The eggs from five adult sea-run silver salmon (Oncorhynchus kisutch) were divided into two lots: 8,239 eggs were buried in gravel in a standard hatchery trough and 7,500 placed in a standard hatching basket as a control. Natural conditions were simulated as closely as possible with the gravel eggs. The eggs required 772.3 temperature units (t.u.) to maximum hatch (control), 1,084.3 t.u. to earliest emergence from the gravel, and 1,155.6 t.u. to maximum emergence from the gravel. Initial to final emergence required at least thirty-eight days. Of the eggs buried, 10.2 per cent emerged from the gravel. In the control, 65.9 per cent of the eggs hatched and 48.2 per cent survived to the time that the experimental fish had finished emerging from the gravel. Examination of the gravel and the dead eggs in it at the conclusion of the experiment and observations made during previous experiments support the view that silt carried by unusually severe floods smothered many of the eggs in the gravel. This fact seems to account in large part for the small percentage of salmon emerging from the gravel. Fifty-six days after initial emergence from the gravel, the experimental fish averaged 23.8 fish per ounce (1.19 grams each, live weight) while the control lot averaged 27.6 fish per ounce (1.13 grams each). During these fifty-six days only forty-eight of the experimental fish died, whereas the mortality in the control lot for the same period totaled 905. In the final two weeks, however, the average daily mortality in the control was only one fish.