UP THE PARANA RIVER TO ROZARIO. 227
houses being concealed behind the water-slope. After this point comes S. Pedro^ where the Parana de las Palmas anastomoses with the Parana Guazii.
The Yi will run up this " Guazu/^ as it is familiarly called. I will first attempt to explain something of the delta formation. The general opinion of the older travellers makes this ringe of the Pampas the easternmost limit of a southern Gulf of Mexico. The limits of this great estuary, a rough quadrilateral, would be Cape S. Antonio to the south-east, Patagonia to the south (limit unknown), westward the line of the Andes, and northwards the Chiquitos country, and the water - sheds which divide the basins of the Amazons and the Plate. Thence the outline would pass eastward of the Xarayes swamps and follow the great spinal cordillera of Paraguay. South of Villa Rica it would trend eastward, embracing the valley of the Parana proper as far as the Salto de la Guayra, and to the south south-east the valley of the Uruguay would complete the circuit. Thus the length would be 1920 geographical miles (betweeu south latitudes 17° and 49°), and the breadth 600 miles (from 58° to 68° longitude) west of Paris. The total area is 1,152,000 square miles — nearly half of South America. This vast estuary is supposed to have been an inland sea with rocky islands, such as the Sierras of Cordoba and S. Luis, gradually warped up by the washings of the Andes and the other highlands, while the ground grew under the influence of secular elevation and deposition. But M. A. Bravard {" Geo- logic des Pampas,^^ a work unhappily incomplete) explains the so-called Pampasian alluvium by atmospheric and terrestrial causes. Secular upheaval produced a shallower sea — upon which sand dunes formed a floor, and subsequently the dust and volcanic ashes were transported by the Pampero builder from the Andes and the arid regions to the west,
15—2
