Post by Teuta1975 on Jul 9, 2008 21:11:12 GMT -5
Illyris, Rome and Macedon in 229-205 B.C.
Author(s): N. G. L. Hammond
Source: The Journal of Roman Studies, Vol. 58, Parts 1 and 2, (1968), pp. 1-21
Published by: Society for the Promotion of Roman Studies
Stable URL: www.jstor.org/stable/299691
In JRS LVI (I966) I gave a description of the Aoi Stena which was based on autopsy, and I discussed the campaigns of Rome against Philip V of Macedon in the years 200 to 198 B.C. In this paper I am concerned with the area farther north which Rome acquired in 229 B.C. and with the actions which took place there before 200 B.c. Many scholars have discussed Rome's early activities in Illyris but practically none of them has trodden the ground. My knowledge of most of the area may help me to advance more down-to- earth views of the extent of Rome's sector in Illyris and of Roman and Macedonian policies.
I include some new evidence on the position of Dimallum.1 The salient feature of Central Albania is the belt of coastal plain which extends from north of Lesh (Lissus) to north of the Gulf of Valona (see fig. i). The widest and richest part of this plain is in the Myzeqija, which extends southwards from Kavaje. The Myzeqija in particular is integral to the economy of Central Albania, the area which was called Southern Illyris in the third century B.C. The transhumance of sheep has always been practised in this part of the Balkans, and the coastal plain of Albania with months of very heavy rainfall in October and March affords exceptionally fine pasturage for the winter period.
The owners of the flocks which are driven from the plain to the mountains for pasture in the summer months have their homes not in the swampy plains but in the hills of the interior. On the other hand during the summer the coastal plain grows the cereals which are consumed in the hinterland, where there is a deficiency of arable land.2 Thus the lord of the plains has a stranglehold on the economy of Central Albania. In the third century B.C. the Myzeqija was held by the two Greek cities, Apollonia and Epidamnus. From a military point of view the coastal plain is difficult to defend. In the spring and early summer and again in the autumn it is so heavily flooded that movement by an army across it is not possible,3 and for purposes of movement and communication an occupying force must possess the rising ground which runs from Apollonia in the south through Kug, Lushnje and Kavaje to Epidamnus in the north. Even so there is no possibility of defence in depth against an enemy who attacks from the hinterland, and the obvious policy for an occupying power would be to hold or threaten the main route of entry from the east, namely the Shkumbi valley, and also to hold the range on the southern rim of the plain. Of this range Mt. Shpiragrit is the most important part because it faces Berat, through which the subsidiary routes from the east pass as if through a funnel. I became aware of these economic and military factors when I travelled in late March and early April from Durres (Epidamnus) via Kavaje to Elbasan in the Shkumbi valley, seeing the river in spate, the coastal plain flooded and the numerous herds of sheep assembling on the higher ground for the move eastwards, and then again in the summer on other occasions when I went from Elbasan to Berat and from there to Poyani (Apollonia), and when I walked from Byllis in the Aous valley to Berat via Metoh, taking six and a half hours.
The defence of the coastal plain against an enemy attacking either from the south or from the north raises different problems. Here the occupying force has depth but his front is narrow and can be easily turned; it is therefore necessary to protect the open flank by holding some of the hill country. In December I940 the Italian army in central Albania held a position facing south and covering what General Papagos called 'the Kelcyre and Tepelene junctions ', i.e. the Aoi Stena together with the passes at either end of the Stena (see JRS LVI, 40, fig. 2). This was always the strongest position for an army in Illyris to hold against an attack delivered from within Epirus. Failing this, such an army would be wise to hold the hilly country just south of and to the east of Apollonia. Both at the Aoi Stena and at Apollonia a flank guard is needed, reaching into the hills towards Mt. Tomor, and covering Berat.5 Defence against an enemy attacking from the north is made difficult by the deep re-entrant through the upper valley of the Black Drin, which rises in Lake Ochrid.
In consequence it is necessary either to hold the area north of the Drin, that is on a line running inland from Scodra, or to hold the hill country to the north of the Shkumbi valley running inland from Durres (Epidamnus). To the mariner Central Albania is distinctly unattractive. The coast between the Gulf of Valona and Ulcinj is generally flat and featureless; it offers few roadsteads and fewer harbours; and it is rendered dangerous not only by the southwesterly gales and in places by the northerly ' Bora ' but also by the changing positions of offshore banks which extend out to sea as far as three miles. The advice of The Mediterranean Pilot 6 to the sailor setting off on the 40 mile stretch from the Bay of Valona to Kep i Gagji is brief and to the point: ' the utmost caution should be exercised when approaching this portion of the coast and at night it should be given a wide berth '. 6 Indeed for the hundred miles between the Gulf of Valona and Ulcinj there are only two sets of harbours. The first was controlled in ancient times by Epidamnus. It consists of Durres harbour, formed by two moles, and of Pali Bay, 6 miles to the north. Each offers shelter from some winds. Between them runs a high ridge (the name Dyrrachium was derived from it), which is a conspicuous feature ' showing from afar like an island with five heads ' and of great value for making a landfall.7
The second set is in the Gulf of the Drin : the mouth of the Drin, navigable to Lesh; the harbour of Shen-Gjin or S. Giovanni di Medua, entered by a narrow channel and suitable for small vessels (the Nymphaeum of Caesar, BC 3, 26, 4, ' qui ab Africo tegebatur, ab Austro non erat tutus '); the navigable river Buene, giving access to the Lake of Scodra, which is only 6 metres above sea level (Livy's ' Barbanna ... ex Labeatide palude oriens', 44, 31, 3); 8 and the small harbour of Ulcinj (the Roman Ulcinium), dangerous in southeasterly winds. Between them this set of harbours offer shelter from any wind. In antiquity Lissus had a harbour but this has silted up and disappeared. And in the south the mouth of the Aous (Vijose), navigable as far as Apollonia, was used but only when conditions of weather were favourable (cf. Plu., Caes. 38).
But then, as now, Epidamnus was the most important station on the coast; anyone holding Epidamnus and denying access to its harbours can make navigation along or towards this coast under sail extremely difficult in bad weather.9 In the days of sail this coast was important for reasons which no longer obtain (see fig. 2). The direct crossing of the Straits of Otranto from the all-weather harbour of Oricum in the Bay of Valona to the heel of Italy was rendered dangerous by exposure to the Bora and the southwesterlies, which create very rough, steep seas, and by the proximity of the formidable Acroceraunian coast. In consequence sailors preferred to cross the Ionian Gulf from Epidamnus to Bari or Brindisi.10 Thus Epidamnus served Greeks trading to the west and Italians trading to the east. At the same time it was the main port of call in the Ionian Gulf for shipping up and down the Adriatic Sea, and so was of paramount importance to those Illyrians who concerned themselves with the sea. The coast of Central Albania has never bred a seafaring tradition, not surprisingly in view of its nature, but the islands off the river Narenta (the ancient Naro) and the fjords such as that of Kotor (' Sinus Rhizonicus ') produced Illyrian navies from time to time. The Illyrian warship, whether developed by the Liburnians or the Ardiaeans, was small and extremely fast under oar and under sail, and it was more fitted for raiding than for set engagements at sea.11
Such Illyrian navies were always anxious to gain possession of Epidamnus. The Liburnians held the site in the seventh century, Monunius early in the third century; and the Ardiaeans, whose naval power was based on the area of the islands and fjords but extended to the southern Ardiaeans by the Gulf of the Drin, were inevitably drawn towards it in the latter part of the third century. The first occasion of Roman intervention across the Adriatic Sea was a naval occasion. It was prompted by the scale and success of the piratical Illyrian raiders, who had made a habit of raiding the coasts of Elis and Messenia, defeated an Aetolian army at Medeon in Acarnania in 231, and then in 230 not only raided the coasts of Elis and Messenia once more but captured Phoenice in Epirus and compelled the Epirote League to enter into alliance. More than piracy at sea was involved; for a land force under Scerdilaidas invaded Epirus at such short notice that we may infer that the Illyrians held the area between Berat and the lower Aous valley as well as 'some parts of Epirus ' (App., Illyr. 7, T-rS 'HrrEipovu rvca).12 In 230 this expanding Illyrian state was in alliance with Macedon, Acarnania and Epirus (Plb. 2, 6, 9).
In spring 229 B.C. an Illyrian fleet, assisted by Acarnanian ships, defeated the fleets of the Aetolian and Achaean Leagues, gained possession of Corcyra and laid siege to Epidamnus. The Illyrians, led by Teuta, widow of Agron, king of the Ardiaei,13 were within an ace of controlling the entire coast from Dubrovnik to the mouth of the Corinthian Gulf.14 Meanwhile Rome was on the move. Italians trading by sea towards the east had been molested repeatedly in recent years by the Illyrians, and in the course of 230 B.C. more Italians than usual had been intercepted at sea, robbed, kidnapped or killed.15 The Senate sent ambassadors to Queen Teuta late in 230 B.C. The conversations ended in a fracas, and one of the ambassadors was assassinated on the way back.16 The Senate then began to organize a naval and military force. However, it did not declare its intentions to Teuta, and she operated during the first part of the campaigning season of 229 B.C. as if her position was not under threat from Rome.
In the first Punic War Rome had become painfully aware of sea power, and she had realized that the long coast of Italy was vulnerable to seaborne attack. Hamilcar Barca had raided Locri and other places in south Italy, and in 246 Rome had planted a colony at Brundisium, which faces the Ionian Gulf. During the war Italy had traded very intensively across the Ionian Gulf,17 partly because the Carthaginian raiders had not penetrated into these waters. If the Illyrians now succeeded in their aims, they would close the mouth of the Adriatic Sea and control the approaches to South Italy through the Ionian Gulf from their bases on the coast and on Corcyra. For a variety of reasons 18 the Senate was not prepared to let this happen. In the second part of the campaigning season Rome sent into action 200 ships, 20,000 infantry and 2,000 cavalry-a large force, but not unduly large for the task the Senate envisaged.19 The campaign was brilliantly conducted (Plb. 2, i).
The Roman fleet sailed first to Corcyra, where Teuta's commander, Demetrius of Pharos, having previously come to a secret understanding with Rome, handed over the island. It then sailed to Apollonia, which joined the Roman cause, and met there the Roman army which had been brought over from Italy. They advanced next towards Epidamnus, the fleet sailing along the coast and the army probably moving along the eastern side of the Myzeqija plain, where the 2,000 cavalry had excellent ground for manoeuvre and plenty of pasture. The Illyrians abandoned the siege of Epidamnus, which now joined the Roman cause. The Roman army then turned further inland and overran the Ardiaei,20 probably in the lowlands and the hinterland between the Mati and the Drin,21 and the Roman fleet, sailing up the coast and taking some places by assault, relieved the island of Issa, which the Illyrians were besieging. Army and fleet then returned to Epidamnus. Casualties had been suffered during the last phase only, for instance at an unidentified place called Noutria. The spectacular success of Roman arms was due to careful planning, the suddenness of the unheralded attack (critics might have called it a treacherous attack),22 the double- dealing of Demetrius of Pharos and the accession of Apollonia.
The strategy was good: the seizure of Corcyra cut off the possibility of naval reinforcements from Acarnania, the occupation of Apollonia and the Myzeqija cut off the possibility of land reinforcements from Epirus, and the rolling up of the Illyrians on a narrow front with naval support and a mobile cavalry force cost Rome few casualties. The whole campaign was over so quickly that Teuta's other ally, Macedon, could not have intervened in time to forestall the Illyrian collapse, even if she had wished to do so. Victory won, Rome had several courses open to her. She could pursue the defeated Illyrians to Arbon and Rhizon, where they had taken refuge, and give a knock-out blow to the Ardiaean monarchy. She could occupy with troops or with colonists the strategic points in the areas she had overrun, e.g. Corcyra, Apollonia, Epidamnus, Issa and Pharos, and thus control the approaches by sea to the outlet of the Adriatic Sea; she would then hold safe ports on the eastern coast of that sea to balance Brundisium.
She could treat the whole operation as a punitive raid and withdraw her forces, having taught Teuta a salutary lesson and leaving her friends in Illyris and the islands to maintain their independence as they had done hitherto. In fact she chose a course intermediate between the second and the third of those which I have suggested. She set up Demetrius as ruler of a kingdom 23 based on Pharos and including most of the Illyrians overrun by Rome (particularly, I imagine, in the area of Scodra), and she no doubt hoped he would act as a buffer between Rome and the Ardiaean monarchy. Issa in the north and Corcyra in the south became in effect dependents of Rome, Issa providing a contingent later to the Roman navy. On the mainland, although the Romans were approached by the envoys of a number of tribes during the advance northwards from Epidamnus (Plb. 2, I, I , cavyjj4av-rTcv 6 TwpEcyUEvrTOV ca-rTOis KCal XrAEovc1v), they accepted only the Parthini and the Atintani into what was or became a position of dependency. They already had Epidamnus and Apollonia on their side. During the winter of 229-8 one of the consuls stayed with forty ships and conscripted a force " from the surrounding states " which kept an eye on the Ardiaei and the other (Illyrians) who had submitted to Rome.
