M. Halévy's theory appeared still less probable when applied to such of the early Sumerian texts as had been recovered at that time by Loftus and Taylor in Southern Babylonia. For these were shown to be short building-inscriptions, votive texts, and foundation-records, and, as they were obviously intended to record and commemorate for future ages the events to which they referred, it was unlikely that they should have been drawn up in a cryptographic style of writing which would have been undecipherable without a key. Yet the fact that very few Sumerian documents of the early period had been found, while the great majority of the texts recovered were known only from tablets of the seventh century B.C., rendered it possible for the upholders of the pan-Semitic theory to make out a case. In fact, it was not until the renewal of excavations in Babylonia that fresh evidence was obtained which put an end to the Sumerian controversy, and settled the problem once for all in accordance with the view of Sir Henry Rawlinson and of the more conservative writers.[2] That Babylonian civilization and culture originated with the Sumerians is no longer in dispute; the point upon which difference of opinion now centres concerns the period at which Sumerians and Semites first came into contact. But before we embark on the discussion of this problem, it will be well to give some account of the physical conditions of the lands which invited the immigration of these early races and formed the theatre of their subsequent history. The lands of Sumer and Akkad were situated in the lower valley of the Euphrates and the Tigris, and corresponded approximately to the country known by classical writers as Babylonia. On the west and south their boundaries are definitely marked by the Arabian desert and the Persian Gulf which, in the earliest period of Sumerian history, extended as far northward as the neighbourhood of the city of Eridu. On the east it is probable that the Tigris originally formed their natural boundary, but this was a direction in which expansion was possible, and their early conflicts with Elam were doubtless provoked by attempts to gain possession of the districts to the east of the river. The frontier in this direction undoubtedly underwent many fluctuations under the rule of the early city-states, but in the later periods, apart from the conquest of Elam, the true area of Sumerian and Semitic authority may be regarded as extending to the lower slopes of the Elamite hills. In the north a political division appears to have corresponded then, as in later times, to the difference in geological structure. A line drawn from a point a little below Samarra on the Tigris before its junction with the Adhem to Hît on the Euphrates marks the division between the slightly elevated and undulating plain and the dead level of the alluvium, and this may be regarded as representing the true boundary of Akkad on the north.