nath if his expedition proved to be a success. Both of them apparently knew it and realized what their situation would be if he were able to conquer the two states, and they were prepared for the attack. Sampsiceramus was the first prospective victim. He sustained the attack and defeated one part of the Persian army. Odenath naturally did not wait for the retiring Persians to attack him. He took the offensive and attacked them, casting in his lot with the Romans, from whom he may have expected as reward for his deed some concessions such as he had not received from Shapuhr. And in fact he was not disappointed. We hear that in A.D. 258 he was in possession of the title vir consularis, which he may have received much earlier.73 It may have been under the pressure of Odenath that the Persians evacuated Dura. This made it possible for the Roman garrison to return to that city.74

Have these hypothetical but probable events left any trace in Dura? I may tentatively mention two paintings, or better drawings, in Dura which may be connected with the temporary occupation of Dura by the Persians and a probable visit to Dura of Odenath. 74a In a private house of Dura was found an unfinished colored drawing occupying almost the whole of the surface of one of the walls of the diwan of the house (fig. i). It shows in the conventional way of the early Sassanian artists a scene of battle. In this scene appear a Sassanian king and his enemy, apparently the leader of the Romans on the one hand (large-scale figures) and members of the Sassanian royal family conquering individual Roman horsemen (smaller figures) on the other, all this in the presence of the gods shown seated on a couch. Dr. A. Little and I have discussed this curious composition several times. I cannot deal with it again at any length in this paper. Suffice it to say that the drawing is certainly the work not of a Greek but of an Iranian "artist," and that it is not an imaginary picture but represents a real historical battle in a conventional way. It is impossible to ascribe the drawing to a time earlier than the third

unit, cohors, ala, or numerus. See H. Seyrig, Syria, XIV, 1933, pp. 152 ff., and his article on the inscriptions of the agora quoted in note 71 (inscriptions nos. 8 and 9).
73. IGR, III, 1031 ; J. Cantineau, Inventaire des Inscriptions de Palmyre, I, no. 17.
74. I realize very well that my reconstruction of the course of events connected with Odenath is very hypothetical. The fragment of Petrus Patricius is not dated and is generally assigned to the time preceding the second invasion of Syria by Shapuhr. However if we assign the episode of the negotiations between Shapuhr and Odenath to A.D. 260 we shall be at a loss to understand the policy of Shapuhr before his first invasion of Syria. In this expedition the neutrality of Odenath was perhaps more important to him than in A.D. 260, since Shapuhr in the first expedition invaded Syria from the Middle Euphrates, while in A.D. 260 he carried out his invasion taking the Mesopotamian road. Why should he not try, before starting his offensive, to win the support of Odenath? He knew that Odenath was ready and willing

to negotiate. If he rejected Odenath's offer it was probably because Odenath offered too little and asked for too much. If negotiations between them came to nought in A.D. 253 and led to an attack by Odenath on the rear of one of Shapuhr's armies, we may understand better the events of A.D. 260. In A.D. 260 Odenath had committed himself too deeply to renew his negotiations with Shapuhr. He merely waited the turn which the events should take. In the light of his experience of A.D. 253 he knew well that, even if the Romans were defeated, he had a fair chance to profit from the victory of Shapuhr by attacking his disorganized rear a second time. Very poorly attested and rather problematic is also the first attack of Odenath on the rear of Shapuhr. No one except Domninus-Malalas (p. 295) mentions it. And it must be granted that Domninus' report is so overgrown with legendary details that even the kernel of it may be regarded as pure legend.
74a. On the evidence yielded by coin hoards found at Dura see A. R. Bellinger, below, pp. 64-65.


Created by the Digital Documentation Center at AUB in collaboration with Al Mashriq of Høgskolen i Østfold, Norway.

990201 MB - Email: hseeden@aub.edu.lb