In spring 228 B.C. Teuta opened negotiations with Rome and concluded a treaty under which she undertook to pay an indemnity, evacuate all " Illyris " except for a few districts, and not sail beyond Lissus with more than two lembi even unarmed.24 On the conclusion of the treaty the Roman consul sent envoys to the Aetolian and Achaean Leagues, explaining the reasons for Rome's initiative and the terms of the treaty she had concluded with the Illyrians. They doubtless emphasized the fact that they had delivered the Greek cities of Corcyra, Apollonia, Epidamnus and Issa from the common enemy, the Illyrians. The envoys received a friendly welcome and then returned to Corcyra. The Roman settlement needs some clarification on the geographical side (see fig. i with Inset). Epidamnus and Apollonia were both wealthy Greek city-states with a very copious silver coinage which had a wide circulation. There is no doubt that each of them had an extensive territory. We have more information about Apollonia.
In her coinage from 229 B.C. onwards she showed the Nymphaeum, which we know was close to her frontier with Byllis and Amantia, independent states which held the north side of the Aous valley above the vicinity of Romes.25 The mass of Apollonia's territory was then not to the south but in the Mizeqija, extending northwards certainly to the Semeni (Apsus) 26 and probably to the Shkumbi (Genusus) and inland towards Mt. Shpiragrit. Epidamnus is likely to have controlled both the plain of the Arzen river and the northern part of the Myzeqija as far south at least as the north bank of the Shkumbi river. As Asparagium (probably Rogozine) was in her territory,27 she formed the outlet through which all trade following the north side of the Shkumbi valley had to pass. The Parthini were a tribe adjacent to Epidamnus (Appian, BC 5, 75, gevos 'ETrBl&a,vcp -rapoiKov; cf. Dio. 41, 49, 2). When Pompey moved from Epidamnus to the north bank of the Semeni (Apsus) and Caesar moved from Apollonia to the south bank of the Semeni (Apsus), they faced each other probably at Ku9. When Caesar entered the Shkumbi (Genusus) valley probably through the sink southwest of Elbasan in order to pursue Pompey down the valley to Asparagium, he captured on his way a Parthinian town which Pompey had garrisoned (Caesar, BC 3, 41, I, and Dio 4I, 47, I). Thus the Parthini held at least the middle valley of the Shkumbi (Genusus) river. Moreover, as Pliny reported that the Parthini had the Dassaretii 'behind them ' (NH 3, I45) and as the Dassaretii extended as far east as at least Lychnidus (Livy 43, 9, 7), the Parthini held the upper valley too.28 The Atintani, as I have shown elsewhere,29 have nothing to do with the Atintanes in the upper Drin valley in Epirus but are an Illyrian tribe about one day's journey from Epidamnus in the direction of Macedonia, and occupying very high country, visible from Epidamnus and near the (Macedonian) frontier of Illyris (Polyaenus 4, 11, 4).
This tribe evidently held the region of ?(ermenike, which extends from just north of Elbasan to the watershed of the highest reaches of the Black Drin. The dispositions of Rome in the northern part of Central Albania are now clear. First Demetrius of Pharos as suzerain of some Illyrian tribes from north of Epidamnus to the vicinity perhaps of Scodra acted as a buffer between the Ardiaean monarchy and the zone of direct dependence on Rome. Second, a continuous line was held by the dependents of Rome-Epidamnus, the Parthini and the Atintani-from the coast to the highest westerly sources of the Black Drin and in sufficient depth to be defensible against attack from the north. At the same time this defensive position not only cut off the Ardiaean monarchy from its ally, Macedon, since the Dardani were pressing down upon Pelagonia at this time,30 but also blocked the main route from Macedonia to the Adriatic coast, that later followed by the Via Egnatia. The chief threat to this defensive position might come from Dassaretis, the high territory between the two arms of the Semeni river, which are called the Devoli and the Osum, and extending inland to the main watershed through which the Tsangon pass and the pass of Vatokhorion lead into the Macedonian canton of Orestis. In 230 B.C. Dassaretis was evidently under the control of the Illyrians, because Agron held ' parts of Epirus ' and Scerdilaidas was able to move swiftly to Antigonea in Epirus at a time when neither Agron nor Scerdilaidas held Apollonia (App., Illyr. 7 and Plb.
2, 5, 6).
The defeat of Teuta's forces and Rome's alliance with Epidamnus, the Parthini and the Atintani left Dassaretis without any direct political affinity. Rome had the good sense not to include it in the zone of direct dependence upon herself. Thus Dassaretis became an independent area and formed a buffer between Macedon and Rome. In the south the zone of direct dependence ended with the territory of Apollonia (the Apolloniatis) which had a footing on the southern bank of the Vijose (Aous) but did not include Aulon, Byllis, Amantia or Oricum. These four small states formed an independent or neutral group situated between Rome and the Epirote League, of which the northernmost cantons were Chaonia and Parauaea. The boundary to the southeast will be defined more closely when we consider the position of Dimallum.31 Rome now had access by right to harbours in Demetrius' realm, at Lissus, Epidamnus, and the Vijose river, then navigable as far inland as Apollonia. On the other hand, the Illyrians of Teuta's kingdom were not only debarred by treaty from sailing south of Lissus with more than two unarmed lembi, but were in effect dependent on Rome's favour for the peaceful use of these indispensable harbours.
The economic interest of the neighbouring coastal areas, that round Scodra32 in the north and that round the Gulf of Valona in the south, drew them strongly to Rome, because Italy offered an excellent market for their products. The inland areas of the Parthini, the Atintani and the Dassaretii depended, as we have seen, upon the Myzeqija for winter pastures and for cereals and also upon Epidamnus and Apollonia for import and export, so that they too had an economic interest in adhering to the political power which was lord of the plain. It is customary to call the zone of direct dependence on Rome 'the Roman protectorate ', a vaguely benevolent and flattering euphemism for an extremely shrewd extension of Roman power. Rome's dependents, whether they enjoyed the title of ally or subject,33 were to be her dependents for good and not free agents, able to transfer their allegiance to other states with impunity. We do not know whether, once the treaty with Teuta was completed, Rome left any troops at Epidamnus or Apollonia, but the point is unimportant, since in a matter of hours she could send an army across to her treaty-ports in order to hold the position defended by her subjects. To use a modern phrase, she intended to maintain ' not a military presence but a military capability-a capability to get troops there if they were needed '.
Having completed her dispositions, Rome disregarded Macedon, the Epirote League and Acarnania, the very states which were most immediately affected by the appearance of Rome as a new constellation on the eastern shore of the Adriatic Sea. Instead, Rome sent embassies to the enemies of those three states, that is to the Aetolian and Achaean Leagues, which had at least the superficial merit of being in opposition to Teuta's Illyrians, and later to Corinth and Athens, both hostile to Macedon. This deliberate and public move by Rome made it clear to Macedon and the Greeks that in any war in the southern Balkans Rome's sympathies would lie initially with Aetolia and Achaea and against Macedon, Epirus and Acarnania. The announcement of this alignment, coupled with the astute organization of Roman interests in southern Illyris, could not fail to cause alarm in Macedon, Epirus and Acarnania. It would be naive to suppose that Rome was unaware of the fact.34
The Roman successes at sea and on land had damaged the prestige but not the power of the Ardiaean monarchy, which was based upon Dubrovnik and the Dalmatian coast; 35 and even within Illyris, an area which began probably around Scodra, a few districts were still in Teuta's hands (Plb. 2, 12, 3). Demetrius of Pharos, who had the skill of a Perdiccas in changing sides, became the successor of Teuta on her death and married Triteuta, mother of the infant king, Pinnes, so that the first or northernmost part of the Roman settlement collapsed completely. There was no longer any buffer between the Ardiaean monarchy and the zone of direct dependence on Rome. Meanwhile Rome became deeply involved in a war with the Gauls in the Po valley which lasted from 225 to 222 B.C., and within these years Demetrius advanced both at sea and on land, engaging in piracy south of the Lissus line and winning over the Atintani from Rome.36 The latter step was the more important; for Demetrius was thereby opening the door to co-operation with Antigonus Doson, king of Macedon, and also making infiltration into Dassaretis possible if he gained Antigonus' co-operation. In 223 B.C.
Demetrius accompanied Antigonus on his invasion of the Peloponnese and in 222 B.C. Demetrius' force of i,6oo Illyrians played an important part in the battle of Sellasia. Polybius enumerates them among Macedon's allies (2, 65, 4).37 Early in 220 Demetrius and Scerdilaidas sailed south of Lissus with 90 lembi and carried their raids into the Cyclades (Plb. 3, i6, 3 ; cf. 4, i6, 6-9, and 4, 19, 7-9). In this year a Roman fleet suppressed some pirates, probably the Istrians (App., Illyr. 8).38 In this year also Demetrius was ravaging the territory of ' the states in Illyris subject to Rome ' and was trying to subdue them (Plb. 3, i6, 3).39 In order to do this effectively, he came through the gap opened by the defection of the Atintani, entered Dassaretis and attacked the Parthini and the Apolloniates from there. In 2i9 the Romans sent the two consuls with an army which was probably as large as that of 229 B.C. to punish Demetrius and his Illyrian collaborators (Appian, ibid.). The attack was again unheralded, but Demetrius had laid his plans in advance. He placed a considerable garrison with suitable supplies in Dimallum, evidently a town of which he already had possession, and he brought about changes of government in his favour 'in the other states ', which presumably now for the first time he took under his influence (Plb. 3, I8, i).40 Having established Dimallum as a strong point, he went to Pharos and prepared to hold Pharos with a picked force. On landing in Illyris, the Romans attacked Dimallum first; knowing it was thought to be impregnable, the consuls hoped to capture it and spread alarm among the Illyrians.
Dimallum fell in a week. Envoys came in from ' all the states ' with offers of submission. The consuls made appropriate agreements in each case and then moved on to attack Pharos. The island fell, but Demetrius escaped. At Actium he joined Philip of Macedon, who had inherited Antigonus' friendship with him and had himself visited Scerdilaidas in the winter of 220-219. The Romans crossed over from Pharos to Illyris, gained control of the rest of Illyris and made a settlement of all its affairs. A triumph was accorded to the consuls on their return to Rome at the end of the summer.41 The Roman intervention of 2I9 needs little explanation. Demetrius had flouted the Roman settlement of 229 all along the line by deserting with the dependents Rome had given him, by detaching the Atintani from their allegiance to Rome, by sailing in strength south of Lissus, by ravaging the territories of Rome's dependents (evidently those of the Parthini, Epidamnus and Apollonia) and by trying to subdue them and so eliminate Rome's holding on the mainland. The Senate must have seen that inaction would be fatal to her position in Illyris. Once in control, with Macedon, Epirus and Acarnania as his accomplices, if not his formal allies, Demetrius rather than Rome would control the Ionian Gulf and the Straits of Otranto at the very time when a renewal of war with Carthage might be expected. 42
The terms of the Roman settlement in 219 are not described in our sources. It should be noted that even less than in 229 did Rome attack the basis of Illyrian power which lay farther north, and it may be doubted whether on this occasion her forces penetrated even to Scodra. Moreover, no steps were taken against Scerdilaidas, though he also had sailed south of Lissus and he also was an ally of Philip. Clearly Rome had no desire at this time to go too far or to increase her commitment in Illyris. Wherever she went, she certainly put pro-Roman parties in power, but it seems as if Dimallum may have been the only addition she made to the number of dependents she had had since 229; for in 215, when Hannibal and Philip made an alliance, one of their aims was to prevent Rome from being in control of Corcyra, of Apollonia and Epidamnus, of Pharos, of ' Dimale' and the Parthini, and of Atintanis (Plb. 7, 9, I3). Thus in 219 the Senate seems to have been content to restore her full control of the 229 group of states, which interposed a barrier between the Illyrians and the Macedonians, blocked the easiest route from Macedonia into this group of states, and left independent or buffer areas north of Epidamnus, in Dassaretis and in the Gulf of Valona and its hinterland.
No doubt she reiterated the ban on Illyrian lembi sailing south of Lissus. Once again Rome did nothing either to ease or to exacerbate relations with Macedon or Epirus. Yet her determination to stand firm in Illyris was itself alarming in view of Rome's record and reputation. On the other side Antigonus' and Philip's acceptance of Demetrius as an ally in good fortune and in bad made it equally clear that Macedon favoured Rome's enemies in Illyris and was prepared to show it.43 In modern terminology Rome and Macedon were now engaged in a ' cold war '. We must pause in the narrative to consider the position of Dimallum. It has been placed by scholars on the coast, not on the coast, near Epidamnus, in the territory of the Parthini and not in the territory of the Parthini.44 Consideration of the texts and of the geographical conditions can reduce the number of alternatives. The coast between Epidamnus and the Gulf of Valona is low and swampy; a broken line of low hills runs parallel to the coast, except at Barderoll where the hill comes down to the coast itself; and then inland of the line of hills there is a swampy plain once again. Now Dimallum was so strongly situated as to be thought impregnable (Plb. 3, i8, 3). It would be difficult to find even at Barderoll an impregnable site along this coast. Moreover, Dimallum was in dispute between Rome and Macedon (Livy 29, i2, 13); we therefore need a place on the Macedonian side of the Roman block of territory and not on this coast which belonged mainly (and had done for centuries) to Epidamnus and Apollonia. As regards the Parthini and Dimallum the one is a tribal state, subsuming under its name the ' urbes Parthinorum ' (Livy 43, 23, 6), and the other is a city, probably autonomous, negotiable as a separate entity between Rome and Macedon.
