Page:Further India; (IA furtherindia00clif).pdf/430

performed by M. Pavie's mission. Similarly the Song Ka, still farther to the south, has been traced to its rise in the mountains which divide Tongking from the valley of the Mekong.

We come now to the Mekong itself, the immense river with the exploration of which we have been so much engaged in the pages of this work. Garnier, it will be remembered, arrived unwillingly at the conclusion that the Mekong was impracticable for navigation by steam- launches above the Khong rapids, but since his day the construction of shallow-draught river-craft has under- gone an immense development, such as he may well be excused for having failed to foresee. The most formid- able obstacle in the Sombor flight of rapids was the fall known as Preatapang, which Garnier himself had made two several attempts to examine, and had pronounced impossible for steam-launches. In 1883, however, Cap- tain Réveillère succeeded after much difficulty in forcing a launch up the flight, and in the course of examinations made during 1891 and 1892, Lieutenant Robaglia discov- ered a channel some six metres in width which is practic- able for steam-launches at all seasons of the year. He further discovered that the island of Khon is in fact a cluster of small islands, and in one of the channels divid- ing these it has been found possible to dig a canal, pro- tected by locks, which gives easy access to the river above the falls. Steam communication between Saigon and the reaches immediately above Khon is thus at last assured. In 1893 an expedition under the command of Lieutenant Simon and Ensign Le Vay was sent with three steamers, the Ham Luong, the Massie and the La Grandière, to