In this as in every other Culture, Spring seeks to express its spirituality in a new ornamentation and, above all, in religious architecture as the sublime form of that ornamentation. But of all this rich form-world the only part that (till recently) has been taken into account has been the Western edge of it, which consequently has been assumed to be the true home and habitat of Magian style-history. In reality, in matters of style as in those of religion, science and social-political life, what we find there is only an irradiation from outside the Eastern border of the Empire.[248] Riegl[249] and Strzygowski[250] have discovered this, but if we are to go further and arrive at a conspectus of the development of Arabian art we have to shed many philological and religious prepossessions. The misfortune is that our art-research, although it no longer recognizes the religious frontiers, nevertheless unconsciously assumes them. For there is in reality no such thing as a Late-Classical nor an Early-Christian nor yet an Islamic art in the sense of an art proper to each of those faiths and evolved by the community of believers as such. On the contrary, the totality of these religions—from Armenia to Southern Arabia and Axum, and from Persia to Byzantium and Alexandria—possess a broad uniformity of artistic expression that overrides the contradictions of detail.[251] _All_ these religions, the Christian, the Jewish, the Persian, the Manichæan, the Syncretic,[252] possessed cult-buildings and (at any rate in their script) an Ornamentation of the first rank; and however different the items of their dogmas, they are all pervaded by an homogeneous religiousness and express it in a homogeneous symbolism of depth-experience. There is something in the basilicas of Christianity, Hellenistic, Hebrew and Baal-cults, and in the Mithræum,[253] the Mazdaist fire-temple and the Mosque, that tells of a like spirituality: it is the _Cavern_-feeling. It becomes therefore the bounden duty of research to seek to establish the hitherto completely neglected architecture of the South-Arabian and Persian temple, the Syrian and the Mesopotamian synagogue, the cult- buildings of Eastern Asia Minor and even Abyssinia;[254] and in respect of Christianity to investigate no longer merely the Pauline West but also the Nestorian East that stretched from the Euphrates to China, where the old records significantly call its buildings “Persian temples.” If in all this building practically nothing has, so far, forced itself specially upon our notice, it is fair to suppose that both the advance of Christianity first and that of Islam later could change the religion of a place of worship without contradicting its plan and style. We know that this is the case with Late Classical temples: but how many of the churches in Armenia may once have been fire-temples?