APPEXDIX 533 ing tlivision of the Empire into military provinces — the 8])heres of the magistri militum ; and a new Greek nomenclature was introduced. The cause of the change was the extreme peril of the Empire from the Saracens. The needs of defence suggested a military organization ; when the frontier was reduced and every province was exposed to the attacks of the enemy, there was a natural tendency to unite civil and military j>ower. In the west, the exarch of Africa and the exarch of Italy are the magistri militiun who have got into their hands the power of the Praetorian prefects of Africa and Italy respectively ; and in the same way in the east, the strategoi of Thrace, the Anatolics. the Armeniacs and the Opsikians, have each a parcel of the prerogatives of the Praetorian Prefect of the East. During the eighth and ninth centtiries the provinces came to be generally called themes, and the list was modified in several ways. (1) It was reduced by losses of territory ; thus Africa was lost. (2) Some of the large provinces were broken up into a number of smaller. (3) Some small frontier districts, which were called clisv.rarchies (Keiaoipa. a mountain jiass), and had been dependent on one of the large districts, were raised to the dignity of independent themes. Thus the Bucellarian theme was formed in the north of Asia Minor between the Opsikian and the Armeniac themes. Then Paphlagonia was cut out as a separate province. The Thracesian theme was cut off the Anatolic. The Marine theme ultimately became three : the Cibyrrhaeot.^ the theme of Samos, and the Aegean Sea. The Helladic province was divided into three (at least) : Hellas, Nicopolis, and the Peloponnesus. The Daln^atian towns were constituted into a separate district ; a separate theme seems to have been formed out of Calabria and the Ionian islands ; but these islands were subsequently detached and constituted as the theme of Cephallenia. In the east of Asia Minor : Colonea, Lycandos, Sebastea, &c. The Armeniac and Anatolic provinces were abridged by the creation of the themes of Charsianon and Cappadocia. We can trace in the chronicles some changes of this kind which were carried out between the seventh and the tenth centuries. But it is not till the beginning of the tenth centiiry that we get any ofiicial list to give us a general view of the divisions of the Empire. The treatise on the themes by the Emperor Constantine (see above, p. 66 sqq.), composed about a.d. 934, is generally taken as the basis of investigation, and, when historians feel themselves called upon to give a list of the Byzantine themes, they always quote his. In my opinion this is a mistake. "We possess better lists than Constantines, of a somewhat earlier date. Emperor though Constantine was, his list is not official ; it is a concoction, in which actual facts are blended with unmethodical antiquarian research. His treatise is valuable indeed ; but it should be criticised in the light of the official lists which we possess. (1) The earliest list is one included in the Cletorologion of Philotheus (see above, p. 517) : Const. Porph. De Cer. Bk. ii. c. .52. p. 71.3-14 and 727-8. The strategoi of the themes are enumerated with other officials in their order of pre- cedence. The list used bj- Philotheus must date from the first years of the tenth century ; it does not mention the themes of Langobardia and Sebastea, which existed before the death of Leo VI., but Cephallenia. which he created, appears in the enumeration.* (2) The second list is a table of the salaries of the governors of themes and clisurae, in the reign of Leo VI., and is included in c. 50 of the Second Book of the De Cerimoniis. But its editor lived in the reign of Romanus I. For he speaks of the governors of Sebastea, Lycandos, Seleucia, Leontocomis, as having been at that time, that is in Leo's reign, clisurarchs (is ^v rrfre KKeicrovpdpxvs)- In other words, a list was used in which these four districts appeared as clisurarchies. Subsequently they were made themes (strategiai) and the editor brought them 'The Cibyrrhaeot Theme was not promoted to thematic dignity till the latter part of the eighth century. This is proved by the seal of " Theophilus. Imperial spathar and turmarch of the Cibyrrhaeots," see Schlumberger, Sigillographie byzantine, p. 261. •» Rambaud, L'empire grec, p. 176.