The two are always mentioned as separate units by Polybius and Livy Plb. 7, 9, I3, UIr8' ETvati Pcoiaious KUvpOUv KEpKupaciov JirS' 'Aro ovIcTrov KiI 'ETricbavicov tr&8 Oa)povu pTr6s Atip&?aAr1s KCI TapeE0ivcov lPr6' 'A-rVravia ; Livy 29, I2, 3, ' Parthinosque et propinquas gentes alias motas esse ad spem novandi res, Dimallumque oppugnare (sc. Romanos) ' ; Livy 29, 12, 13 'ut Parthini et Dimallum et Bargullum et Eugenium Romanorum essent '. I conclude then that Dimallum (and also Bargullum and Eugenium) are not Parthinian towns 45 nor in Parthinian territory, which extended from the upper valley of the Shkumbi to somewhere inland of Asparagium (probably Rogozine), which was in the territory of Epidamnus. It follows then that Dimallum lay either north of the line of Epidamnian and Parthinian territory, that is facing northern Illyris, or south of Parthinian territory and east of Apolloniatis, that is facing Dassaretis. Those who place Dimallum near Epidamnus do not rely on any ancient evidence. But there is a clue to the contrary in Livy 29, I2, 5, where the Roman consul who had gone from Epidamnus to Dimallum, abandoning the siege, retired to Apollonia (' quo Sempronius se receperat '), and from this I infer that Dimallum lay nearer to Apollonia than to Epidamnus.
But the decisive point is that Dimallum was in dispute between Rome and Macedon at a time when Macedon held nothing north of Epidamnus but did hold Dassaretis. Dimallum, then, lay between Apolloniatis and Dassaretis, to the south of the Shkumbi valley, and therefore probably on the range of Mt. Shpiragrit. ' La decouverte de la cite illyrienne de Dimale' is the title of an article published by Burhan Dautaj in Studia Albanica 1965, I, 65-7I, of which an offprint was very kindly sent to me in 1967 by Professor Frano Prendi. In I963 and I964 Dautaj excavated the fortress of Krotine,46 which is situated on ' a fine peak' of the Shpiragrit range (see fig. 3). This peak, being 404 metres above sea-level, is the highest of the western outliers of the range; it has ravines on three sides of it, and there are traces of a circuit-wall on the steep slopes. It was an exceptionally strong place and might well have been thought impregnable. There are two summits inside the circuit, the higher being the acropolis and the lower providing accommodation for three-quarters of the population. Praschniker estimated the circuit of the wall to be 2,400 metres. When the acropolis was excavated the most important discovery was that of numerous stamped tiles of Hellenistic date, rectangular, red to grey in colour, some long and narrow with squared edges, others broad with slightly rounded edges. Within an excavated area of some 500 square metres Dautaj found no less than 150 tiles stamped on the edge with the monogram shown in fig. 4(a), and he deduced that these tiles were local to the site. In addition to the monogram i6 tiles were stamped with the name NEYTQP or NE22TQP in the nominative, which appears to be that of a local potter since it is not found on stamped tiles elsewhere in Albania. Four tiles bear stamped monograms and the stamped word AIMAAAITAN, cf. (fig. 4(b).
Twelve tiles are stamped HPAIQN, a name which appears also on tiles of Apollonia; but these tiles are smaller and poorly made compared with those of Apollonia. They were made presumably for members of a cult in honour of Hera or for a group so-called, like the Heraeis of early Megara.47 One tile had a genitive plural ending in - OANIQN. Names in the genitive, evidently of magistrates, were stamped on other tiles: AMYNTA, APMHNOC, API:TOMENEOS and - MAXOY.48 The word AIMAAAITAN written with the broken-bar alpha has the same form of genitive and much the same lettering as AlX:ITAN on coins of Lissus attributable to some time within the period 250-200 B.C.49 The name of the city at this time was evidently Ai|caXAoS or AhiaAAXov as we find Dimallum in Livy. It is a Greek word meaning ' of double fleece ', very suitable to a place overlooking the rich sheep-pastures of the Myzeqija.50 Another independent city, Eugenium, in this region has a Greek name, used of a type of laurel.51 Coins were found on the acropolis of Dimallum. The earliest of these were coins of Epidamnus dating to the period 320-270 B.C. The city may have been founded c. 290 B.C. by Pyrrhus or his successors, to control the route along the side of the plain to the sink between the Semeni and the Shkumbi rivers, just as Antigonea was founded by him farther south.52 In any case it seems beyond doubt that this is the city into which Demetrius put a garrison in 219 when he had overrun Dassaretis.
We do not know its earlier history. It may have been taken by Rome already in 229, because it threatened the route along the side of the Myzeqija, and it may then have gone over to or been acquired by Demetrius at some time before 219. In any case, once captured by assault in 219, Dimallum was regarded by Rome as her possession, as Plb. 7, 9, 13 shows. The Roman settlement of 219 was treated with scant respect. In 218 Demetrius may have indulged in raiding again (App., Illyr. 8 fin.), and Scerdilaidas certainly sailed south of Lissus and supported Philip in his naval enterprises that year. Rome may have sent a punitive force against Demetrius at or near Pharos, and she kept on good terms with the young king, Pinnes, ruler of the Ardiaean state (App., Illyr. ibid.). Scerdilaidas had troubles in Illyris with' city-dynasts ' (Plb. 5, 4, 3 -roAtluvv&orras) and these troubles limited the help he gave to Philip in this year; the ' city-dynasts ' may have been persons set up under the Roman settlement of 219, but many other explanations for their activities are possible in an area so split by tribal feuds.
In 217 Rome sent envoys to Philip of Macedon demanding the surrender of Demetrius (Livy 22, 33, 3). If this was her first diplomatic contact with Macedon, it was an inauspicious one; for the demand was of the kind made to a subject state and not to an equal and independent state, and Philip naturally refused to comply. At this time Pinnes was visited by Roman envoys, who demanded arrears of 'tribute ', evidently a war-indemnity payable in instalments (Livy, ibid.); one wonders if they got any change out of Pinnes. These actions are probably to be connected with Rome's alarm when Hannibal reached Picenum and then Apulia on the Adriatic coast, having won the battle of Lake Trasimene in summer 217. At the same time Scerdilaidas had struck out on his own. By sea he sent 15 lembi south of Lissus to attack Philip's friends off Leucas (Plb. 5, 95, i f., and 5, IOI) ; by land he plundered Pissaeum, a Macedonian town in Pelagonia, won over by threats or promises 53 three towns in Dassaretis, namely Antipatrea, Chrysondyon and Gertous, and overran much of that part of Macedonia which was conterminous with Pelagonia and Dassaretis (Plb. 5, io8, I-2).
As Demetrius had done in 220, Scerdilaidas passed through the high country by Lake Lychnitis between the Parthinian territory of the Roman sector and Macedonia and so entered Dassaretis, which Rome had left as a weak and unassigned buffer area in her settlement of 219. Now Scerdilaidas used Dassaretis as a springboard for attacking Macedonia, as the Romans were to do later in 200 B.C.54 Philip's counter-stroke was the obvious one, to cut the line of entry from northern Illyris into Dassaretis by capturing the district round Lake Lychnitis. Philip therefore ' won back' (Plb. 5, io8, 8, avEKTilacro) Antipatrea, Chrysondyon and Gertous, and then ' captured' (KcTErEAasTo) a number of places: Creonium and Gerous in Dassaretis; of the people round Lake Lychnitis Enchelanae, Cerax, Sation and Boii; in the territory of the Caloecini Bantia; and also of the so-called Pisantini Orgyssus.55 The winter of 217 was now upon him and he disbanded his troops. He had put a stopper on Scerdilaidas, but he had done something which might be more dangerous than anything Scerdilaidas could ever do. He had made himself the immediate neighbour of Rome by occupying Dassaretis. On her side Rome accepted the defeated Scerdilaidas as an ally, just as Philip had recently accepted Demetrius. Polybius believed that in acting against Scerdilaidas in 2I7 Philip wished to consolidate his position in Illyris and then to cross over to Italy, a course advised especially by Demetrius (Plb. 5, Ioi, 6 f. ; 5, 105, i and 5 ; 5, io8, 4 f.).
It is possible that Polybius formed this belief in view of Philip's actions in 216 and 214; for in 217, whatever ideas he did or did not have about Italy, Philip was compelled by his interests at home to stop Scerdilaidas' raids on Macedonian territory. When Philip occupied Dassaretis, then the question of a move westwards became possible for the first time. But to anyone harbouring such an ambition the strength of Rome's position in Illyris was clear. As we have seen, the coast offers very few harbours. Not one of them was in Philip's hands. Scerdilaidas, who was now seeking help from Rome, held the coast from Lissus northwards; Epidamnus and Apollonia held the coast as far as the entry to the Gulf of Valona; the buffer states of Byllis, Amantia and Oricum held the coast of the Gulf; the Acroceraunian shore offered no anchorages at all; and even the harbours of the Epirote League by Buthrotum were blanketed by Rome's naval base on Corcyra. Yet, unless Philip could establish himself on this coast, he had little or no chance of gaining control of Illyris; for so long as the harbours of Apollonia and Epidamnus were open and accessible to reinforcements and supplies, Philip could not hope to reduce the cities by blockade or by siege. Any plan to cross over to Italy would be pointless, if Philip did not already hold ports on the coast of Illyris from which to despatch, supply and reinforce any troops he might want to land in Italy.56 In addition, any lembi which he proposed to build would not be capable of meeting Rome's quinqueremes in a naval action; therefore he could not challenge Rome's naval supremacy. So far as Illyris was concerned, he might be able to deliver an assault force unexpectedly at a strategic place,57 provided that there were no quinqueremes within range, and so capture a port which he could then defend. The obvious target was Apollonia; it was farthest away from Scerdilaidas' bases in the Adriatic Sea, it was closest to Macedonian troops in Dassaretis, and the navigable river, the Aous, was not only easy for an assault force to seize by surprise but also would give harbourage to an attacking fleet.
Philip made two attempts. In the winter of 217/6 he used Illyrian shipwrights to build him a hundred lembi in Macedonian ports. Setting out from Macedonia he rounded the Peloponnese and reached Leucas in early summer 216. There he ascertained that the Roman fleet was lying off Lilybaeum in western Sicily. He sailed on northwards, passed Corcyra and at night time was close to the mouth of the river Aous and ready to deliver his assault force,58 when the report reached him that Roman quinqueremes were actually crossing over, bound for Apollonia (Plb. 5, 109-I IO). It appears that Scerdilaidas, knowing that Illyrian shipwrights were being employed by Philip, had informed Rome and had asked for help in order to stiffen his own naval forces, and it was in answer to this request that some quinqueremes were detached from the fleet at Lilybaeum and sent across the Ionian Gulf (Plb. 5, IIo, 8-9). The timing was in fact fortuitous. But Philip could not know that, nor could he tell how many quinqueremes were on the way (in fact there were only ten). He therefore withdrew hastily.59 An actual clash between Macedonian and Roman troops was thus averted. The expedition had two effects. In 215 Philip approached Hannibal and obtained an alliance, under which, if they made peace with Rome, one condition would be that Rome would never make war on Macedon and Rome should no longer be in control of Corcyra, or of Apollonia and Epidamnus, or of Pharos, or of Dimallum and the Parthini, or of Atintanis, and should surrender to Demetrius of Pharos those of his friends who were interned on Roman soil (Plb. 7, 9, I3). This would give Philip what he wanted most, a guarantee against Roman attack and the removal of Roman power from Illyris. Meanwhile he may have hoped for naval help from Carthage in another attack on Apollonia and Hannibal may have hoped for some military help in Italy from Macedon.
But nothing specific was arranged and nothing came of it. The phrase in Polybius' report of the treaty (7, 9, II) pOT]i0e'UETE 5E IITV cS &v XpEia 1i Kiai cos &v vp90 vilocopEv provided for a future rather than an immediate contingency.60 On the Roman side the attempt by Philip on Apollonia had revealed Philip's intentions; moreover, the treaty between Philip and Hannibal became known to Rome when Philip's envoy fell into Roman hands. Consequently Rome placed a fleet at Tarentum to guard the coast and watch Macedon (Livy 23, 38, 9). A state of war now existed in fact between Macedon and Rome. In 214 Hannibal and Philip undertook concerted, if not simultaneous, actions against Tarentum on the one hand and Apollonia on the other hand. The action at Tarentum engaged the attention of the Roman fleet and enabled Philip to reach the Aous river without being intercepted. This time he sailed up the Aous with izo lembi and attacked the town. On the landward side an army which had marched up through the territory of his ally the Epirote League,61 and also probably some troops from Dassaretis, joined in the attack.
