Glacial, but Gagel describes it as Second Glacial, and does not recognize any First Glacial in Germany. This, in Leverett's opinion, is an error. The matter of the extent of the Fourth Glacial drift in Germany is not fully settled, but there is an unanimous agreement that it extends much farther than the Baltic moraine, which J. Geikie considered as its limit. [Soergel (1919.1, figure 12, page 100) presents a map (figure 14) show

FIGURE 14.-Retreat Phases of the Fourth Glaciation in North Central Germany

After Soergel.

ing the stages which have been determined for the retreat of the ice of the Fourth Glaciation in north central Germany.] Leverett regards as very instructive the map in Werth's book, "Das Eiszeitalter," published in 1909, which sets forth four of the chief morainic systems of the Fourth Glaciation in Europe.

(9) Leverett considers that there is distinct advantage in the simple fourfold classification of the Ice Age, such as is presented in the Osborn

Reeds diagram (figure 13), but unfortunately there is in America the difficult "Iowan" question to complicate correlation; until this is settled, we do not know whether there were four or five glacial stages in America. Our last, or Wisconsin, stage is so complex that it is not an easy matter to correlate any part of it with the corresponding part of the Würm stage of Europe or with the last glaciation of the north of Europe. Leverett is unable to come to a definite opinion as to whether the Third Glaciation of Europe is the correlative of the Illinoian or of the Iowan of America; the drift of the Third Glaciation in Germany and in Holland is thin or scanty like our Iowan. This leaves the Second and First glaciations of Europe, which seem to correlate with the Second and First glaciations of America.

(10) Coming to the question of harmonizing Penck's interpretations with those of De Geer, Leverett is not clear as to the precise way in which the one interpretation would fit into the other. There is in this case the need for further careful field-work to clear up the matter satisfactorily. He states that he would not venture an opinion as to whether the morainic system which separates De Geer's "Gothiglacial" from his "Finiglacial" is the correlative of Penck's moraines of the Daun substage, or whether the Daun moraines find their equivalent in later moraines of Sweden. It is unlikely that the Daun moraines are older than those which mark the division between Gothiglacial and Finiglacial in Sweden. In De Geer's "Quaternary sea bottoms in western Sweden" [and Antev's (1917.1) "Post-Glacial marine shell-beds in Bohuslän"] we have evidence of remarkable diastrophic complexity, which would serve to show the danger of following Depéret in his idea of simplicity of earth movements.

(11) Leverett notices that De Geer seems to think that his determination of 11,000 years for the time since the ice disappeared from the northern part of the Baltic region is inconsistent with the American estimate of over 30,000 years for the duration of Niagara Falls, and says it is perhaps four times too long. This leads Leverett to remark that it is likely that the ice in North America was reduced to a small area on the east side of Hudson Bay at the time it was melting out of the north end of the Baltic basin. In that case Niagara history is much longer. The part that corresponds to the life of Lake Algonquin would entirely precede it, and probably a considerable part of the life of Lake Nipissing. It seems to Leverett, therefore, that we are under no necessity of discarding the estimate of about 30,000 years for the life of Niagara. The estimate is, however, of a less refined nature than that made by De Geer, and is so understood by those of us who have followed Spencer and also

Taylor (1913.1) in the methods of making the estimates. Leverett is of the opinion that the Niagara estimates can not be more than 5,000, or at most 10,000, years out of the way; that the life of Niagara must have been at least 20,000 years, and that it may have been 35,000 to 40,000 years. The Niagara Falls Folio enables us to judge as to the quality of Taylor's work and the value of his estimates. Spencer's methods are brought out in his work on Niagara published by the Canadian Geological Survey.

(12) Leverett considers that, inasmuch as the ice of North America is likely to have been reduced to a small field in the part of Canada that is difficult of exploration, it may be some time, in the present sparsely settled conditions, before we can correlate the position of the ice with the Daun substage, or even with substages corresponding to the Gschnitz and the Bühl of Penck's classification. These may all be found in the wilds of Canada.

VIEWS OF WILLIAM C. ALDEN, OF THE U. S. GEOLOGICAL SURVEY

(13) William C. Alden agrees with Leverett (February 17, 1922) that it is not possible to substitute eustatic (sea movement) for diastrophic (land movement) as Depéret has done. While it is probably true that the net result of all crustal movements was downward, under gravity action, yet there have undoubtedly been upward earth movements both widespread and local; to these diastrophic movements Alden believes all changes of sealevel are attributable, except those actually due to the abstraction of water for storage in ice-sheets, and the subsequent return. of the water to the sea-basins.

(14) As regards the beach levels, Alden remarks that it is very interesting to have the data brought together in the way Depéret has done, but that the results can not be regarded as final; nor is it probable that we can establish the Pleistocene classification on the basis of sea-beach levels. Only where direct connection is traceable between the glaciofluvial terraces and the marine terraces might there be a basis for at least tentative correlation between the glacial and marine phenomena. There is an opportunity to test out the matter of correlating glacial and marine features right in the vicinity of New York, and it may be hoped. that the coordination made by Depéret will be a stimulus to American glaciologists to work out the relations carefully.

