elevated zone (Figs. 2 and 3). While mature willow and alder bushes grow above the abandoned strand, the elevated beaches and benches support only scattered annual plants and occasional young alders and willows (Figs. 4 and 5). Many of these bushes were cut down and the rings of annual growth counted; but in no case was one found on the raised shore-line with more than five rings, while most of them had but three or four. This fact points quite clearly to the season of 1899 as the time of uplift.

Bearing upon this point of the exact period of change of level is the human testimony. In the first place, Prof. I. C. Russell spent some time in this fiord in 1890 and 1891, and he reports no change of level. Dr. G. K. Gilbert spent several days in the Yakutat bay inlet in June, 1899, landing at several points in the fiord; in one place on a coast where the uplift amounts to 18 feet. Moreover, the ship in which he travelled sailed twice close by one of the new uncharted reefs, which could hardly have escaped the attention of sailors. Both Prof. Russell and Dr. Gilbert have made a special study of abandoned shore-lines, and it is inconceivable that these geographers could have failed to observe the uplift if it had existed during their visit. A photograph taken by a United States Fish Commission expedition in July, 1901, shows clearly the presence at that time of an uplifted bench at one point (Haenke island) in the inlet.

The testimony of natives is definite as to the time of occurrence of the change of level. Every spring they spend two months or more in the inlet hunting the hair seal, so that they are familiar with the region in detail. They assert positively that the uplift occurred in September, 1899, during a series of earthquake shocks which lasted for seventeen days.

Of the occurrence of this earthquake there is unquestioned proof. Entirely aside from the testimony of the natives, there were three prospectors encamped on the shores of the inlet during the heaviest of the shocks; and at the village of Yakutat, at the mouth of the bay, there are a number of white men whose account tallies with that of the natives and of the prospectors. The shocks began on September 3 and ceased on the 20th, reaching greatest intensity September 10 and 15. On September 10, between 9.20 a.m. and 3 p.m., there were fifty-two notable shocks, culminating in one of great violence. The prospectors report a great water-wave in the fiord during the most vigorous shocks; and the inhabitants of Yakutat, 15 or 20 miles away from the centre of greatest disturbance, were so alarmed by the shaking that they abandoned their houses and retreated to tents on the neighbouring hills.

Signs of the earthquake are still visible at various points in the inlet. In the first place, the mountain slopes are scarred with great avalanches, far more numerous than in other fiords on the Alaskan coast. In the second place, there are wave-swept areas (Fig. 7), in two sections reaching to an elevation of 40 feet above sea-level, throughout which the forest is completely destroyed. Here trees are overturned, twisted, broken, and uprooted, giving rise to such a scene of devastation as only rushing water could produce.

In this sparsely settled region it is natural that the extent of the earthquake should not be exactly known. That it was not confined to this limited area is probable; and in this connection it is noteworthy that in September, 1899, an earthquake of great violence occurred in Glacier bay, 135 miles from Yakutat bay. This earthquake was so

violent that it shattered the front of the Muir glacier to such an extent that for several years the tourist steamer was not able to penetrate the bay to the Muir glacier. In all probability the uplift which accompanied this shock affected the entire region between Yakutat and Glacier bays, and doubtless for a considerable distance beyond this area. As yet, however, no evidence of change of level has been reported outside of Yakutat bay.

In connection with our other studies, we made measurements of the amount of change of level at various points along the 150 miles of coast-line which we visited, making in all over one hundred careful measurements, besides numerous observations for checking these. The numerical results of these observations are plotted on the accompanying map (Fig. 1). From this map it will be seen that the amount of uplift varies greatly from place to place. One of the most noteworthy features on the map is that there is a zone just outside the mountain front, on one side of which there is uplift, and on the other side, in general, either no change of level or else a depression. This is true at the head of the inlet and on both shores of Yakutat bay. From this and other facts it is inferred that a fault-line extends here as indicated by the line A. The very variable amount of change of level along this line is believed to be the result of a drag along the fault. It should be stated that the straight mountain front which rises above the foreland close by this inferred fault was interpreted by Russell as a fault scarp in 1890 on the basis of its form alone. We reached the same conclusion when we first saw this mountain front, and before we knew of the recent change of level.

The complex condition of change of level in the archipelago of islands north of Yakutat Cannery and Mission may in part be due to a shaking of the gravels of which these islands are made; but this would not account for the uplifted sections. We have, therefore, inferred a minor fault along the axis of these islands; but the evidence of this fault-line is less definite than that of the others.

On the basis of its form, Prof. Russell also classed the straight mountain front along the eastern shore of Yakutat bay as a fault scarp; and here again our evidence tends to verify his interpretation. The evidence of faulting along this line, which we have indicated by the letter B, is of several kinds. In the first place, the straight, steep, truncated mountain spurs are suggestive of fault origin. In the second place, there is an unusual abundance of avalanches along this mountain front. In the third place, the change of level along this line varies greatly, as would be expected along a line of faulting with updragging of the strata on the downthrow side. In the fourth place, at one point (marked Old Beach) there is a still earlier uplifted strand on which a mature forest is now growing, with trees over seventyfive years old, proving a former period of uplift.