His siege engines tried to breach the circuit wall, which exceeded four kilometres in length. When he did not obtain immediate success, he switched his attack one night to Oricum and captured that city and its capacious harbour.62 By now news had reached the Roman fleet at Brundisium. Laevinus crossed over, probably with 50 warships and a legio classica (Livy 24, 1i, 3), and captured Oricum. He then passed a relieving force into Apollonia at night unobserved. This force together with the Apolloniates made a sortie the following night, killed almost 3,000 men, captured rather more, and brought the Macedonian siege train inside the walls. The Roman fleet then entered the mouth of the Aous river. Philip burnt his fleet and withdrew his army through Dassaretis into Macedonia. Rome now took Oricum into its zone of dependent states and stationed a fleet there, in order to extend its naval holdings and to patrol both sides of the straits (Livy 24, 40,17 ; Plb. 8, I).63 The first clash between Macedonian and Roman troops had resulted in a severe defeat for Macedon. The failure of his two attempts on Apollonia caused Philip to change his strategy. He set out now to reduce the area of the Roman sector gradually and to open up an entry into the territory of the Ardiaei. This he achieved in the course of 213 and 212. He captured Dimallum (now, if not in the campaign of 214) and probably Gerunium and Orgessus; he consolidated his control of Dassaretis; he brought the Atintani and the Parthini over to his side and so opened up the way into northern Illyris. He was probably operating in Atintanis, when he marched in two days to Lissus and Acrolissus, which were his next objective (Plb. 8, I3-I4).
These two strong places were defended not only by their inhabitants but also by troops from the neighbouring parts of Illyris. Philip captured both places by a brilliant stratagem and by hard fighting, and his success led to the surrender or reduction of all the Illyrians of the neighbourhood (&CravTc-r Tro0S Trrppi). Now or soon afterwards his rule extended over the southern group of Ardiaei around Scodra, the subjects hitherto of Rome's friend Scerdilaidas.64 These successes enabled Philip to isolate Epidamnus and Apollonia and to put economic pressure upon them. It was now possible for him to build and man a fleet of lembi on the Adriatic coast, instead of in the Thermaic Gulf; if the earliest coinages of Lissus and Scodra are correctly dated c. 211, they may have been issued to pay shipwrights and purchase timber for Macedon. With such a fleet he could attempt to capture Apollonia or Epidamnus or both and so eliminate the Roman holding in Illyris. He might also make contact at sea with his Carthaginian allies; for the Carthaginian fleet was operating off Syracuse, and Hannibal held Tarentum and most of the harbours in Magna Graecia. A combined attack by Philip's lembi and the Carthaginian fleet upon the Roman fleet based at Oricum was well within the realm of possibility. The commander of the Roman fleet took the initiative by forming an alliance probably in 21i between Rome and the Aetolian League.
The alliance was directed against Macedon 65 and defined the spheres of looting (these divided at Corcyra, Livy 26, 24, i) ; Rome was to act at sea with not less than 25 quinqueremes, the hope being expressed that Scerdilaidas and his son Pleuratus would join the alliance (and provide their fleet of lembi) ; and the Aetolians were to attack Macedon on land. When Philip heard at Pella of the alliance, he made a sudden attack upon the territory of Oricum and Apollonia, and when the army of Apollonia made a sortie he routed it (Livy 26, 25, I-2).66 It is clear that the Roman fleet and troops were away at the time of Philip's attack, and that Philip's army was not large if
the Apolloniates alone made a sortie. It is possible, then, that Philip made his attack both from the land and from the sea, using his Illyrian lembi based on Lissus at a time when the Roman fleet was away. This suggestion gains some support from the statement in the annalistic tradition that after the pact with Aetolia Philip advanced as far as Corcyra but was frightened away by Laevinus, the commander of the Roman fleet (Zonaras 9, 6).
It also fits into Livy's account of Laevinus after the conclusion of the Aetolian treaty operating against Zacynthos and Acarnania and then returning to Corcyra (Livy 26, 24, 15-i6).67 Immediately after these operations Philip laid waste the' nearest part of Illyricum', translating the Greek word 'XlAupis (' vastatis proximis Illyrici '), a very vague phrase, which may refer to the tribes of the upper Drilon valley, as he then moved into Pelagonia. Thereafter we hear no more of him in Illyris, although Scerdilaidas and Pleuratus threatened trouble in 207 (Plb. 10, 41, 4; Livy 28, 5, 7). For a time the Aetolian League served Rome's interest well. At an unsuccessful peace conference in 208 the League demanded inter alia that Atintanis should be restored to Rome and the Ardiaei to Scerdilaidas and Pleuratus (thus excluding Philip from northern Illyris and its coast). In 206 the League made a separate peace with Philip and naturally made no demands on Rome's behalf as regards Illyris. The peace left Philip free to concentrate on Illyris, where only Epidamnus and Apollonia and farther south Oricum were hostile to him. However, the Romans moved first, probably in spring 205, sending io,ooo infantry, I,ooo cavalry and 35 warships to land at Epidamnus (Livy 29, 12, i). The Myzeqija was then flooded and the rivers were in spate. The first impact of the army was on the Parthini in the Shkumbi valley, through which it had to move in order to lay siege to Dimallum. The first objective of the Romans was to secure the line of communication between Epidamnus and Apollonia, which runs along the inland side of the plain on rising ground overlooked by Dimallum and other places situated on the range of Mt. Shpiragrit. The news reached Philip, who was probably in Macedonia, that the Roman force had come to Dyrrachium (Epidamnus), that the Parthini and other tribes in the vicinity were moved to hope for a revolutionary change in the situation, and that the Romans were besieging Dimallum.68 If Philip took the shortest route from Lower Macedonia, he went via Florina and through the Tsangon pass and descended to Antipatrea (Berat).69 He found that the enemy had withdrawn to Apollonia. He laid waste the territory of Apollonia, and he offered battle; but the Romans and the Apolloniates remained behind the walls. The Roman commander had sent part of his force by sea to Aetolia, asking the League to break its recently sworn treaty of peace with Macedon. The Aetolians refused. The Epirote League then took the initiative and negotiated a peace under which the Parthini, Dimallum, Bargullum and Eugenium were to belong to Rome and Atintanis was to belong to Macedon, if the Senate agreed (as it later did).70
Rome bargained well to get more than she possessed at the time,71 but Philip was able to keep the door open towards northern Illyris through holding Atintanis and he may still have controlled the southern Ardiaei, as no mention was made of any concessions to Pleuratus 72 who was a signatory on the Roman side (Livy 29, I2, I4). Nevertheless Rome retained the essential bases at Epidamnus, Apollonia and Oricum and the ability to defend them at sea and on land. One factor made this settlement more likely to lead to war than the settlement of 228. The buffer zones between the Roman sector and Macedonia had disappeared. For Philip had taken control of northern Illyris and of Dassaretis, and his acquisition of Atintanis made him an immediate neighbour of the Parthini in the upper Shkumbi valley. We may think Philip would have been wiser now to have withdrawn and disengaged. But we know from recent experience that that is difficult to achieve.
Even if Philip had disengaged in 205, it is doubtful if Rome would have played a different tune in 200. When we consider the springs of Roman and Macedonian policy in Illyris, we must remember that the Roman state and the Macedonian state alike were imperialist in the proper sense of the term, that is in desiring power, the power of commanding other states, and were not at any time in their history quietist or pacific states. This imperialistic quality in Rome is obvious from the first act in Illyris. If Rome had desired only to punish Teuta for the kidnapping or killing of Italian merchants or for the killing of her ambassador allegedly or actually at Teuta's command, Rome would have attacked Teuta's kingdom at its centre, that is in the region of Dubrovnik.
In fact, Teuta escaped lightly. What Rome took was not revenge on Teuta but command of a strategic area in Illyris, strategic not only in a military and naval sense but also in an economic sense; and to this command she clung consistently until in 200 her other major commitments were so much reduced that she could exploit her strategic position against Macedonia. If it is claimed that Rome took command of this strategic area in Illyris in order to stop Illyrian piracy, the facts are that neither earlier nor now nor later did the independence of this sector of Illyris prevent or even hinder substantially the practice of piracy by the Illyrians, nor did Rome ever use her bases in her sector of Illyris to try to stop Illyrian piratical expeditions. An accurate consideration of the geographical situation does much to make these conclusions clear and convincing. Between 228 and 205 Rome made enormous calls upon her manpower to meet her very numerous commitments elsewhere, and the fact that she did not exploit her position in Illyris in order to obtain further positions of command until 2II in the case of Greece and until 200 in that of Macedonia is due to the overstraining of her resources in other fields and not to a deep-seated pacificism or indifference. The deliberate choice of Rome in 228 to send envoys not to her new near-neighbours, Macedon and the Epirote League, but to the Aetolian League and the Achaean League, who were enemies of her near- neighbours, showed once and for all that she was concerned not with the establishment of pacific relations but with her future intentions.
Again Rome's choice of an occasion on which to open diplomatic negotiations with Macedon was not calculated to ease relations; for it was a straight demand in 2I7 to surrender to Rome a man who had been for many years an ally of Macedon. And her first positive step in Greece was to incite the Aetolian League to attack Macedon by the treaty of alliance which Rome herself initiated in 211. The Macedonian position was in many respects similar. From the outset Macedon took no steps to cultivate diplomatic relations with Rome. When Demetrius defected from Rome, Macedon accepted him as an ally and remained faithful to him even when Rome had defeated him in 219. In the same way in 2 7 or 216 Rome accepted as an ally Scerdilaidas who had attacked Macedon and been defeated by Macedon, and Rome remained faithful to him. In the early years Macedon was careful not to exacerbate Rome; thus she did not occupy Dassaretis or support Demetrius in his attacks on the Roman sector in Illyris or in his piratical expeditions.
The outbreak of the Second Punic War changed the situation on both sides. Rome was alarmed lest Macedon enter the War, and her demand to Philip to surrender Demetrius in 217 forced the issue for Macedon of compliance or resistance. Macedon chose resistance because she aspired to power not only in Illyris but elsewhere even as Rome did, and Macedon went on to occupy Dassaretis which invited friction with Rome as an immediate neighbour. Rome's support of Scerdilaidas after his attack on Macedon did nothing to lessen the chance of friction. Indeed it was at this point, if not earlier, that Philip of Macedon must have become convinced that Rome's intention was to make war sooner or later on Macedon. The decisive step which involved a serious risk of an armed clash between Rome and Macedon was taken by Philip of Macedon. His attempt to take Apollonia by surprise in 216 before the battle of Cannae and before any understanding with Carthage was only part of a plan to eliminate the Roman sector in Illyris, whether Rome reinforced her dependents there or not. It was chance rather than design which averted a clash on this occasion between Macedonian troops and Roman troops. Again, Macedon took the first step in constructing an alliance directed specifically against Rome in 215 ; the riposte to this was made by Rome in 211 when she allied herself with the Aetolian League against Macedon.
And it was Macedon's attack on Apollonia in 214 which led to the first clash between Macedonian troops and Roman troops; in this instance Macedon was certainly the aggressor. From then on until the peace of Phoenice in 205 hostilities continued if and when their other commitments allowed. There are some who excuse the desire for power, which is the basis of imperialism, on the grounds that it is a form of defence against domination or subjugation by some other power. When they apply this doctrine to Rome, they see in her acts this form of defence magnified to the nth degree. In this context, however, we must excuse not Rome but Macedon on these grounds. For in occupying a sector of Illyris Rome was not averting any threat of domination or destruction by Illyrians or by anyone else in 229; and again in keeping this sector of Illyris in the final settlement she was not averting such a threat by Macedon or anyone else in 205. On the other hand in 228 and in 2I7 Macedon had good reason to suspect that Rome intended to dominate her and in the end destroy her independence.
This basic fear came to the surface in the clause which Philip must have inspired in the Punic-Macedonian pact of 215, under which the first condition of a joint peace with Rome was (Plb. 7, 9, I3) 'that it shall not be possible ever for Rome to begin a war against Macedon'. Moreover Macedon's fear was a realistic one; for Rome's resources far exceeded Macedon's on any estimate,73 and Rome enrolled against Macedon not only in 216 the Ardiaean monarchy, which she had originally set out in 229 to chastise, but also in 211 a strong coalition of Greek states led by the Aetolian League. In the final analysis the incidents which we have been studying in Illyris arose from a Roman settlement in 229 which had little to do with precautions against piracy and everything to do with her own desire for power. If there had been no imperialistic power in that part of the Balkans, Rome would have stood still in Illyris until after 200, not from lack of desire to go further but from the pressure of other commitments. But there was another imperialistic power in the vicinity, aware of Rome's desire for power and alarmed by Rome's technique in diplomacy. Under these conditions a conflict between these two imperialistic states, both strong in the desire for power, was as likely to break out as the conflict between Athens and Sparta had been in 432.
The attendant circumstance of the second Punic War delayed the ultimate conflict but helped to incite actions on both sides which led to the state of war which culminated in 205. Once Carthage was defeated the delay was brief indeed, not because Macedon wanted the ultimate conflict in 200 but because Rome did.
