In Part I the environment of the coastal dunes of Oregon and Washington is analyzed. Most of the substratum is a narrow foreland or terrace, in part submerged, that borders the mountain front. Temperature is relatively low in summer and rarely reaches the freezing point in winter. Winter precipitation is heavy, and there is almost no snow; there is a pronounced deficiency in precipitation in summer. The summer wind is a very constant afternoon sea breeze from north to northwest; winter winds are variable but include frequent southwest gales. Longshore currents are governed by the seasonal winds, moving southward in summer and northward in winter. The regional vegetation is dense, tall conifer forest. The influence of man has been comparatively slight: in prehistoric times the starting of forest fires; in historic time moderate disturbance due to grazing and recent efforts to control the movement of the sand. Part II deals with forms and processes. The simplest combination of elements comprises sand, wind, and water. Interaction of these produces two patterns. In the transverse-ridge pattern the individual unit is a ridge essentially normal to the summer wind and moving with it; it is asymmetric in profile with gentle windward slope and steep leeward slope (slipface). Origin and maintenance of this profile are explained in accordance with principles developed in the field of aerodynamics, and conclusions arrived at deductively are confirmed by slow-motion photography of smoke streams on the dunes. The oblique-ridge pattern occurs only where there is full exposure to both seasonal winds, plenty of space, and plenty of sand. These conditions are met only in Region III (Coos Bay dune sheet). The units are much more massive than the transverse ridges and are essentially stationary. Their trend lies between the means of summer and winter winds. A theoretical explanation of their origin and maintenance is presented, and the oblique ridges and the “longitudinal dunes” of certain desert regions are compared. Neither transverse nor oblique ridge is found in stabilized condition. The factor vegetation added to the other three—sand, wind, and water—promotes stabilization. On a prograding shore, successive beach ridges are quickly captured and fixed, and retain their initial form indefinitely. On a retrograding shore the processes are exceedingly complex and involve repeated stabilization and rejuvenation. Two cases are presented: flat shore, with and without abundant sand supply; stabilized dune masses undergoing erosion. In the latter case, the commoner, development involves the following phases: trough blowout, merging of troughs, reduction to deflation base, and precipitation ridge, which may in time become completely stabilized while still retaining its characteristic form. In places sheltered from one or the other of the seasonal winds, usually the summer wind, giant parabola dunes, which may likewise become fixed by vegetation, develop. Part III is a description of the dune localities of the Oregon-Washington coast and an account of their history. Forty per cent of this coast bears dunes of greater or lesser magnitude. Thirty dune localities are grouped in four regions. Region I includes four localities north of Tillamook Head, Oregon, making a continuous strip 53 km long, that bears the parallel beach-dune-ridge pattern associated with progradation. The forms in Regions II, III, and IV, heterogeneous and complex, are those characteristic of retrograding shores. The two principal stabilized forms are the precipitation ridge and the parabola dune. In Regions II, III, and IV the existing features came into being mainly during the last grand period of sea-level rise, and the seaward portions of massive parabola complexes have been sliced away in varying degree by the advancing sea. The beach-ridge dunes of Region I were formed during the period of comparatively stable sea level, with a probable small net lowering, which followed the maximum of sea advance and has extended to the present. Progradation here during this period, in contrast with almost none south of Tillamook Head, has been possible because of the ample bed load carried to the coast by the Columbia River. South of Tillamook Head there is, in a number of well-distributed localities, evidence of three episodes of advance. The first is represented by the strip of thoroughly stabilized dunes that nearly everywhere forms the inner marginal part of the dune complexes. This episode reached its culmination before the sea had attained its maximum of advance—attested by the slicing away of portions of completely stabilized masses. The second advance for the most part fell, short of the first, though in a few places it overpassed the limits of the latter; present condition ranges from complete stabilization to vigorous activity. The third episode is represented by active dunes with open access to the shore. In certain localities there are only an inner strip of stabilized dunes and an outer zone of active dunes. It is assumed that in these the visible effects of the second and third episodes have merged. An earlier cycle of dune development, similar to the modern one in character and extent, is proved by eolian sediments containing altered podzolic soils, which make a minor part of the mantle of unconsolidated materials that lies upon the rock platform of the 30-m terrace. The dunes of which these masses are remnants were formed during the next-to-the-last grand period of submergence. Development of the dunes of Regions II, III, and IV, associated with the grand period of submergence, is assigned to the period of deglaciation that followed the Wisconsin maximum, and mainly to the period of rapid deglaciation that began after the Valders-Mankato advance. The beach-ridge dunes of Region I, on the other hand, have developed in their entirety in the time since sea-level rise was succeeded by stability. By analogy,...