and river by elaborate systeins of fortification and entreuchment.' Their kingdom, almost girt about by an artificial circumvallation, might be compared to an entrenched camp, and the stages in its territorial expansion are marked by successive ramparts. Beyond the Danube, il ditch and carthen wall connected the Pruth with the Dniester in northern Bessarabii, ind i similar fence protected the angle between the months of the Sereth, the Danube, and the Pruth." The early settlement of Isperikli at Little Preslav, near the mouth of the Danube, was fortitied by a

rampart across the Dobrudzha," following the line of older Roman walls of earth and stone, but turned to confront a foc avancing from the south, while the Roman defences had been designed against barbarians descending froin the north. When the royal residence was moved to Pliskit, at line of fortifications was constructed along the heights of Haemus; and a trench and rampart from the mountains to the Danube marked the Western froutier. When their successes at the expense of the Empire enabled the conquerors to bestride the mountains, a new fence, traversing Thrace, marked the third position in their southwarel advance.' The westward expansion is similarly separated by two more entrenchments connecting the Haemus with the Danube, while the right bank of that river was defended by il series of fortresses and entrenchuments from Little Preslav to the neighbourhood of Nicopolis.

The main road from Constantinople to the capital of the Bulgarian kings (Tosse the frontier, cast of the Tundzha, near the conspicuous heights of Meleona,' which, still covered with

The following brief description is in Southern Bessarabia between the based on Shkorpil's, in bobni, c: XX. Pruth and Luke huduk ; ib. 524. 503 . ; C also l'rilo:k. ii. 566-509. See Schuchhardt, Arch..op. Mitthei. Masudi describes the "clominioll lungen, ix. 216 817. (1885). of the Bulgarians its surrounded by : Schuchlardt, ib. 87 849. ; Tocilesco, a thorn fence, with openings like Fouilles et recherches archéologiques en wooden windows, and resembling it Roumanie, 1900 (Bucharest). wall and canal (llarkavi, Shuzuniin, • See below, p. 301. 126). Uspenski (atbulin, 10) tikes 6 bobu, 501-505, 514, the heights of “ dominion' to mean the royal aula, Bakuzhik.. , Slikorpil remarks that and relates the description to Alubi. they “could

ist natural This is it strained interpretation ; but boundary, before the construction of possibly Masui's source mentioned The Erkisiin." It is certain that by both the circumvallation of the king. the middelle of the eighth century itt dom and the fortifications of Pliskio, Tintest the Bulgarian frontier huu it!12 Masri confused chem.

the remains of Bulgarian fortifications, marked an important station on the frontier, since they commanded the road. Το the north-west of Meleona, the Bulgarians held Diampolis, which preserves its old name as Jambol, situated on the Tundzlia. The direct road to Pliska did not go by Diampolis, but ran northward in a direct course to the fortress of Marcellue, which is the modern Karnobudl.' This stronghold possessed a high strategic importance in the early period of Bulgarian history, guarding the southern end of the pass of Veregava,' which led to the gates of the Bulgarian king. Not far to the west of Veregava is the prisis of Verbits, through which the rouil lay from Pliska to Diampolis. The whole route from Marcellae to Pliska was flanked by a succession of fortresses of carth and stone.

1

$ 2. Krum and Nicephorus I. In the wars during the reign of Irene and Constantine VI., the Bulgarians had the upper hand; king Kardam repeatedly routed Roman armies, and in the end the Erpress submitted to the humiliation of paying an annual tribute to the lord of Pliskal. A period of peace ensued, lasting for about ten year's (A.1). 797-807). We may surmise that the 497, who relates that krumi sought to kormisos, Jireček in the ninth century riew with Michael I. (see below) (en sboba, 508). Seo below }, 301. the trouty concluded in the reign atbobu, bol, cp. 562. Jirecchi (arch.. of Theolosius of Adramyttion and op. Wilth. X. 158) wished to place the patriarchate of Germanis with Narcellino at kainbash. His identifica Korrisos, “tlocus ruler of Bulgaria." tion is based on Ama ('omnienia, i. 241 There is an arror here, ils Tervol was and ii. 71 (ed. Reillir'sclarid), and the Bulg:ırian king in the reign of he princes Lordeas it harnoband. But Theodosius III., in Constantine V. Shkorpil finds Landens at the grass of was Emplor in the reign of Korimisos Marsh (505). Both place Goloe also (7-43-760). If we accept Theodosius, mentioned by Auna) near Kirtakli. the treaty was in A.1. 716 : if we hattakli, kaialasli, tid the Marash accepit kormisos, it was a generation detile lie in this order on the soutlı. later. ily view is that the treaty on ward road from the Verbits poiss to which hrum based his negotiations Jamibol. was between Kommisos iud Constantine % The identification of the Kleooipa V., but that in the text of that treaty Bepegaßww with the Rishi Pass is lin. ali okler treaty between Theolosius questionably vight. C. albobu, 564; and Torvel was referred to, The Jireček, locrosstrasse, 149-150. Jirecek decision of this question does not, of also identifies Voregava with the aidat course, cloride the lite of the Erkosiia, σιδηραι or Σιδη of Grech historians, ils Melona (τους όρους από Μηλεώνων but Slikorpil (dbubu, 505) takes Soompa tñs Opçians, ib.) may have been the

to be the Verbits prase

1 am inclinice? howary many years before its con. to agree with Jireček.

