532 APPENDIX {i.e. Andalusia) iu the year 116 '. The mint is not given on the early gold coins, probabl}^ becavise they were usually struck at the Khabf s capital, Damas- cus. "These original dinars (a name formed from the Roman denarius) and dirhema (di-achma) of the Khalif of Damascus formed the model of all Muslim coinages for many centuries ; and their respective weights — 05 and 43 grains — served as the standard of all subsequent issues up to comparatively recent times. The fineness was about "979 gold in the dinars, and "960 to "970 silver in the dirhem. The Mohammadan coinage was generally very pure. ... At first ten dirhems went to the dinftr, but the relation varied from age to age." Thus the dinar of the Omayyad Caliphs, weighing on the average 65'3 grains of almost pure gold, was worth about lis. 6d. In later times there were double dinars, and under the Omayyads there were thirds of a dinar, which weighed less than half a dirhem. As to a coin which Gibbon supposes (p. 5, note 9) to be preserved in the Bod- leian Library, Mr. S. Lane-Poole kindh' informs me that no such coin e.;ists there. "The Wasit coins there preserved were acquired long after Gibbon's time and none has the date 8S a.h. There is a dirhem of that year in the British Museum weighing 44 6 grains. [S. Lane-Poole, Catalogue of Mohammadan Coins in the Bodleian Library, 1888 ; Catalogue of Oriental Coins in the British Museum, vol. i. no. 174 (1875).] " 3. THE THEMES OF THE ROMAN EMPIRE— (P. 62, 66 sqq.) In the tenth century we find the Empire divided into a number of themes, each of which is governed by a strategos. Not only the title of the governor, but the word theme [difxa, a regiment) shows their military origin. These themes existed in the eighth and ninth centuries ; they originated in the seventh. In the latter part of the seventh centur}' we find the empire consisting of a number of large military provinces, not 3'et called themes, but probably kuown as (Tt parriyiai. We have no ofiicial list of them ; but from literarj- notices we can reconstruct an approximate list of the provinces c. 700 a.d. : — ^ 1. The Armeniacs. 6. The Helladics 2. The Anatolics. 7. Italv. 3. The Opsikion. 8. Sicily. 4. The Marines. 9. Africa. 5. Thrace. We have to consider first how this sj-stem originated, and secondly how it de- veloped into the system of themes which we find two centuries later. The identification of the strategoi of the seventh century with the magistri militum of the sixth century gives the clue to the origin of the thematic system. (This was pointed out in Bury's Later Roman Empire, if. 346-8.) The strate- gos of the Armeniacs is the magister militum of Armenia, instituted by Jus- tinian ; the strategos of the Anatolics is the magister militum per Orientem ; the ' ' count " of the Opsikians corresponds to the mag. mil. praesentalis ; ^ the strategos of Thrace is the raag. mil. per Thraciam ; the strategos of the Hella- dics is probably the representative of the mag. mil. per lUyricum. The magistri militum of Africa and Italy remain under the title of exarchs. The maritime provinces arose probably, as M. Diehl attractively suggests, from the province of Caria, Cyprus, Rhodes, the Cyclades and Scj-thia, instituted by Justinian, and placed by him under a quaestor Justinianus. Thus, what happened was this. In the seventh century the old s}-stem of dioceses and provinces was swept away. Its place was taken by the already exist- 1 Diehl, L'origine des Thimes, p. g; Bury, Later Roman Empire, ii. p. 345. 2 Diehl, ib. p. 15. M. Diehl has developed this explanation more fully.
