From about the 7 t h /6 t h centuries BC ancient Chorasmia was located south of the Aral Sea, in the delta of the Classical Oxus River (mod. Amu-dar'ya). To the north lay the Inner Asian steppe (now Kazakhstan), to the west the cliffs of the inhospitable Ustiurt Plateau (further west, the Caspian Sea), to the east the delta of the Classical Jaxartes (mod. S'ir-dar'ya), and to the south two deserts, the Kara-kum and Kz'il-kum which separated Chorasmia from Margiana and Sogdiana. Its geographical isolation form the civilsed ancient Indo-Iranian world resulted in virtually independent cultural development for much of its early history and, later on, after the devastation caused by the Mongols and particularly Timur, remarkable preservation of pre-Islamic monuments the like of which cannot be found anywhere else in Central Asia. Long before archaeological explorations began, Chorasmia was known from Persian and Greek texts as a province (satrapy) of the Achaemenid Persian empire; it also stands as the possible area of the Aryan Expanse of the Avesta, as the best land created by Ahura Mazda and therefore of signal importance regarding the early stages of the Zoroastrian faith. By the time of Alexander the Great Chorasmia was independent and had a king. This is the last textual mention of Chorasmia until the early medieval period, although it may have had relations with the Kushan empire at least from the 2 n d century AD onward. Exploration began in the I930s under the leadership of S. P. Tolstov who founded the Chorasmian Archaeological Expedition whose work continued up to the collapse of the Soviet Union in I991. Since I995 the University of Sydney Central Asian Programme (USCAP) and the Karakalpak Branch of the Uzbek Academy of Sciences have conducted excavations in eastern Chorasmia, at two key sites: Kazakl'i-yatkan (Akcha-khani-kalesi), a heavily fortified site (Rus. gorodishche) which may have been one of the sacral centres of ancient Chorasmia; and at Tash-kirman-tepe, one of the earliest, undisputed (Zoroastrian) fire temples yet discovered. Both sites may date back to the early 4 t h century BC and some standing remains may be even more ancient. Both sites were abandoned during the Kushan period after about the 2 n d century AD.

The Gathas of Zoroaster does not refer to any iconography or temples. Indeed, a doctrinal aversion to such rituals can be detected in his teachings. His seven abstract and amorphic Amesha Spentas, for example, are probably the first purely conceptual deities with a built-in deterrent to artistic representation. His prescribed places of daily rituals, moreover, are mountaintops and riverbanks, and not temples. Such devotional purity and simplicity is also confirmed by the lack of hardly any archaeological evidence to the contrary in the prophet's homeland in Eastern Iran. This is also the case in the pre-imperial period in Western Iran, where the centre of religion moved early in the first millennium BC. Assyrian annals and many classical writers also confirm that Medes and early Achaemenians did not have any cult statues and temples. The latter, however, soon after forming the first world empire, developed the many aspects of by now, familiar Zoroastrian iconography. This was inspired, no doubt, by the artistic repertoire of their western subjects, such as the Babylonians, Elamites, Greeks and Egyptians. Nor could the Achaemenians refrain, towards the end of their rule, from building temples, both to house their own sacred fire, and the alien-inspired divine images. These two distinct temple cults, established around 400 BC, appear to have spread throughout the Achaemenian, and later, the Parthian Empires. Their religious significance was not challenged until the iconoclastic Sasanians (224-637 AD) removed the divine images from their temples. However, the fire-temples and religious iconography have remained permanent features of the Zoroastrian devotional life since the Achaemenian times.

This paper examines translation patterns for Biblical Hebrew nominal and verbal clauses in the Arabic translation of Saadya Gaon. The examination is based on three major versions (Derenbourg edition, Hasid edition, and Ms. St. Petersburg), but considers other versions as well (Ms. Add. 1008 held in Cambridge, Constantinople, London and Paris Polyglots, and Geniza fragments). Although the data reveals differences among the various versions, it is possible to trace consistent syntactical patterns of nominal and verbal clauses typical of Saadya's translation. These patterns shed light on the syntactical characteristics of Saadya's Arabic, and they reflect well the original Biblical Hebrew word order.

L'auteur s'interroge dans le present article sur les raisons qui pousserent l'empereur Julien a reconstruire le temple juif de Jerusalem en 362 ou 363 de notre ere. Les autres questions posees auxquelles l'auteur tente d'apporter une reponse sont les suivantes : la reconstruction du temple fut-elle effectivement due a Julien ? Si la reconstruction fut debutee, quand fut-elle arretee ? Pourquoi les sources juives contemporaines ne mentionnent-elles pas cette reconstruction ?