VIEWS OF EMMANUEL DE MARTONNE AND DOUGLAS W. JOHNSON

(15) Emmanuel de Martonne writes March 29, 1922, in reply to a number of questions raised by Douglas W. Johnson, that he was at first

skeptical of the eustatic movement theory and of the conclusions reached by General de Lamothe, but that the studies of terrace and shoreline altitudes have multiplied in the last few years in the Mediterranean and France with such concordant results that one can not escape being impressed by them. Everywhere are found the terraces of 15-20 meters, 30-35 meters, 50-60 meters, and 90-100 meters. Everywhere are found the valley ways carved below the present level of the sea near mouths of great rivers and for a considerable distance upstream. Particularly striking are the observations of Gignoux (now professor in the University of Strasbourg), who has found in Italy that the levels of Depéret and de Lamothe are very well marked where there have been no recent dislocations. Chaput, although he was rather skeptical when he began his studies upon the terraces of the Loire, was convinced by the evidence, and found the 15 and the 30 meter level everywhere. There seems to be some question as to the upper levels in the Sahel of Algeria indicated by de Lamothe. In developing his theory, Depéret, it appears, has not been strict enough in the choice of evidence supporting the trace of the upper shorelines; that he is too systematically opposed to continental movements (diastrophism); that there are some contradictions in his argument. He considers the Saxonian (Second glaciation of the north) as the equivalent of the Mindelian (Second Alpine glaciation) and of the Milazzian (epoch of the shore of 55-60 meters, on the Mediterranean); but he tells of terraces of 33 meters referred to the Saxonian.

(16) There is something rather astonishing in the statements of Depéret about the altitudes of cold and warm faunas; for example, the deposits with Yoldia arctica at 13 meters, near Hamburg, would be contemporaneous with those of the warm fauna of Neudeck at +100 meters, the first being formed at the bottom of the sea, which was cold, the second near the shore; so that the deposits with Yoldia which one finds at sealevel in England, would indicate an old shore of +100 meters. . . . However, when Depéret defines the shore of 100 feet (33 meters), as later than the Polonian (Third glaciation of the north), it is by the deposits with Yoldia which are at this altitude.

(17) De Martonne recognizes difficulties in the application of the eustatic theory. The lowering of scalevel must be very rapid, being due to the sinking of submarine depths; but the rising of the water must be very slow, since it is due to sedimentation; the result is that the shoreline, during the transgression, must be constantly displaced. Depéret notes, however, after the regression below present sealevel, an uplift of +33 meters, which maintains itself at this level during two glacial periods and an interglacial period at least. Concerning the Côtes de

Provence, with which de Martonne is acquainted, it appears that there is in this region, as in Algeria, a very constant littoral terrace at 15 meters, visible on all the promontories, which he attributes to marine abrasion; also that shore deposits have been found in place there. There are also indications of higher levels, but these are not very clear. According to de Martonne, the terraces of the valley of the Rhone and that of the Durance are more numerous than de Lamothe and Depéret say, and are manifestly influenced by glaciers; but it is not true, as Penck says and as de Martonne states that he himself has believed at times, that they all join downstream. There are, particularly near the delta of the Rhone, some very high terraces. De Martonne has not personally studied the famous grotto of Baoussé Roussé near Monaco, where there are human remains, but he has confidence in Boule's very searching monograph, the conclusions of which are favorable to displacements of the sealevel.

(18) De Martonne sums up as follows: We can not deny modifications of sealevel since Pliocene time, but they may be attributable to earth movements (diastrophic), not only regressive but transgressive; nor are these earth movements limited to sea-bottoms. The result is that we can not find everywhere shorelines at the same levels. There is, without doubt, much to accept and much to abandon in the argument advanced in favor of the eustatic theory, but the observations are too numerous not to make a strong impression. We must remain capable of giving fair consideration to earth movements (diastrophism) and to changes in sealevel due to the accumulation and the melting of glaciers. (19) Professor Johnson doubts whether the eustatic theory of a rise and fall of sealevel, advocated by de Lamothe and supported in part by Depéret and Gignoux, can adequately explain the elevated marine shorelines, marine terraces, and river terraces, attributed to the Pliocene and Pleistocene periods, occurring at successive elevations on the shores of the Mediterranean and Atlantic and along the rivers of Europe and northwestern Africa. His own observations in America and Europe do not indicate the existence of a fairly definite series of marine terraces and river terraces, at more or less uniform elevations above the present sealevel and present river channels, either on the shores of the Atlantic or in parts of the Mediterranean seen by him; and he accordingly believes, with the majority of geologists, that differential movements of the land, together with the normal degrading of alluvium-filled valleys, better account for the observed facts. [Johnson inclines rather to the diastrophic theory. It appears from the above résumé of opinion that the new classification of the Quaternary proposed by Depéret is being