Mapping of landscape surfaces and their relative ages documents 5 phases in the retreat of the Puget lobe of the Cordilleran ice sheet. The MAIN phase is marked by multiple meltwater outlets through the hills at the southern margin of the ice sheet. With ice retreat, the RUSSELL phase is marked by formation of pro-glacial Lake Russell and drainage via a single outlet at Black Lake, south of Olympia. Further thinning and retreat of the ice sheet opened a lower outlet via the Chimacum valley, south of Port Townsend, initiating the BRETZ phase. Modest further retreat opened Admiralty Inlet and salt water gained access to the pro-glacial basin, beginning the EVERSON phase. Submerged end moraines (Partridge Point to Cattle Point, south end Rosario Strait, mouth of East Sound) demonstrate stepwise retreat of the Everson ice front across the San Juan Islands. I arbitrarily place onset of the SUMAS phase at formation of the oldest subaerial moraine near Ferndale, western Whatcom County. Landforms define multiple successive positions of the ice sheet margin during Russell, Bretz, Everson, and Sumas phases. Easterbrook, Thorson, Dethier, Kovanen, and others have told parts of this story.

Substantially more erosion and deposition are associated with initial Everson shorelines than with younger Everson shorelines, or with the modern shoreline. I infer that during deglaciation the rate of erosion was slowed by revegetation in the wake of the retreating ice sheet. Thus Everson-age deltas are likely not coeval, nor are Bretz or Russell deltas, and earlier estimates of the amount and rate of isostatic rebound may be too low. Early-Everson change in ice flow direction, drainage of Lake Bretz via the Chimacum valley instead of the lower-elevation Leland Creek valley to the west, and rapid isostatic rebound suggest catastrophic collapse of the Juan de Fuca lobe of the ice sheet. The oldest Sumas moraine was built when local relative sea level was at +90 m near Ferndale and relative sea level dropped to its modern level, or lower, by the end of Sumas time. Evidently much glaciomarine (“Everson”) deposition was coeval with Sumas glaciation. Available topographic evidence does not indicate significant readvance of the ice sheet during Sumas time.