Author(s): N. G. L. Hammond
Source: The Journal of Roman Studies, Vol. 58, Parts 1 and 2, (1968), pp. 1-21
Published by: Society for the Promotion of Roman Studies
Stable URL: www.jstor.org/stable/299691
In JRS LVI (I966) I gave a description of the Aoi Stena which was based on autopsy, and I discussed the campaigns of Rome against Philip V of Macedon in the years 200 to 198 B.C. In this paper I am concerned with the area farther north which Rome acquired in 229 B.C. and with the actions which took place there before 200 B.c. Many scholars have discussed Rome's early activities in Illyris but practically none of them has trodden the ground. My knowledge of most of the area may help me to advance more down-to- earth views of the extent of Rome's sector in Illyris and of Roman and Macedonian policies.
I include some new evidence on the position of Dimallum.1 The salient feature of Central Albania is the belt of coastal plain which extends from north of Lesh (Lissus) to north of the Gulf of Valona (see fig. i). The widest and richest part of this plain is in the Myzeqija, which extends southwards from Kavaje. The Myzeqija in particular is integral to the economy of Central Albania, the area which was called Southern Illyris in the third century B.C. The transhumance of sheep has always been practised in this part of the Balkans, and the coastal plain of Albania with months of very heavy rainfall in October and March affords exceptionally fine pasturage for the winter period.
The owners of the flocks which are driven from the plain to the mountains for pasture in the summer months have their homes not in the swampy plains but in the hills of the interior. On the other hand during the summer the coastal plain grows the cereals which are consumed in the hinterland, where there is a deficiency of arable land.2 Thus the lord of the plains has a stranglehold on the economy of Central Albania. In the third century B.C. the Myzeqija was held by the two Greek cities, Apollonia and Epidamnus. From a military point of view the coastal plain is difficult to defend. In the spring and early summer and again in the autumn it is so heavily flooded that movement by an army across it is not possible,3 and for purposes of movement and communication an occupying force must possess the rising ground which runs from Apollonia in the south through Kug, Lushnje and Kavaje to Epidamnus in the north. Even so there is no possibility of defence in depth against an enemy who attacks from the hinterland, and the obvious policy for an occupying power would be to hold or threaten the main route of entry from the east, namely the Shkumbi valley, and also to hold the range on the southern rim of the plain. Of this range Mt. Shpiragrit is the most important part because it faces Berat, through which the subsidiary routes from the east pass as if through a funnel. I became aware of these economic and military factors when I travelled in late March and early April from Durres (Epidamnus) via Kavaje to Elbasan in the Shkumbi valley, seeing the river in spate, the coastal plain flooded and the numerous herds of sheep assembling on the higher ground for the move eastwards, and then again in the summer on other occasions when I went from Elbasan to Berat and from there to Poyani (Apollonia), and when I walked from Byllis in the Aous valley to Berat via Metoh, taking six and a half hours.
The defence of the coastal plain against an enemy attacking either from the south or from the north raises different problems. Here the occupying force has depth but his front is narrow and can be easily turned; it is therefore necessary to protect the open flank by holding some of the hill country. In December I940 the Italian army in central Albania held a position facing south and covering what General Papagos called 'the Kelcyre and Tepelene junctions ', i.e. the Aoi Stena together with the passes at either end of the Stena (see JRS LVI, 40, fig. 2). This was always the strongest position for an army in Illyris to hold against an attack delivered from within Epirus. Failing this, such an army would be wise to hold the hilly country just south of and to the east of Apollonia. Both at the Aoi Stena and at Apollonia a flank guard is needed, reaching into the hills towards Mt. Tomor, and covering Berat.5 Defence against an enemy attacking from the north is made difficult by the deep re-entrant through the upper valley of the Black Drin, which rises in Lake Ochrid.
In consequence it is necessary either to hold the area north of the Drin, that is on a line running inland from Scodra, or to hold the hill country to the north of the Shkumbi valley running inland from Durres (Epidamnus). To the mariner Central Albania is distinctly unattractive. The coast between the Gulf of Valona and Ulcinj is generally flat and featureless; it offers few roadsteads and fewer harbours; and it is rendered dangerous not only by the southwesterly gales and in places by the northerly ' Bora ' but also by the changing positions of offshore banks which extend out to sea as far as three miles. The advice of The Mediterranean Pilot 6 to the sailor setting off on the 40 mile stretch from the Bay of Valona to Kep i Gagji is brief and to the point: ' the utmost caution should be exercised when approaching this portion of the coast and at night it should be given a wide berth '. 6 Indeed for the hundred miles between the Gulf of Valona and Ulcinj there are only two sets of harbours. The first was controlled in ancient times by Epidamnus. It consists of Durres harbour, formed by two moles, and of Pali Bay, 6 miles to the north. Each offers shelter from some winds. Between them runs a high ridge (the name Dyrrachium was derived from it), which is a conspicuous feature ' showing from afar like an island with five heads ' and of great value for making a landfall.7
The second set is in the Gulf of the Drin : the mouth of the Drin, navigable to Lesh; the harbour of Shen-Gjin or S. Giovanni di Medua, entered by a narrow channel and suitable for small vessels (the Nymphaeum of Caesar, BC 3, 26, 4, ' qui ab Africo tegebatur, ab Austro non erat tutus '); the navigable river Buene, giving access to the Lake of Scodra, which is only 6 metres above sea level (Livy's ' Barbanna ... ex Labeatide palude oriens', 44, 31, 3); 8 and the small harbour of Ulcinj (the Roman Ulcinium), dangerous in southeasterly winds. Between them this set of harbours offer shelter from any wind. In antiquity Lissus had a harbour but this has silted up and disappeared. And in the south the mouth of the Aous (Vijose), navigable as far as Apollonia, was used but only when conditions of weather were favourable (cf. Plu., Caes. 38).
But then, as now, Epidamnus was the most important station on the coast; anyone holding Epidamnus and denying access to its harbours can make navigation along or towards this coast under sail extremely difficult in bad weather.9 In the days of sail this coast was important for reasons which no longer obtain (see fig. 2). The direct crossing of the Straits of Otranto from the all-weather harbour of Oricum in the Bay of Valona to the heel of Italy was rendered dangerous by exposure to the Bora and the southwesterlies, which create very rough, steep seas, and by the proximity of the formidable Acroceraunian coast. In consequence sailors preferred to cross the Ionian Gulf from Epidamnus to Bari or Brindisi.10 Thus Epidamnus served Greeks trading to the west and Italians trading to the east. At the same time it was the main port of call in the Ionian Gulf for shipping up and down the Adriatic Sea, and so was of paramount importance to those Illyrians who concerned themselves with the sea. The coast of Central Albania has never bred a seafaring tradition, not surprisingly in view of its nature, but the islands off the river Narenta (the ancient Naro) and the fjords such as that of Kotor (' Sinus Rhizonicus ') produced Illyrian navies from time to time. The Illyrian warship, whether developed by the Liburnians or the Ardiaeans, was small and extremely fast under oar and under sail, and it was more fitted for raiding than for set engagements at sea.11
Such Illyrian navies were always anxious to gain possession of Epidamnus. The Liburnians held the site in the seventh century, Monunius early in the third century; and the Ardiaeans, whose naval power was based on the area of the islands and fjords but extended to the southern Ardiaeans by the Gulf of the Drin, were inevitably drawn towards it in the latter part of the third century. The first occasion of Roman intervention across the Adriatic Sea was a naval occasion. It was prompted by the scale and success of the piratical Illyrian raiders, who had made a habit of raiding the coasts of Elis and Messenia, defeated an Aetolian army at Medeon in Acarnania in 231, and then in 230 not only raided the coasts of Elis and Messenia once more but captured Phoenice in Epirus and compelled the Epirote League to enter into alliance. More than piracy at sea was involved; for a land force under Scerdilaidas invaded Epirus at such short notice that we may infer that the Illyrians held the area between Berat and the lower Aous valley as well as 'some parts of Epirus ' (App., Illyr. 7, T-rS 'HrrEipovu rvca).12 In 230 this expanding Illyrian state was in alliance with Macedon, Acarnania and Epirus (Plb. 2, 6, 9).
In spring 229 B.C. an Illyrian fleet, assisted by Acarnanian ships, defeated the fleets of the Aetolian and Achaean Leagues, gained possession of Corcyra and laid siege to Epidamnus. The Illyrians, led by Teuta, widow of Agron, king of the Ardiaei,13 were within an ace of controlling the entire coast from Dubrovnik to the mouth of the Corinthian Gulf.14 Meanwhile Rome was on the move. Italians trading by sea towards the east had been molested repeatedly in recent years by the Illyrians, and in the course of 230 B.C. more Italians than usual had been intercepted at sea, robbed, kidnapped or killed.15 The Senate sent ambassadors to Queen Teuta late in 230 B.C. The conversations ended in a fracas, and one of the ambassadors was assassinated on the way back.16 The Senate then began to organize a naval and military force. However, it did not declare its intentions to Teuta, and she operated during the first part of the campaigning season of 229 B.C. as if her position was not under threat from Rome.
In the first Punic War Rome had become painfully aware of sea power, and she had realized that the long coast of Italy was vulnerable to seaborne attack. Hamilcar Barca had raided Locri and other places in south Italy, and in 246 Rome had planted a colony at Brundisium, which faces the Ionian Gulf. During the war Italy had traded very intensively across the Ionian Gulf,17 partly because the Carthaginian raiders had not penetrated into these waters. If the Illyrians now succeeded in their aims, they would close the mouth of the Adriatic Sea and control the approaches to South Italy through the Ionian Gulf from their bases on the coast and on Corcyra. For a variety of reasons 18 the Senate was not prepared to let this happen. In the second part of the campaigning season Rome sent into action 200 ships, 20,000 infantry and 2,000 cavalry-a large force, but not unduly large for the task the Senate envisaged.19 The campaign was brilliantly conducted (Plb. 2, i).
The Roman fleet sailed first to Corcyra, where Teuta's commander, Demetrius of Pharos, having previously come to a secret understanding with Rome, handed over the island. It then sailed to Apollonia, which joined the Roman cause, and met there the Roman army which had been brought over from Italy. They advanced next towards Epidamnus, the fleet sailing along the coast and the army probably moving along the eastern side of the Myzeqija plain, where the 2,000 cavalry had excellent ground for manoeuvre and plenty of pasture. The Illyrians abandoned the siege of Epidamnus, which now joined the Roman cause. The Roman army then turned further inland and overran the Ardiaei,20 probably in the lowlands and the hinterland between the Mati and the Drin,21 and the Roman fleet, sailing up the coast and taking some places by assault, relieved the island of Issa, which the Illyrians were besieging. Army and fleet then returned to Epidamnus. Casualties had been suffered during the last phase only, for instance at an unidentified place called Noutria. The spectacular success of Roman arms was due to careful planning, the suddenness of the unheralded attack (critics might have called it a treacherous attack),22 the double- dealing of Demetrius of Pharos and the accession of Apollonia.
The strategy was good: the seizure of Corcyra cut off the possibility of naval reinforcements from Acarnania, the occupation of Apollonia and the Myzeqija cut off the possibility of land reinforcements from Epirus, and the rolling up of the Illyrians on a narrow front with naval support and a mobile cavalry force cost Rome few casualties. The whole campaign was over so quickly that Teuta's other ally, Macedon, could not have intervened in time to forestall the Illyrian collapse, even if she had wished to do so. Victory won, Rome had several courses open to her. She could pursue the defeated Illyrians to Arbon and Rhizon, where they had taken refuge, and give a knock-out blow to the Ardiaean monarchy. She could occupy with troops or with colonists the strategic points in the areas she had overrun, e.g. Corcyra, Apollonia, Epidamnus, Issa and Pharos, and thus control the approaches by sea to the outlet of the Adriatic Sea; she would then hold safe ports on the eastern coast of that sea to balance Brundisium.
She could treat the whole operation as a punitive raid and withdraw her forces, having taught Teuta a salutary lesson and leaving her friends in Illyris and the islands to maintain their independence as they had done hitherto. In fact she chose a course intermediate between the second and the third of those which I have suggested. She set up Demetrius as ruler of a kingdom 23 based on Pharos and including most of the Illyrians overrun by Rome (particularly, I imagine, in the area of Scodra), and she no doubt hoped he would act as a buffer between Rome and the Ardiaean monarchy. Issa in the north and Corcyra in the south became in effect dependents of Rome, Issa providing a contingent later to the Roman navy. On the mainland, although the Romans were approached by the envoys of a number of tribes during the advance northwards from Epidamnus (Plb. 2, I, I , cavyjj4av-rTcv 6 TwpEcyUEvrTOV ca-rTOis KCal XrAEovc1v), they accepted only the Parthini and the Atintani into what was or became a position of dependency. They already had Epidamnus and Apollonia on their side. During the winter of 229-8 one of the consuls stayed with forty ships and conscripted a force " from the surrounding states " which kept an eye on the Ardiaei and the other (Illyrians) who had submitted to Rome.
