Since the indecisive battle of Qarqar, the western frontier of the empire had receded as far as the Euphrates, and Shalmaneser had been obliged to forego the collection of the annual Syrian tribute. It would have been an excellent opportunity for the Khati, while they enjoyed this accidental respite, to come to an understanding with Damascus, for the purpose of acting conjointly against a common enemy; but they let the right moment slip, and their isolation made submission inevitable. The effort to subdue them cost Shalmaneser dear, both in time and men; in the spring of each year he appeared at the fords of Tul-barsip and ravaged the environs of Carchemish, then marched upon the Orontes to accomplish the systematic devastation of some fresh district, or to inflict a defeat on such of his adversaries as dared to encounter him in the open field. In 850 B.C. the first blow was struck at the Khati; Agusi* was the next to suffer, and its king, Arame, lost Arnie, his royal city, with some hundred more townships and strongholds.** * Historians have up to the present admitted that this campaign of the year 850 took place in Armenia. The context of the account itself shows us that, in his tenth year, Shalmaneser advanced against the towns of Arame, immediately after having pillaged the country of the Khati, which inclines me to think that these towns were situated in Northern Syria. I have no doubt that the Arame in question is not the Armenian king of that name, but Arame the sovereign of Bit-Agusi, who is named several times in the Annals of Shalmaneser. ** The text of Bull No. 1 adds to the account of the war against Arame, that of a war against the Damascene league, which merely repeats the account of Shalmaneser's eleventh year. It is generally admitted that the war against Arame falls under his tenth year, and the war against Ben-hadad during his eleventh year. The scribes must have had at their disposal two different versions of one document, in which these two wars were described without distinction of year. The compiler of the inscription of the Bulls would have considered them as forming two distinct accounts, which he has placed one after the other.