Along the shores of Disenchantment bay there are great differences in the amount of uplift. Throughout the greater part of the eastern shore the uplift is between 7 and 9 feet. On Haenke island and on the peninsula just north-east of it, the uplift ranges from 17 feet to 19 feet 3 inches. On the western shore of Disenchantment bay the uplift varies from 33 feet 11 inches to 47 feet 4 inches, the latter being the greatest change of level in the inlet, and, in fact, the greatest uplift ever recorded as having occurred at a single period of time. To account for the difference in level of the upraised strand in Disenchantment bay, we have inferred two fault-lines (C and D), as indicated on the map.

Along the north-west arm of Russell fiord the south-west shore shows very slight uplift, while the north-east shore is raised from 7 feet 6 inches to 9 feet. Moreover, along the north-east shore there is an older beach covered by a growth of mature alders. This shore of the fiord is made of crystalline rocks, while the younger Yakutat series forms the south-west shore. Geological evidence proves the presence of a fault-line between the Nunatak fiord and the valley of Hidden glacier, which, extended, would pass along the axis of the north-west arm of Russell fiord, where, from the evidence of the upraised shore-lines, we have inferred fault E.

From the region of slight uplift in lower Russell fiord there is a gradual rise of the abandoned strand toward the mountain front, where it reaches a maximum of 9 to 10 feet.

In inferring our fault-lines we have endeavoured to be conservative, and to postulate no more than the evidence definitely calls for. Our studies, however, prove conclusively that there was pronounced dislocation in other parts of the region. At several points we found small recent faults. These were best shown on the Gannett Nunatak, at the head of Nunatak fiord, and a description of these may serve for the rest. This nunatak is a glaciated rock hill rising about 1450 feet above the fiord, and composed of steeply dipping gneisses and schists, striking approximately parallel to the major axis of the St. Elias chain-that is, north-westward. In its southern half it is crossed by scores of small faults extending from a few feet to over 100 yards, and with throws varying from an inch to 3 feet, but usually of less than a foot (Figs. 8 and 9). They generally extend along the strike of the rock, but in some instances diverge from it, and a few short faults strike at right angles to the main series. The hade is nearly vertical, and in almost all cases the south-western side of the fault is the upthrow side. There are also some fissures, and a few instances of small graben blocks (from 3 to 30 feet wide) between parallel faults.

While it cannot be demonstrated that these minor faults were formed at the same period as the uplifted strands, this inference is nevertheless apparently warranted, for the faults are evidently very recent, and no other notable earthquakes have recently occurred here. The fault scarps are steep, and in many instances have striæ extending up to their very edges. Their recency is proved by the sharp angle formed where the fault-plane intersects the surface, and by the absence of notable talus slopes at the base of most of the tiny fault-scarps, even though the rocks are friable, thin-bedded schists. It seems incredible that these fault-scarps can have been exposed to the sharp frost-action of this climate longer than six years.

From our observations we draw the conclusion that this part of the St. Elias chain is still actively growing, and that this growth is being accomplished by movements along a series of fault-lines by which several mountain blocks are being differentially uplifted, as first suggested by Prof. Russell. At least one period of earlier uplift is demonstrated by the presence of the older forest-covered beaches; but a much more notable mountain growth occurred in September, 1899. This uplift involved the entire mountain region inside of the Yakutat foreland as far as the bay reaches, and to an unknown distance beyond. It consisted of a general uplift along the mountain front, and of a differential uplift along several secondary lines of faulting: in other words, the mountains are tilted in a series of fault blocks. In addition to these major lines of faulting there was a minor fracturing, apparently due to local adjustments in the tilted blocks.

This recent change of level is noteworthy from three standpoints: Its age can be definitely determined; it is the greatest uplift so far recorded as having occurred at a given time; it contains a lesson of importance as to the mode of formation of mountains, representing as it does a step in the development of the loftiest range on the American continent.

ADDENDUM. - THE SAN FRANCISCO EARTHQUAKE OF APRIL 18.

Since this article was written, and just as it is being made ready for mailing (April 19), news comes of the terribly destructive earthquake which has caused such devastation in California. While as yet the meagre dispatches give us little information of scientific value regarding this most recent of vigorous earthquakes, enough is known, both from the past history of California and from the newspaper accounts of the present shock, to make it certain that the conditions described above are distinctly applicable to the San Francisco earthquake. American geologists have long been familiar with the fact that the Coast Ranges of California, like the mountains of Alaska, are still in the process of evolution. The frequent earthquakes which have been felt in California, numbering from one to three score a year, are proof of this. Geological study demonstrates that the rocks of the Coast Ranges are crossed by many fault-lines; and physiographic investigations have shown that recent earth-movements have upraised parts of the Californian coast and depressed others, including the bay of San Francisco, which is the drowned mouth of the Sacramento river, into which the sea has been admitted by local subsidence.

It

Where the fault-lines along which the slipping occurred to cause the San Francisco earthquake are located, whether there is but one line of slipping or several as in Yakutat bay-and whether the movements have produced visible signs of uplift or depression, as in Alaska, are questions which future investigation must answer. seems evident, however, that the shock is the result of a normal process of mountain-building here as in Alaska. The strains to which the mountain rocks are subjected have locally become too great, and relief has been gained by a slipping of the rocks over one another -probably on one or more planes of older faulting, along which previous