None of the tributaries to the Washita, above the hot springs, have hitherto been explored. The Little Missouri and the Fourche-au-Cadeau, enter it in succession from the west, in the course of a considerable bend which it makes to the south, after receiving the waters of Hot Spring creek.[44] These two streams run mostly in a mountainous country, though some fertile lands and some settlements occur on each. On the Little Missouri, Hunter and Dunbar found the maclura, a tree confined to fertile soils. The first considerable stream entering the Washita from the north is the Saline, rising in three principal branches, twenty or thirty miles north-west of the hot springs. The road from Derdonai to the springs crosses these streams near their sources, in an extremely rugged and mountainous region. The Saline, like the Washita itself in this part, and the other tributaries already enumerated, is liable to great and sudden floods, and also to great depression in seasons of drought. Originating in a mountainous tract, and in the continuation of the range so profusely supplied with springs in the country about the sources of White river, we might expect that the Washita would be fed by numerous and unfailing fountains. It appears, however, to derive the greater part of its supplies from the water of rains, and consequently to rise and fall according to the time of year and the state of the weather. Where Major Long crossed the Washita, on the 31st December 1817, six miles south-west of the hot springs, the river was one hundred and fifty yards wide, about four feet deep, and extremely rapid. In the latter part of October 1820, at the time of our journey, the Washita at Keisler's settlement, about fifteen miles below the springs, was something {157} less than one hundred yards in width, flowing in a deep and unequal channel over a bed of clay-slate. The water is here ten or fifteen feet deep in many places, and the currents scarce perceptible; as we looked down upon the river from the elevated banks it appeared like a quiet lake, and the unusual blackness of the waters suggested the idea of its great depth. Little groups of naked rocky islands were disclosed here and there in different parts of the channel. On examination we found the apparent dark colour of the water to depend upon the complexion of the rocks which form the bottom and sides of the bed, they being principally of a dark-coloured argillite; and not only these, but the small fragments of quartz and other whitish stones, had acquired, from lying in the water, a peculiar tinge of dark brown. We expected to find an incrustation covering the surfaces of these stones, but upon examination the colouring matter seemed inseparably blended with the rock itself. The water, seen by transmitted light, was entirely transparent, and had no perceptible saltness.