forces once more upon this region, it is still heated to near the temperature of crystallization; rock mashing and recrystallization take place in it. The development of differentiation to its final stages gives an acidic magma in the upper part of the reservoir. The escape of the gases in whose presence only this magma can remain molten hastens the crystallization to great depth. The crust is once more strengthened, rock mashing becomes ineffective, the period of orogenic activity as marked by igneous intrusion and dynamometamorphism has come to a close.

ART. XVII.-The Post-Glacial Terraces of Anticosti Island; by W. H. TWENHOFEL and W. H. CONINE.

Introduction.

Detailed description of the terraces.

The sea-level terrace.

The terraces above sea-level.

Origin of the terraces.

Time of origin of the terraces.
Glaciation of the island.
Champlain submergence.
Deposits of post-Champlain time.
Age of the river valleys.

Conclusions as to time of origin.
Correlation with terraces elsewhere.

INTRODUCTION.

There are probably no more impressive features in the physiography of Anticosti Island than its terraces. Like stairways for giants, they begin at the level of the seapossibly below sea-level-and continue into the interior as far as the eye is able to follow. The highest measured exceeds 400 feet. The cliffs or steep slopes which front them vary greatly both in height and in extent. In places a terrace may have a width of a mile or more; in other places it narrows to disappearance, with its cliff or front slope merging with the cliff or front slope of one of those above. The places of greatest width are about the indentations of the coast; the places of least width are on the salients.

Anticosti Island is a cuesta with the escarpment facing north, the dip slope facing to the south. The northern channel occupies the inner lowland. The trend of the island is not parallel to the structure nor is the strike of the strata identical with the trend of the island. gressively younger strata are met in proceeding eastward along the north side, progressively older strata in going westward and eastward from Southwest Point along the south side.

Pro

The asymmetry of the island's surface with respect to its north and south slopes is reflected in its terraces. On the south side, some of them are several miles wide, and to reach the summit of the highest, it is necessary, in most places, to go somewhere near the middle of the island. On the north side, the highest observed terrace may be reached in many places within a couple of miles

of the sea and there are a few places where it is within a mile. The greater width of the terraces on the south slope lessens the impression they make and, except where a high headland reaches the sea, not more than three or four may be seen in most localities; on the north side, on the other hand, the terraces can not fail to attract the attention of even the most casual observer and there are many places where more than a half-dozen may be seen.

At the time of the senior writer's first visit to Anticosti in 1908, the terraces were noted-it would have been impossible not to have seen them—and a few estimates were made as to their number and heights. On the second visit in 1919, a consistent effort was made to obtain exact data relating to them and Mr. Conine was instructed to give all the time at his disposal to this end. In addition, measurements were made at many places by the senior writer and other members of his party. The result is that elevations and width of terraces have been obtained at several dozen localities of which only a few are shown in the diagrams. Measurements were ordinarily made by hand leveling and pacing, but in a very few instances heights and widths were estimated. Wherever possible, the nature of the materials on the surfaces of the terraces was determined. This could only be done, however, on the edges of cliffs and about the roots of upturned trees, neither of which is everywhere present.

Collection of data relating to the terraces is simple, but not always easy; nor are the results obtained altogether accurate. As soon as one leaves the beach, he finds himself in an almost impenetrable forest through which he rarely can see more than a couple of hundred feet at most. The problem for most of the terraces was confined to the obtaining of their heights and widths at localities which appeared likely to give the most accurate results. Except for the lowest terrace, little effort was made toward correlation, as this is impossible except by the actual walking out on each terrace entirely around the island, and to do this for each one, or for any one, was out of the question.

DETAILED DESCRIPTION OF THE TERRACES.

The sea-level terraces.-The lowest exposed terrace of Anticosti is the one whose width is at present being ex

tended by the work of the waves. This is known as the "reef" by the people of the island and by the sailors of the Gulf. Its presence is a menace to navigation and it renders many parts of the coast of Anticosti inaccessible from the sea except to the smallest of boats. Its width varies from nothing to about three miles, the greatest width being on the south side at the mouth of Dauphine River, but it is almost everywhere present to some degree except in those bays which have barrier beaches about their heads as is the case at the mouths of Jupiter River, Salmon River, Fox River, etc. On the south side there are not many places where the reef does not have a width of at least an eighth of a mile, and a quarter- to a half-mile width is an extremely common occurrence. On the north side the reef is less marked than on the south, but even there it is commonly from a sixteenth to an eighth of a mile wide.

The reef shows little other than barnacle- and seaweedcovered rocks. Pebbles are generally wanting except at the shore, where they are apt to be in quantity unless the rock of the coast is composed of material from which they could not readily be derived. Were the shells and seaweeds removed from the reef, there would be little or nothing present to show that it is of wave-cut origin; morever, it is probable that if the waves should cut entirely across the island so as to develop a reef over the entire extent of its present area, there would be no conglomerate to show the wave-cut origin, and, were the reef then submerged, there would be no basement conglomerate to mark the unconformity, but sands and muds would rest directly across the truncated edges of the limestones, shales, and sandstones which constitute the rocks of the island.

The terraces above sea-level.-The terraces above sealevel are tabulated in the list which follows. It is probable that the elevations which are given contain some error for all terraces above the eighth, this being a necessary consequence of the method of measurement and the conditions under which measurements were made. Another factor introducing an error is the fact that every terrace is covered with vegetable mould and peat of which the thickness was not always determinable. So far as possible, however, an estimate has been made and a deduction allowed therefor.

Terraces whose elevations are considered most accurate are designated by an asterisk. Those which have been observed in the greatest number of places are the 1st, 2nd, 3rd, 6th, 7th, and 10th. For statement of elevation at the various localities, the reader should consult the diagram (fig. 1).

The 5 to 6-foot terrace has been seen at hundreds of localities and in many places it is fronted by a littoral barrier which in some instances is 12 to 15 feet above tide

FIG. 1.-Profiles of terraces at various localities. Each vertical space equals 25 feet, each horizontal 25 feet.

1, 5-6 feet*; 2, 10-13*; 3, 20 feet*; 4, About 29 feet; 5, 40 feet*; 6, About 51 feet*; 7, 60-65 feet; 8, 74 feet; 9, 85 feet*; 10, About 95 feet; 11, 105 feet; 12, 115 feet; 13, About 120 feet; 14, About 130 feet; 15, 145-150 feet; 16, About 180 feet; 17, 210 feet; 18, About 300 feet; 19, 344 feet; 20, 380 feet; 21, 409 feet; 22, 442 feet.

level, thus attesting to the power of the waves on those portions of the coast. This terrace is ordinarily not of great width, but it is commonly a hundred or more feet wide and there are places where it approaches a width of a halfmile. On its landward margin there are not uncommonly hooded or overhanging cliffs at the feet of which are talus slopes related to the cliffs in the manner shown in figure 2, which represents the elevated cliff between West Point