The two st'ulation. Zlatarski dites it in the neighbouring passes together rvign of Tervol, Slikorpil in that of known as the Gyrlurski l'ass (ib, 518).

l'e

attention of the Bulgarian king was at this time preoccupied by the political sitütion which had arisen in the regions adjacent to the Middle Danube by the advance of the Frank power and the overthrow of the Avars. On the other hand, Nicephorus who, soon after his accession, was embroiled in war with the Saracens, may have taken some pains to avoid hostilities on his northern frontier. It is at all events significant that he did not become involved in war with Bulgaria until the tide of the eastern war had abated. We do not know what cause of provocation was given, but so far ils our record goes, it was the Roman Emperor who began hostilities Kardam had in the meantime been succeeded by Krum,' a strong, crafty, ilud ambitious barbarian, whose short reign is memorable in the annals of his country.

It was in A.D. 807 that Nicephorus set forth at the head of all irmy to invade Bulgaria." But when he reached Hadrianople it mutiny broke out, and he was compelled to abandon his expedition. The next hostile movement of which we hear--we cannot say which occurred—was the appearance of a Bulgarian army in Macedonia, in the regions of the Strymon, towards the close of the following year." regiments of the garrison of the province, with the strategos himself and the oflicers, were cut to pieces, and the treasury of the khan was enriched by the capture of 1100 lbs. of gold (£17,520) which had been destined to pay the soldiers. It would seem that the Romans had not expected an attack so

3

Many late in the year; but the presence of a considerable force in the Strymon regions points to the fact that the Bulgarians had already betrayed their desigus against Macedonia. In the ensuing spring (809) Krum followed up his success on the Strymon by an attack on the town of Sardicu, whicli seems at this time to have been the most northerly outpost of the Empire towards the Danube. He captured it not by violence, but by wily words, and put to death a garrison of six thousand soldiers and (it is said) the population of the place. It does not appear that he had conceived the idea of annexing the plain of Sardica to his realm. He dismantled the fortifications and perhaps burned the town, which was one day to be the capital of the Bulgarian name. When the tidings of the calamity irrived, Niccphorus loft Constantinople in haste on the Tuesday before Easter (April 3). Although the monk, who has related these events, Silys nothing of his route, we can have no doubt that he marched straight to the mountains by Meleona and Marcellae, and descended on Pliska from the Veregava l'ass. For he dispatched to the city an Imperial letter in which he mentioned that he spent Easter day in the palace of the Bulgarian king. The plunder of Pliska was a reprisal for the sack of Sardica, to which Nicephorus then proceeiled for the purpose of rebuilding it. We are not told what load he took, but he ivoided meeting the victorious army of the enemy. It is said that some oflicers who had estaju el the massacre asked Nicephorus in vain for a promise that he Would not punish them, and were forced to desert to the Bulgarians.

The Emperor desired to rebuild Sardica ils speedily and is cheaply ils possible, and, fearing that the soldiers would be unwilling to submit to a labour which they might sily was not it soldier's business, he prompted the generals and officers to induce the soldiers to address a spontaneous request to the Emperor that the city might be rebuilt. But the men saw through this stratagem, and were filled with indignation. They tore down the tents of their superiors, and, standing in front of the Emperor's pavilion, cried that they would endure

6

his rapacity no more. It was the hour of noon and Nicephorus Wils dining. Ho directed two patricians to attempt to tranquillise the army; the noise abated; the soldiers formed a company on a hillock hard by, “and, forgetting the matter in hand, kept cryilig, 'Lord, have mercy !" This unorganized mutiny Wils soon quelled by Imperial promises,' and the. oflicers were all on the Emperor's side. Punishment, however, was afterwards intlicted on the ringlenders.

Nicephorus viewed with illxiety the western provinces of his Empire in Macedonia and Thessaly. The Slavs, on whose tidelity no reliance could be placed, were predlominant there, and it was the aim of the Bulgarians to bring the Macedonian Slavs under their dominion. To meet the langers in this quarter the Emperor determined co translate a large number of his subjects from other parts of the Empire and establisha them is Roman colonists in what was virtually a Slavonic land. They could keep the Slavs in check and help in l'epulsing Bulgarian nggression. The transmigration began . in September 809 and continued until Easter 810. It seems to have been an impopular measure.

Men did not like to leave the homes to which they were attached, to sell their property, and say farewell to the tombs of their fathers, The poor cling far inore to places than the rich and elucated, and it was to the poor agriculturists that this measure exclusively applied. Some, we are told, were driven to desperation and committed suicide rather than go into a strange and distant land; and their richer brethren sympathized with them; in fact, the act was described as nothing short of "il captivity.” But though it may have been haril on individuals, it was it measure of sound policy; and those who on other grounds were ill-disposed to the government exaggerated the odium which it aroused. Nicephorus, who, as we are told, porideil himself greatly on this act, seems to have realised the danger that the Slavonic settlements in Macedonia and Greece might eventually be gathered into a Bulgarian empire; and these new colonies were designed to obviate such a possibility. i On the next day Nicephorus made

"mos!"

were punishoil by & speech full of asseverations of his stripes, bunishment, or compulsory good will to the soldiers and their tonsure, and the rest children. He then returned to Cple., Chrysopolis (486). leaving Theodosins Salibaras

Theoph. 496. discover the ringleaders. Theophanes

says

were sent to

to