Many of the crevasses are filled with water, but there are no surface streams and no lakes. Melting is rapid during the warm {187} summer days, but the water finds its way down into the glacier and joins the general subglacial drainage. It is evident that the streams beneath the surface must be of large size, as they furnish the only means of escape for the waters flowing beneath the Agassiz, Seward and Marvine glaciers, as well as for the waters formed by the melting of the great Malaspina glacier. The outer borders of the Malaspina glacier are practically stationary, but there are currents in its central part. Like the expanded ends of some of the Alpine glaciers, as the Galiano and Lucia glaciers, for example, this glacier is of the nature of a delta of ice, analogous in many of its features to river deltas. As a stream in meandering over its delta builds up one portion after another, so the currents in an expanded ice-foot may now follow one direction and deposit loads of débris, and then slowly change so as to occupy other positions. This action tends to destroy the individuality of morainal belts and to form general sheets of débris. The presence of such currents as here suggested has not been proved by measurements, but the great swirls in the Malaspina glacier and the tongues of clear ice in the upper portions of the débris fields on the smaller glaciers strongly suggest their existence. The Malaspina glacier is evidently not eroding its bed; any records that it is making must be by deposition. Should the glacier melt away completely, it is evident that a surface formed of glacial débris, and very similar to that now existing in the forested plateau east of Yakutat bay, would be revealed. The former extent of the Malaspina glacier cannot be determined, but it is probable that during its greatest expansion it extended seaward until deep water was reached, and broke off in bergs in the same manner as do the Greenland glaciers at the present day. Soundings in the adjacent waters might possibly determine approximately the former position of the ice-front, and it is possible that submarine moraines might be discovered in this way. The Pimpluna reefs, reported by Russian navigators and indicated on many maps, may possibly be a remnant of the moraine left by the Piedmont glacier from the adjacent coast. The glaciers west of Icy bay were seen from the top of Pinnacle pass cliffs, and are evidently of the same character as the Malaspina glacier and fully as extensive. A study of these {188} Piedmont glaciers will certainly throw much light on the interpretations of the glacial records over northeastern North America. Their value in this connection is enhanced by the fact that they are now retreating and making deposits rather than removing previous geological records. * * * * *