URING the month of October, 1938, the writer observed a large number of chinook salmon (Oncorhynchus tsckawytscha) spawning in the main Columbia River below Kettle Falls, Washington. The completion of Grand Coulee Dam has flooded this area so that the phenomenon will not be seen again. Since the observations are somewhat unusual when compared with the spawning habits of this species as recorded elsewhere, the following record is made of them. Kettle Falls were located in the Columbia River in Washington about 35 miles south of the Canadian border. Although the Falls were normally nearly flooded out during the period of the spring run-off, when the amount of water carried by the river increases tremendously, in the summer and autumn a drop in the neighborhood of 15 feet in height was presented. While the crest was broken by numerous shoots and shallow spots, the Falls did constitute something of a natural barrier to the upstream migration of salmon, and since long before the time of recorded history, have been the fishing grounds of that tribe now known as the Colville Indians. Immediately below the Falls the river, until the area was flooded, was compressed between high rock walls and had formed a long, deep hole. Below this hole the river widened and shallowed, until in periods of low water it was little more than 15 feet deep. In this area, which extends for several miles downstream, the bottom is entirely covered with gravel. One of the Indian fishermen, Pete Lemery, was employed by the Washington Department of Fisheries to keep count of the number of fish caught by the Indians and make general observations. Early in October he reported fish spawning below the Falls and the writer was sent to confirm his report. On October 19 a survey was made of the area. In the immediate vicinity five separate spawning areas were discovered. About a mile downstream from the Falls a V-shaped gravel island, with the apex pointing downstream, appears at low water. In the deep water adjacent to each arm of the V there were salmon spawning. Because of the lack of accessibility, no estimation could be made of the number of salmon on either of these spawning grounds. Mr. Lemery reported later, however, that he counted 22 redds in this stretch, although he was not able to observe the entire area. Below this on the east side of the river two other large spawning areas were found. The upper one of these was north of the town of Kettle Falls near the old highway ferry landing. The lower one was separated from the upper by about one-fourth of a mile and extended down to the site of the former private ferry landing a little south of the town. Both of these grounds were approximately a half mile long and extended from the bank of the river