In spring 228 B.C. Teuta opened negotiations with Rome and concluded a treaty under which she undertook to pay an indemnity, evacuate all " Illyris " except for a few districts, and not sail beyond Lissus with more than two lembi even unarmed.24 On the conclusion of the treaty the Roman consul sent envoys to the Aetolian and Achaean Leagues, explaining the reasons for Rome's initiative and the terms of the treaty she had concluded with the Illyrians. They doubtless emphasized the fact that they had delivered the Greek cities of Corcyra, Apollonia, Epidamnus and Issa from the common enemy, the Illyrians. The envoys received a friendly welcome and then returned to Corcyra. The Roman settlement needs some clarification on the geographical side (see fig. i with Inset). Epidamnus and Apollonia were both wealthy Greek city-states with a very copious silver coinage which had a wide circulation. There is no doubt that each of them had an extensive territory. We have more information about Apollonia.
In her coinage from 229 B.C. onwards she showed the Nymphaeum, which we know was close to her frontier with Byllis and Amantia, independent states which held the north side of the Aous valley above the vicinity of Romes.25 The mass of Apollonia's territory was then not to the south but in the Mizeqija, extending northwards certainly to the Semeni (Apsus) 26 and probably to the Shkumbi (Genusus) and inland towards Mt. Shpiragrit. Epidamnus is likely to have controlled both the plain of the Arzen river and the northern part of the Myzeqija as far south at least as the north bank of the Shkumbi river. As Asparagium (probably Rogozine) was in her territory,27 she formed the outlet through which all trade following the north side of the Shkumbi valley had to pass. The Parthini were a tribe adjacent to Epidamnus (Appian, BC 5, 75, gevos 'ETrBl&a,vcp -rapoiKov; cf. Dio. 41, 49, 2). When Pompey moved from Epidamnus to the north bank of the Semeni (Apsus) and Caesar moved from Apollonia to the south bank of the Semeni (Apsus), they faced each other probably at Ku9. When Caesar entered the Shkumbi (Genusus) valley probably through the sink southwest of Elbasan in order to pursue Pompey down the valley to Asparagium, he captured on his way a Parthinian town which Pompey had garrisoned (Caesar, BC 3, 41, I, and Dio 4I, 47, I). Thus the Parthini held at least the middle valley of the Shkumbi (Genusus) river. Moreover, as Pliny reported that the Parthini had the Dassaretii 'behind them ' (NH 3, I45) and as the Dassaretii extended as far east as at least Lychnidus (Livy 43, 9, 7), the Parthini held the upper valley too.28 The Atintani, as I have shown elsewhere,29 have nothing to do with the Atintanes in the upper Drin valley in Epirus but are an Illyrian tribe about one day's journey from Epidamnus in the direction of Macedonia, and occupying very high country, visible from Epidamnus and near the (Macedonian) frontier of Illyris (Polyaenus 4, 11, 4).
This tribe evidently held the region of ?(ermenike, which extends from just north of Elbasan to the watershed of the highest reaches of the Black Drin. The dispositions of Rome in the northern part of Central Albania are now clear. First Demetrius of Pharos as suzerain of some Illyrian tribes from north of Epidamnus to the vicinity perhaps of Scodra acted as a buffer between the Ardiaean monarchy and the zone of direct dependence on Rome. Second, a continuous line was held by the dependents of Rome-Epidamnus, the Parthini and the Atintani-from the coast to the highest westerly sources of the Black Drin and in sufficient depth to be defensible against attack from the north. At the same time this defensive position not only cut off the Ardiaean monarchy from its ally, Macedon, since the Dardani were pressing down upon Pelagonia at this time,30 but also blocked the main route from Macedonia to the Adriatic coast, that later followed by the Via Egnatia. The chief threat to this defensive position might come from Dassaretis, the high territory between the two arms of the Semeni river, which are called the Devoli and the Osum, and extending inland to the main watershed through which the Tsangon pass and the pass of Vatokhorion lead into the Macedonian canton of Orestis. In 230 B.C. Dassaretis was evidently under the control of the Illyrians, because Agron held ' parts of Epirus ' and Scerdilaidas was able to move swiftly to Antigonea in Epirus at a time when neither Agron nor Scerdilaidas held Apollonia (App., Illyr. 7 and Plb.
2, 5, 6).
The defeat of Teuta's forces and Rome's alliance with Epidamnus, the Parthini and the Atintani left Dassaretis without any direct political affinity. Rome had the good sense not to include it in the zone of direct dependence upon herself. Thus Dassaretis became an independent area and formed a buffer between Macedon and Rome. In the south the zone of direct dependence ended with the territory of Apollonia (the Apolloniatis) which had a footing on the southern bank of the Vijose (Aous) but did not include Aulon, Byllis, Amantia or Oricum. These four small states formed an independent or neutral group situated between Rome and the Epirote League, of which the northernmost cantons were Chaonia and Parauaea. The boundary to the southeast will be defined more closely when we consider the position of Dimallum.31 Rome now had access by right to harbours in Demetrius' realm, at Lissus, Epidamnus, and the Vijose river, then navigable as far inland as Apollonia. On the other hand, the Illyrians of Teuta's kingdom were not only debarred by treaty from sailing south of Lissus with more than two unarmed lembi, but were in effect dependent on Rome's favour for the peaceful use of these indispensable harbours.
The economic interest of the neighbouring coastal areas, that round Scodra32 in the north and that round the Gulf of Valona in the south, drew them strongly to Rome, because Italy offered an excellent market for their products. The inland areas of the Parthini, the Atintani and the Dassaretii depended, as we have seen, upon the Myzeqija for winter pastures and for cereals and also upon Epidamnus and Apollonia for import and export, so that they too had an economic interest in adhering to the political power which was lord of the plain. It is customary to call the zone of direct dependence on Rome 'the Roman protectorate ', a vaguely benevolent and flattering euphemism for an extremely shrewd extension of Roman power. Rome's dependents, whether they enjoyed the title of ally or subject,33 were to be her dependents for good and not free agents, able to transfer their allegiance to other states with impunity. We do not know whether, once the treaty with Teuta was completed, Rome left any troops at Epidamnus or Apollonia, but the point is unimportant, since in a matter of hours she could send an army across to her treaty-ports in order to hold the position defended by her subjects. To use a modern phrase, she intended to maintain ' not a military presence but a military capability-a capability to get troops there if they were needed '.
Having completed her dispositions, Rome disregarded Macedon, the Epirote League and Acarnania, the very states which were most immediately affected by the appearance of Rome as a new constellation on the eastern shore of the Adriatic Sea. Instead, Rome sent embassies to the enemies of those three states, that is to the Aetolian and Achaean Leagues, which had at least the superficial merit of being in opposition to Teuta's Illyrians, and later to Corinth and Athens, both hostile to Macedon. This deliberate and public move by Rome made it clear to Macedon and the Greeks that in any war in the southern Balkans Rome's sympathies would lie initially with Aetolia and Achaea and against Macedon, Epirus and Acarnania. The announcement of this alignment, coupled with the astute organization of Roman interests in southern Illyris, could not fail to cause alarm in Macedon, Epirus and Acarnania. It would be naive to suppose that Rome was unaware of the fact.34
The Roman successes at sea and on land had damaged the prestige but not the power of the Ardiaean monarchy, which was based upon Dubrovnik and the Dalmatian coast; 35 and even within Illyris, an area which began probably around Scodra, a few districts were still in Teuta's hands (Plb. 2, 12, 3). Demetrius of Pharos, who had the skill of a Perdiccas in changing sides, became the successor of Teuta on her death and married Triteuta, mother of the infant king, Pinnes, so that the first or northernmost part of the Roman settlement collapsed completely. There was no longer any buffer between the Ardiaean monarchy and the zone of direct dependence on Rome. Meanwhile Rome became deeply involved in a war with the Gauls in the Po valley which lasted from 225 to 222 B.C., and within these years Demetrius advanced both at sea and on land, engaging in piracy south of the Lissus line and winning over the Atintani from Rome.36 The latter step was the more important; for Demetrius was thereby opening the door to co-operation with Antigonus Doson, king of Macedon, and also making infiltration into Dassaretis possible if he gained Antigonus' co-operation. In 223 B.C.
Demetrius accompanied Antigonus on his invasion of the Peloponnese and in 222 B.C. Demetrius' force of i,6oo Illyrians played an important part in the battle of Sellasia. Polybius enumerates them among Macedon's allies (2, 65, 4).37 Early in 220 Demetrius and Scerdilaidas sailed south of Lissus with 90 lembi and carried their raids into the Cyclades (Plb. 3, i6, 3 ; cf. 4, i6, 6-9, and 4, 19, 7-9). In this year a Roman fleet suppressed some pirates, probably the Istrians (App., Illyr. 8).38 In this year also Demetrius was ravaging the territory of ' the states in Illyris subject to Rome ' and was trying to subdue them (Plb. 3, i6, 3).39 In order to do this effectively, he came through the gap opened by the defection of the Atintani, entered Dassaretis and attacked the Parthini and the Apolloniates from there. In 2i9 the Romans sent the two consuls with an army which was probably as large as that of 229 B.C. to punish Demetrius and his Illyrian collaborators (Appian, ibid.). The attack was again unheralded, but Demetrius had laid his plans in advance. He placed a considerable garrison with suitable supplies in Dimallum, evidently a town of which he already had possession, and he brought about changes of government in his favour 'in the other states ', which presumably now for the first time he took under his influence (Plb. 3, I8, i).40 Having established Dimallum as a strong point, he went to Pharos and prepared to hold Pharos with a picked force. On landing in Illyris, the Romans attacked Dimallum first; knowing it was thought to be impregnable, the consuls hoped to capture it and spread alarm among the Illyrians.
Dimallum fell in a week. Envoys came in from ' all the states ' with offers of submission. The consuls made appropriate agreements in each case and then moved on to attack Pharos. The island fell, but Demetrius escaped. At Actium he joined Philip of Macedon, who had inherited Antigonus' friendship with him and had himself visited Scerdilaidas in the winter of 220-219. The Romans crossed over from Pharos to Illyris, gained control of the rest of Illyris and made a settlement of all its affairs. A triumph was accorded to the consuls on their return to Rome at the end of the summer.41 The Roman intervention of 2I9 needs little explanation. Demetrius had flouted the Roman settlement of 229 all along the line by deserting with the dependents Rome had given him, by detaching the Atintani from their allegiance to Rome, by sailing in strength south of Lissus, by ravaging the territories of Rome's dependents (evidently those of the Parthini, Epidamnus and Apollonia) and by trying to subdue them and so eliminate Rome's holding on the mainland. The Senate must have seen that inaction would be fatal to her position in Illyris. Once in control, with Macedon, Epirus and Acarnania as his accomplices, if not his formal allies, Demetrius rather than Rome would control the Ionian Gulf and the Straits of Otranto at the very time when a renewal of war with Carthage might be expected. 42
The terms of the Roman settlement in 219 are not described in our sources. It should be noted that even less than in 229 did Rome attack the basis of Illyrian power which lay farther north, and it may be doubted whether on this occasion her forces penetrated even to Scodra. Moreover, no steps were taken against Scerdilaidas, though he also had sailed south of Lissus and he also was an ally of Philip. Clearly Rome had no desire at this time to go too far or to increase her commitment in Illyris. Wherever she went, she certainly put pro-Roman parties in power, but it seems as if Dimallum may have been the only addition she made to the number of dependents she had had since 229; for in 215, when Hannibal and Philip made an alliance, one of their aims was to prevent Rome from being in control of Corcyra, of Apollonia and Epidamnus, of Pharos, of ' Dimale' and the Parthini, and of Atintanis (Plb. 7, 9, I3). Thus in 219 the Senate seems to have been content to restore her full control of the 229 group of states, which interposed a barrier between the Illyrians and the Macedonians, blocked the easiest route from Macedonia into this group of states, and left independent or buffer areas north of Epidamnus, in Dassaretis and in the Gulf of Valona and its hinterland.
No doubt she reiterated the ban on Illyrian lembi sailing south of Lissus. Once again Rome did nothing either to ease or to exacerbate relations with Macedon or Epirus. Yet her determination to stand firm in Illyris was itself alarming in view of Rome's record and reputation. On the other side Antigonus' and Philip's acceptance of Demetrius as an ally in good fortune and in bad made it equally clear that Macedon favoured Rome's enemies in Illyris and was prepared to show it.43 In modern terminology Rome and Macedon were now engaged in a ' cold war '. We must pause in the narrative to consider the position of Dimallum. It has been placed by scholars on the coast, not on the coast, near Epidamnus, in the territory of the Parthini and not in the territory of the Parthini.44 Consideration of the texts and of the geographical conditions can reduce the number of alternatives. The coast between Epidamnus and the Gulf of Valona is low and swampy; a broken line of low hills runs parallel to the coast, except at Barderoll where the hill comes down to the coast itself; and then inland of the line of hills there is a swampy plain once again. Now Dimallum was so strongly situated as to be thought impregnable (Plb. 3, i8, 3). It would be difficult to find even at Barderoll an impregnable site along this coast. Moreover, Dimallum was in dispute between Rome and Macedon (Livy 29, i2, 13); we therefore need a place on the Macedonian side of the Roman block of territory and not on this coast which belonged mainly (and had done for centuries) to Epidamnus and Apollonia. As regards the Parthini and Dimallum the one is a tribal state, subsuming under its name the ' urbes Parthinorum ' (Livy 43, 23, 6), and the other is a city, probably autonomous, negotiable as a separate entity between Rome and Macedon.
