age to Alexander the great. London.
Goldsworthy, A. 1996. The Roman army at war, 100 B.C.–A.D. 200. New York.
———. 2000. The Punic wars. London.
———. 2003. The complete Roman army. London.
———. 2006. Caesar: Life of a colossus. New Haven.
Hanson, V. D. 1989. The Western way of war: Infantry battle in ancient Greece. New York.
———. 2005. A war like no other: How the Athenians and Spartans fought the Peloponnesian war. New York.
Heckel, W. 1992. The marshals of Alexander’s empire. London.
Labisch, A. 1975. Frumentum Commeatusque: Die Nahrungsmittelversongung der Heere Caesars. Berlin.
Maloney, E. S. 1987. Chapman piloting. New York.
Maurice, F. 1930. “The size of the army of Xerxes in the invasion of Greece, 480 B.C.” JHS 50: 210–35.
Peddie, J. 2005. Hannibal’s war. Phoenix Mill.
Roth, J. 1999. The logistics of the Roman army at war, 264 B.C.–A.D. 235. Leiden.
Strauss, B. 2004. The battle of Salamis. New York.
Taylor, I. D. 1997. The logistics of the Roman army in North Africa. Diss., University of Arkansas.
Wells, P. S. 2003. The battle that stopped Rome. New York.