called Bhago Toro. Below this point the westing movement of centuries has now brought the stream to the extreme edge of the alluvial land, and into contact with the gravel slopes bordering the hill-country. As the gravel tracts project in a bow into the alluvial land of Lower Sindh, the river, unable to erode them, is forced to conform to their contour, and to run in a great curve for nearly 180 miles to Thata. This curve continues through the Delta to the sea, so that from Bhago Toro to the river-mouth the course of the Indus forms an arc of some 260 miles, of which the chord is about 160 miles, and the maximum depth nearly 50 miles. The general result is to give the course of the river in Sindh the form of the letter S. And, as its abandoned channels attest, _such has been the form in which the river has run in past ages_ as it approached the sea. The lower curve of the S had a still bolder sweep _eastward_ when the river ran far east of its present course, unchecked by rock or gravel bed, than it has now, when this part of the course has been shaped by a resistance which the current cannot overcome. This S-shaped course of the river _in all ages_ should be remembered in considering questions of ancient local topography, such, for instance, as that of the site of Patala. It will then be seen to be impossible that the river can have run at the same period in its present course near Haidarâbâd, and, lower down through the Ghâro, or ancient Sindh Sâgara; also that if Patala was at Haidarâbâd, the western river-mouth of Alexander’s time must have lain, not at the western extremity of the sea-face of the Delta, but much to the east of that point. From these remarks (which I have abbreviated from the text), it will be seen that Haidarâbâd can no longer be taken to be the modern representative of Patala. Where then was the point at which, in Alexander’s time, the Indus bifurcated, and Patala was situated? Major-General Haig says that any precise identification of this site is hardly within the limits of possibility; but, for reasons for which his work itself must be consulted, he is of opinion that “the ancient capital of the Delta was most likely not far from a spot 35 miles south-east of Haidarâbâd”—a spot which happens to be 160 miles distant from each extremity of the Delta coast, as supposed to have existed in Alexander’s time. With regard to places which lie farther north than Patala, the views set forth in this volume do not differ from those of Major-General Haig. He is, however, of opinion that the kingdom of Mousikanos was of greater extent than is usually supposed, and must have embraced the district of Bahawulpur, which answers better to the description of that kingdom, as _the most flourishing in all India_, than the country around Alôr.