The two are always mentioned as separate units by Polybius and Livy Plb. 7, 9, I3, UIr8' ETvati Pcoiaious KUvpOUv KEpKupaciov JirS' 'Aro ovIcTrov KiI 'ETricbavicov tr&8 Oa)povu pTr6s Atip&?aAr1s KCI TapeE0ivcov lPr6' 'A-rVravia ; Livy 29, I2, 3, ' Parthinosque et propinquas gentes alias motas esse ad spem novandi res, Dimallumque oppugnare (sc. Romanos) ' ; Livy 29, 12, 13 'ut Parthini et Dimallum et Bargullum et Eugenium Romanorum essent '. I conclude then that Dimallum (and also Bargullum and Eugenium) are not Parthinian towns 45 nor in Parthinian territory, which extended from the upper valley of the Shkumbi to somewhere inland of Asparagium (probably Rogozine), which was in the territory of Epidamnus. It follows then that Dimallum lay either north of the line of Epidamnian and Parthinian territory, that is facing northern Illyris, or south of Parthinian territory and east of Apolloniatis, that is facing Dassaretis. Those who place Dimallum near Epidamnus do not rely on any ancient evidence. But there is a clue to the contrary in Livy 29, I2, 5, where the Roman consul who had gone from Epidamnus to Dimallum, abandoning the siege, retired to Apollonia (' quo Sempronius se receperat '), and from this I infer that Dimallum lay nearer to Apollonia than to Epidamnus.
But the decisive point is that Dimallum was in dispute between Rome and Macedon at a time when Macedon held nothing north of Epidamnus but did hold Dassaretis. Dimallum, then, lay between Apolloniatis and Dassaretis, to the south of the Shkumbi valley, and therefore probably on the range of Mt. Shpiragrit. ' La decouverte de la cite illyrienne de Dimale' is the title of an article published by Burhan Dautaj in Studia Albanica 1965, I, 65-7I, of which an offprint was very kindly sent to me in 1967 by Professor Frano Prendi. In I963 and I964 Dautaj excavated the fortress of Krotine,46 which is situated on ' a fine peak' of the Shpiragrit range (see fig. 3). This peak, being 404 metres above sea-level, is the highest of the western outliers of the range; it has ravines on three sides of it, and there are traces of a circuit-wall on the steep slopes. It was an exceptionally strong place and might well have been thought impregnable. There are two summits inside the circuit, the higher being the acropolis and the lower providing accommodation for three-quarters of the population. Praschniker estimated the circuit of the wall to be 2,400 metres. When the acropolis was excavated the most important discovery was that of numerous stamped tiles of Hellenistic date, rectangular, red to grey in colour, some long and narrow with squared edges, others broad with slightly rounded edges. Within an excavated area of some 500 square metres Dautaj found no less than 150 tiles stamped on the edge with the monogram shown in fig. 4(a), and he deduced that these tiles were local to the site. In addition to the monogram i6 tiles were stamped with the name NEYTQP or NE22TQP in the nominative, which appears to be that of a local potter since it is not found on stamped tiles elsewhere in Albania. Four tiles bear stamped monograms and the stamped word AIMAAAITAN, cf. (fig. 4(b).
Twelve tiles are stamped HPAIQN, a name which appears also on tiles of Apollonia; but these tiles are smaller and poorly made compared with those of Apollonia. They were made presumably for members of a cult in honour of Hera or for a group so-called, like the Heraeis of early Megara.47 One tile had a genitive plural ending in - OANIQN. Names in the genitive, evidently of magistrates, were stamped on other tiles: AMYNTA, APMHNOC, API:TOMENEOS and - MAXOY.48 The word AIMAAAITAN written with the broken-bar alpha has the same form of genitive and much the same lettering as AlX:ITAN on coins of Lissus attributable to some time within the period 250-200 B.C.49 The name of the city at this time was evidently Ai|caXAoS or AhiaAAXov as we find Dimallum in Livy. It is a Greek word meaning ' of double fleece ', very suitable to a place overlooking the rich sheep-pastures of the Myzeqija.50 Another independent city, Eugenium, in this region has a Greek name, used of a type of laurel.51 Coins were found on the acropolis of Dimallum. The earliest of these were coins of Epidamnus dating to the period 320-270 B.C. The city may have been founded c. 290 B.C. by Pyrrhus or his successors, to control the route along the side of the plain to the sink between the Semeni and the Shkumbi rivers, just as Antigonea was founded by him farther south.52 In any case it seems beyond doubt that this is the city into which Demetrius put a garrison in 219 when he had overrun Dassaretis.
We do not know its earlier history. It may have been taken by Rome already in 229, because it threatened the route along the side of the Myzeqija, and it may then have gone over to or been acquired by Demetrius at some time before 219. In any case, once captured by assault in 219, Dimallum was regarded by Rome as her possession, as Plb. 7, 9, 13 shows. The Roman settlement of 219 was treated with scant respect. In 218 Demetrius may have indulged in raiding again (App., Illyr. 8 fin.), and Scerdilaidas certainly sailed south of Lissus and supported Philip in his naval enterprises that year. Rome may have sent a punitive force against Demetrius at or near Pharos, and she kept on good terms with the young king, Pinnes, ruler of the Ardiaean state (App., Illyr. ibid.). Scerdilaidas had troubles in Illyris with' city-dynasts ' (Plb. 5, 4, 3 -roAtluvv&orras) and these troubles limited the help he gave to Philip in this year; the ' city-dynasts ' may have been persons set up under the Roman settlement of 219, but many other explanations for their activities are possible in an area so split by tribal feuds.
In 217 Rome sent envoys to Philip of Macedon demanding the surrender of Demetrius (Livy 22, 33, 3). If this was her first diplomatic contact with Macedon, it was an inauspicious one; for the demand was of the kind made to a subject state and not to an equal and independent state, and Philip naturally refused to comply. At this time Pinnes was visited by Roman envoys, who demanded arrears of 'tribute ', evidently a war-indemnity payable in instalments (Livy, ibid.); one wonders if they got any change out of Pinnes. These actions are probably to be connected with Rome's alarm when Hannibal reached Picenum and then Apulia on the Adriatic coast, having won the battle of Lake Trasimene in summer 217. At the same time Scerdilaidas had struck out on his own. By sea he sent 15 lembi south of Lissus to attack Philip's friends off Leucas (Plb. 5, 95, i f., and 5, IOI) ; by land he plundered Pissaeum, a Macedonian town in Pelagonia, won over by threats or promises 53 three towns in Dassaretis, namely Antipatrea, Chrysondyon and Gertous, and overran much of that part of Macedonia which was conterminous with Pelagonia and Dassaretis (Plb. 5, io8, I-2).
As Demetrius had done in 220, Scerdilaidas passed through the high country by Lake Lychnitis between the Parthinian territory of the Roman sector and Macedonia and so entered Dassaretis, which Rome had left as a weak and unassigned buffer area in her settlement of 219. Now Scerdilaidas used Dassaretis as a springboard for attacking Macedonia, as the Romans were to do later in 200 B.C.54 Philip's counter-stroke was the obvious one, to cut the line of entry from northern Illyris into Dassaretis by capturing the district round Lake Lychnitis. Philip therefore ' won back' (Plb. 5, io8, 8, avEKTilacro) Antipatrea, Chrysondyon and Gertous, and then ' captured' (KcTErEAasTo) a number of places: Creonium and Gerous in Dassaretis; of the people round Lake Lychnitis Enchelanae, Cerax, Sation and Boii; in the territory of the Caloecini Bantia; and also of the so-called Pisantini Orgyssus.55 The winter of 217 was now upon him and he disbanded his troops. He had put a stopper on Scerdilaidas, but he had done something which might be more dangerous than anything Scerdilaidas could ever do. He had made himself the immediate neighbour of Rome by occupying Dassaretis. On her side Rome accepted the defeated Scerdilaidas as an ally, just as Philip had recently accepted Demetrius. Polybius believed that in acting against Scerdilaidas in 2I7 Philip wished to consolidate his position in Illyris and then to cross over to Italy, a course advised especially by Demetrius (Plb. 5, Ioi, 6 f. ; 5, 105, i and 5 ; 5, io8, 4 f.).
It is possible that Polybius formed this belief in view of Philip's actions in 216 and 214; for in 217, whatever ideas he did or did not have about Italy, Philip was compelled by his interests at home to stop Scerdilaidas' raids on Macedonian territory. When Philip occupied Dassaretis, then the question of a move westwards became possible for the first time. But to anyone harbouring such an ambition the strength of Rome's position in Illyris was clear. As we have seen, the coast offers very few harbours. Not one of them was in Philip's hands. Scerdilaidas, who was now seeking help from Rome, held the coast from Lissus northwards; Epidamnus and Apollonia held the coast as far as the entry to the Gulf of Valona; the buffer states of Byllis, Amantia and Oricum held the coast of the Gulf; the Acroceraunian shore offered no anchorages at all; and even the harbours of the Epirote League by Buthrotum were blanketed by Rome's naval base on Corcyra. Yet, unless Philip could establish himself on this coast, he had little or no chance of gaining control of Illyris; for so long as the harbours of Apollonia and Epidamnus were open and accessible to reinforcements and supplies, Philip could not hope to reduce the cities by blockade or by siege. Any plan to cross over to Italy would be pointless, if Philip did not already hold ports on the coast of Illyris from which to despatch, supply and reinforce any troops he might want to land in Italy.56 In addition, any lembi which he proposed to build would not be capable of meeting Rome's quinqueremes in a naval action; therefore he could not challenge Rome's naval supremacy. So far as Illyris was concerned, he might be able to deliver an assault force unexpectedly at a strategic place,57 provided that there were no quinqueremes within range, and so capture a port which he could then defend. The obvious target was Apollonia; it was farthest away from Scerdilaidas' bases in the Adriatic Sea, it was closest to Macedonian troops in Dassaretis, and the navigable river, the Aous, was not only easy for an assault force to seize by surprise but also would give harbourage to an attacking fleet.
Philip made two attempts. In the winter of 217/6 he used Illyrian shipwrights to build him a hundred lembi in Macedonian ports. Setting out from Macedonia he rounded the Peloponnese and reached Leucas in early summer 216. There he ascertained that the Roman fleet was lying off Lilybaeum in western Sicily. He sailed on northwards, passed Corcyra and at night time was close to the mouth of the river Aous and ready to deliver his assault force,58 when the report reached him that Roman quinqueremes were actually crossing over, bound for Apollonia (Plb. 5, 109-I IO). It appears that Scerdilaidas, knowing that Illyrian shipwrights were being employed by Philip, had informed Rome and had asked for help in order to stiffen his own naval forces, and it was in answer to this request that some quinqueremes were detached from the fleet at Lilybaeum and sent across the Ionian Gulf (Plb. 5, IIo, 8-9). The timing was in fact fortuitous. But Philip could not know that, nor could he tell how many quinqueremes were on the way (in fact there were only ten). He therefore withdrew hastily.59 An actual clash between Macedonian and Roman troops was thus averted. The expedition had two effects. In 215 Philip approached Hannibal and obtained an alliance, under which, if they made peace with Rome, one condition would be that Rome would never make war on Macedon and Rome should no longer be in control of Corcyra, or of Apollonia and Epidamnus, or of Pharos, or of Dimallum and the Parthini, or of Atintanis, and should surrender to Demetrius of Pharos those of his friends who were interned on Roman soil (Plb. 7, 9, I3). This would give Philip what he wanted most, a guarantee against Roman attack and the removal of Roman power from Illyris. Meanwhile he may have hoped for naval help from Carthage in another attack on Apollonia and Hannibal may have hoped for some military help in Italy from Macedon.
But nothing specific was arranged and nothing came of it. The phrase in Polybius' report of the treaty (7, 9, II) pOT]i0e'UETE 5E IITV cS &v XpEia 1i Kiai cos &v vp90 vilocopEv provided for a future rather than an immediate contingency.60 On the Roman side the attempt by Philip on Apollonia had revealed Philip's intentions; moreover, the treaty between Philip and Hannibal became known to Rome when Philip's envoy fell into Roman hands. Consequently Rome placed a fleet at Tarentum to guard the coast and watch Macedon (Livy 23, 38, 9). A state of war now existed in fact between Macedon and Rome. In 214 Hannibal and Philip undertook concerted, if not simultaneous, actions against Tarentum on the one hand and Apollonia on the other hand. The action at Tarentum engaged the attention of the Roman fleet and enabled Philip to reach the Aous river without being intercepted. This time he sailed up the Aous with izo lembi and attacked the town. On the landward side an army which had marched up through the territory of his ally the Epirote League,61 and also probably some troops from Dassaretis, joined in the attack.
