This expedition was for Tiglath-pileser merely an interlude between two more serious campaigns; and with the beginning of his fifth year he reappeared in the provinces of the Upper Euphrates to complete his conquest of them. He began by attacking and devastating Musri, which lay close to the territory of Milid. While thus occupied he was harassed by bands of Kumani; he turned upon them, overcame them, and imprisoned the remainder of them in the fortress of Arini, at the foot of Mount Aisa, where he forced them to kiss his feet. His victory over them, however, did not disconcert their neighbours. The bulk of the Kumani, whose troops had scarcely suffered in the engagement, fortified themselves on Mount Tala, to the number of twenty thousand; the king carried the heights by assault, and hotly pursued the fugitives as far as the range of Kharusa before Musri, where the fortress of Khunusa afforded them a retreat behind its triple walls of brick. The king, nothing daunted, broke his way through them one after another, demolished the ramparts, razed the houses, and strewed the ruins with salt; he then constructed a chapel of brick as a sort of trophy, and dedicated within it what was known as a copper thunderbolt, being an image of the missile which Eammân, the god of thunder, brandished in the face of his enemies. An inscription engraved on the object recorded the destruction of Khunusa, and threatened with every divine malediction the individual, whether an Assyrian or a stranger, who should dare to rebuild the city. This victory terrified the Kumani, and their capital, Kibshuna, opened its gates to the royal troops at the first summons. Tiglath-pileser completely destroyed the town, but granted the inhabitants their lives on condition of their paying tribute; he chose from among them, however, three hundred families who had shown him the most inveterate hostility, and sent them as exiles into Assyria.* * The country of the Kumani or Kammanu is really the district of Comana in Cataonia, and not the Comana Pontica or the Khammanene on the banks of the Halys. Delattre thinks that Tiglath-pileser penetrated into this region by the Jihun, and consequently seeks to identify the names of towns and mountains, e.g. Mount Ilamuni with Jaur-dagh, the Kharusa with Shorsh-dagh, and the Tala with the Kermes-dagh; but it is difficult to believe that, if the king took this route, he would not mention the town of Marqasi-Marash, which lay at the very foot of the Jaur-dagh, and would have stopped his passage. It is more probable that the Assyrians, starting from Melitene, which they had just subdued, would have followed the route which skirts the northern slope of the Taurus by Albistan; the scene of the conflict in this case would probably have been the mountainous district of Zeitûn.