His siege engines tried to breach the circuit wall, which exceeded four kilometres in length. When he did not obtain immediate success, he switched his attack one night to Oricum and captured that city and its capacious harbour.62 By now news had reached the Roman fleet at Brundisium. Laevinus crossed over, probably with 50 warships and a legio classica (Livy 24, 1i, 3), and captured Oricum. He then passed a relieving force into Apollonia at night unobserved. This force together with the Apolloniates made a sortie the following night, killed almost 3,000 men, captured rather more, and brought the Macedonian siege train inside the walls. The Roman fleet then entered the mouth of the Aous river. Philip burnt his fleet and withdrew his army through Dassaretis into Macedonia. Rome now took Oricum into its zone of dependent states and stationed a fleet there, in order to extend its naval holdings and to patrol both sides of the straits (Livy 24, 40,17 ; Plb. 8, I).63 The first clash between Macedonian and Roman troops had resulted in a severe defeat for Macedon. The failure of his two attempts on Apollonia caused Philip to change his strategy. He set out now to reduce the area of the Roman sector gradually and to open up an entry into the territory of the Ardiaei. This he achieved in the course of 213 and 212. He captured Dimallum (now, if not in the campaign of 214) and probably Gerunium and Orgessus; he consolidated his control of Dassaretis; he brought the Atintani and the Parthini over to his side and so opened up the way into northern Illyris. He was probably operating in Atintanis, when he marched in two days to Lissus and Acrolissus, which were his next objective (Plb. 8, I3-I4).
These two strong places were defended not only by their inhabitants but also by troops from the neighbouring parts of Illyris. Philip captured both places by a brilliant stratagem and by hard fighting, and his success led to the surrender or reduction of all the Illyrians of the neighbourhood (&CravTc-r Tro0S Trrppi). Now or soon afterwards his rule extended over the southern group of Ardiaei around Scodra, the subjects hitherto of Rome's friend Scerdilaidas.64 These successes enabled Philip to isolate Epidamnus and Apollonia and to put economic pressure upon them. It was now possible for him to build and man a fleet of lembi on the Adriatic coast, instead of in the Thermaic Gulf; if the earliest coinages of Lissus and Scodra are correctly dated c. 211, they may have been issued to pay shipwrights and purchase timber for Macedon. With such a fleet he could attempt to capture Apollonia or Epidamnus or both and so eliminate the Roman holding in Illyris. He might also make contact at sea with his Carthaginian allies; for the Carthaginian fleet was operating off Syracuse, and Hannibal held Tarentum and most of the harbours in Magna Graecia. A combined attack by Philip's lembi and the Carthaginian fleet upon the Roman fleet based at Oricum was well within the realm of possibility. The commander of the Roman fleet took the initiative by forming an alliance probably in 21i between Rome and the Aetolian League.
The alliance was directed against Macedon 65 and defined the spheres of looting (these divided at Corcyra, Livy 26, 24, i) ; Rome was to act at sea with not less than 25 quinqueremes, the hope being expressed that Scerdilaidas and his son Pleuratus would join the alliance (and provide their fleet of lembi) ; and the Aetolians were to attack Macedon on land. When Philip heard at Pella of the alliance, he made a sudden attack upon the territory of Oricum and Apollonia, and when the army of Apollonia made a sortie he routed it (Livy 26, 25, I-2).66 It is clear that the Roman fleet and troops were away at the time of Philip's attack, and that Philip's army was not large if
the Apolloniates alone made a sortie. It is possible, then, that Philip made his attack both from the land and from the sea, using his Illyrian lembi based on Lissus at a time when the Roman fleet was away. This suggestion gains some support from the statement in the annalistic tradition that after the pact with Aetolia Philip advanced as far as Corcyra but was frightened away by Laevinus, the commander of the Roman fleet (Zonaras 9, 6).
It also fits into Livy's account of Laevinus after the conclusion of the Aetolian treaty operating against Zacynthos and Acarnania and then returning to Corcyra (Livy 26, 24, 15-i6).67 Immediately after these operations Philip laid waste the' nearest part of Illyricum', translating the Greek word 'XlAupis (' vastatis proximis Illyrici '), a very vague phrase, which may refer to the tribes of the upper Drilon valley, as he then moved into Pelagonia. Thereafter we hear no more of him in Illyris, although Scerdilaidas and Pleuratus threatened trouble in 207 (Plb. 10, 41, 4; Livy 28, 5, 7). For a time the Aetolian League served Rome's interest well. At an unsuccessful peace conference in 208 the League demanded inter alia that Atintanis should be restored to Rome and the Ardiaei to Scerdilaidas and Pleuratus (thus excluding Philip from northern Illyris and its coast). In 206 the League made a separate peace with Philip and naturally made no demands on Rome's behalf as regards Illyris. The peace left Philip free to concentrate on Illyris, where only Epidamnus and Apollonia and farther south Oricum were hostile to him. However, the Romans moved first, probably in spring 205, sending io,ooo infantry, I,ooo cavalry and 35 warships to land at Epidamnus (Livy 29, 12, i). The Myzeqija was then flooded and the rivers were in spate. The first impact of the army was on the Parthini in the Shkumbi valley, through which it had to move in order to lay siege to Dimallum. The first objective of the Romans was to secure the line of communication between Epidamnus and Apollonia, which runs along the inland side of the plain on rising ground overlooked by Dimallum and other places situated on the range of Mt. Shpiragrit. The news reached Philip, who was probably in Macedonia, that the Roman force had come to Dyrrachium (Epidamnus), that the Parthini and other tribes in the vicinity were moved to hope for a revolutionary change in the situation, and that the Romans were besieging Dimallum.68 If Philip took the shortest route from Lower Macedonia, he went via Florina and through the Tsangon pass and descended to Antipatrea (Berat).69 He found that the enemy had withdrawn to Apollonia. He laid waste the territory of Apollonia, and he offered battle; but the Romans and the Apolloniates remained behind the walls. The Roman commander had sent part of his force by sea to Aetolia, asking the League to break its recently sworn treaty of peace with Macedon. The Aetolians refused. The Epirote League then took the initiative and negotiated a peace under which the Parthini, Dimallum, Bargullum and Eugenium were to belong to Rome and Atintanis was to belong to Macedon, if the Senate agreed (as it later did).70
Rome bargained well to get more than she possessed at the time,71 but Philip was able to keep the door open towards northern Illyris through holding Atintanis and he may still have controlled the southern Ardiaei, as no mention was made of any concessions to Pleuratus 72 who was a signatory on the Roman side (Livy 29, I2, I4). Nevertheless Rome retained the essential bases at Epidamnus, Apollonia and Oricum and the ability to defend them at sea and on land. One factor made this settlement more likely to lead to war than the settlement of 228. The buffer zones between the Roman sector and Macedonia had disappeared. For Philip had taken control of northern Illyris and of Dassaretis, and his acquisition of Atintanis made him an immediate neighbour of the Parthini in the upper Shkumbi valley. We may think Philip would have been wiser now to have withdrawn and disengaged. But we know from recent experience that that is difficult to achieve.
Even if Philip had disengaged in 205, it is doubtful if Rome would have played a different tune in 200. When we consider the springs of Roman and Macedonian policy in Illyris, we must remember that the Roman state and the Macedonian state alike were imperialist in the proper sense of the term, that is in desiring power, the power of commanding other states, and were not at any time in their history quietist or pacific states. This imperialistic quality in Rome is obvious from the first act in Illyris. If Rome had desired only to punish Teuta for the kidnapping or killing of Italian merchants or for the killing of her ambassador allegedly or actually at Teuta's command, Rome would have attacked Teuta's kingdom at its centre, that is in the region of Dubrovnik.
In fact, Teuta escaped lightly. What Rome took was not revenge on Teuta but command of a strategic area in Illyris, strategic not only in a military and naval sense but also in an economic sense; and to this command she clung consistently until in 200 her other major commitments were so much reduced that she could exploit her strategic position against Macedonia. If it is claimed that Rome took command of this strategic area in Illyris in order to stop Illyrian piracy, the facts are that neither earlier nor now nor later did the independence of this sector of Illyris prevent or even hinder substantially the practice of piracy by the Illyrians, nor did Rome ever use her bases in her sector of Illyris to try to stop Illyrian piratical expeditions. An accurate consideration of the geographical situation does much to make these conclusions clear and convincing. Between 228 and 205 Rome made enormous calls upon her manpower to meet her very numerous commitments elsewhere, and the fact that she did not exploit her position in Illyris in order to obtain further positions of command until 2II in the case of Greece and until 200 in that of Macedonia is due to the overstraining of her resources in other fields and not to a deep-seated pacificism or indifference. The deliberate choice of Rome in 228 to send envoys not to her new near-neighbours, Macedon and the Epirote League, but to the Aetolian League and the Achaean League, who were enemies of her near- neighbours, showed once and for all that she was concerned not with the establishment of pacific relations but with her future intentions.
Again Rome's choice of an occasion on which to open diplomatic negotiations with Macedon was not calculated to ease relations; for it was a straight demand in 2I7 to surrender to Rome a man who had been for many years an ally of Macedon. And her first positive step in Greece was to incite the Aetolian League to attack Macedon by the treaty of alliance which Rome herself initiated in 211. The Macedonian position was in many respects similar. From the outset Macedon took no steps to cultivate diplomatic relations with Rome. When Demetrius defected from Rome, Macedon accepted him as an ally and remained faithful to him even when Rome had defeated him in 219. In the same way in 2 7 or 216 Rome accepted as an ally Scerdilaidas who had attacked Macedon and been defeated by Macedon, and Rome remained faithful to him. In the early years Macedon was careful not to exacerbate Rome; thus she did not occupy Dassaretis or support Demetrius in his attacks on the Roman sector in Illyris or in his piratical expeditions.
The outbreak of the Second Punic War changed the situation on both sides. Rome was alarmed lest Macedon enter the War, and her demand to Philip to surrender Demetrius in 217 forced the issue for Macedon of compliance or resistance. Macedon chose resistance because she aspired to power not only in Illyris but elsewhere even as Rome did, and Macedon went on to occupy Dassaretis which invited friction with Rome as an immediate neighbour. Rome's support of Scerdilaidas after his attack on Macedon did nothing to lessen the chance of friction. Indeed it was at this point, if not earlier, that Philip of Macedon must have become convinced that Rome's intention was to make war sooner or later on Macedon. The decisive step which involved a serious risk of an armed clash between Rome and Macedon was taken by Philip of Macedon. His attempt to take Apollonia by surprise in 216 before the battle of Cannae and before any understanding with Carthage was only part of a plan to eliminate the Roman sector in Illyris, whether Rome reinforced her dependents there or not. It was chance rather than design which averted a clash on this occasion between Macedonian troops and Roman troops. Again, Macedon took the first step in constructing an alliance directed specifically against Rome in 215 ; the riposte to this was made by Rome in 211 when she allied herself with the Aetolian League against Macedon.
And it was Macedon's attack on Apollonia in 214 which led to the first clash between Macedonian troops and Roman troops; in this instance Macedon was certainly the aggressor. From then on until the peace of Phoenice in 205 hostilities continued if and when their other commitments allowed. There are some who excuse the desire for power, which is the basis of imperialism, on the grounds that it is a form of defence against domination or subjugation by some other power. When they apply this doctrine to Rome, they see in her acts this form of defence magnified to the nth degree. In this context, however, we must excuse not Rome but Macedon on these grounds. For in occupying a sector of Illyris Rome was not averting any threat of domination or destruction by Illyrians or by anyone else in 229; and again in keeping this sector of Illyris in the final settlement she was not averting such a threat by Macedon or anyone else in 205. On the other hand in 228 and in 2I7 Macedon had good reason to suspect that Rome intended to dominate her and in the end destroy her independence.
This basic fear came to the surface in the clause which Philip must have inspired in the Punic-Macedonian pact of 215, under which the first condition of a joint peace with Rome was (Plb. 7, 9, I3) 'that it shall not be possible ever for Rome to begin a war against Macedon'. Moreover Macedon's fear was a realistic one; for Rome's resources far exceeded Macedon's on any estimate,73 and Rome enrolled against Macedon not only in 216 the Ardiaean monarchy, which she had originally set out in 229 to chastise, but also in 211 a strong coalition of Greek states led by the Aetolian League. In the final analysis the incidents which we have been studying in Illyris arose from a Roman settlement in 229 which had little to do with precautions against piracy and everything to do with her own desire for power. If there had been no imperialistic power in that part of the Balkans, Rome would have stood still in Illyris until after 200, not from lack of desire to go further but from the pressure of other commitments. But there was another imperialistic power in the vicinity, aware of Rome's desire for power and alarmed by Rome's technique in diplomacy. Under these conditions a conflict between these two imperialistic states, both strong in the desire for power, was as likely to break out as the conflict between Athens and Sparta had been in 432.
The attendant circumstance of the second Punic War delayed the ultimate conflict but helped to incite actions on both sides which led to the state of war which culminated in 205. Once Carthage was defeated the delay was brief indeed, not because Macedon wanted the ultimate conflict in 200 but because Rome did.