journey of all drift a long one, and adds greatly to the destruction of the bergs by stoppages and by causing them to break up. Disintegration is also hastened by their breaking away from the floe ice, for detached bergs will melt and break up rapidly even in high latitudes during the summer.

549. THE ICE-BEARING CURRENTS.-The Labrador Current passes to the southward along the coasts of Baffin Land and Labrador, and, although it occasionally ceases altogether, its usual rate is from 10 to 36 miles per day. Near the coast it is very much influenced by the winds, and reaches its maximum rate after those from

the northward. The general drift of the current is to the southward, as shown by the passage of many icebergs, although occasions have arisen on which these have been observed to travel northward without any apparent reason. The breadth and depth of the current are not known, but it is certain that it pours into the Atlantic enormous masses of water for which compensation is derived from the warm waters of the Atlantic and from the East Greenland Current that flows around Cape Farewell. The flow of the Polar Current down the east coast of Greenland has been abundantly demonstrated by the drift of vessels that have been beset in the ice pack to the east

60 miles wide, and then takes a northwesterly direction along the Greenland coast as far as the Arctic Circle, where it meets the southerly current from Baffin Bay.

550. DRIFT AND CHARACTERISTICS OF ICEBERGS.-Not all the bergs made in any one season find their way south during the following one, for only a small percentage of them ever reach trans-Atlantic routes. So many delays attend their journey and so irregular and erratic is it that many bergs seen in any one season may have been made several seasons before. If bergs on their calving at once drifted to the southward and met with no obstructions their journey of about 1,200 to 1,500 miles would occupy from 4 to 5 months, reckoning the drift of the Labrador Current at 10 miles a day, which may be making it too little. Then, if bergs were liberated principally in July and August they should reach trans-Atlantic routes in December and January, while we know this to be the rare exception. It is then seen what an important bearing the shores of Labrador have in arresting their flow, when it is known that bergs are generally most plentiful in the late spring and early summer months off the Bank.

It should not be supposed that all bergs follow the same course when set adrift from their parent glaciers, for, like floating bodies at the head of a river, some will go direct to the mouth, others will go but a short distance and lodge, others still will accomplish half the journey and remain until another freshet again floats them, so that in the end the debris will be composed in part of that of several years' production.

Bergs, when first liberated on the west Greenland shore, are out of the strongest sweep of the southerly current, and they may take some months to find their way out of Davis Strait, while again others may at once drift into the current and move unobstructed until dissipated in the Gulf Stream. The difference in time of two bergs reaching a low latitude, which were set adrift the same day, may cover a period of one or two years.

Field ice also offers an obstruction to bergs, and a close season in the Arctic may prevent their liberation to a great extent, though, from their deep submersion, they act as ice plows and aid materially in breaking up the vast fields of ice which so often close the Arctic Basin.

Ice fields are more affected by wind than bergs. Bergs owe their drift almost entirely to current, so that they will often be noticed forcing their way through immense fields of heavy ice and going directly to windward. Advantage is taken of this by vessels in ice fields, which often moor to bergs and are towed for miles through ice in which they could not otherwise make any headway. This is accomplished by sinking an ice anchor into them and using a strong towline, and as the berg advances open water is left to leeward while the loose ice floats past on both sides. For the same reason vessels, when beset by field ice, run from the lee of one berg to that of another, as leads may offer themselves.

Instances are not rare where icebergs were seen to drift toward north, making 15 to 24 miles a day, near the tail of the Bank and to the eastward of Cape Race. All ice is brittle, especially that in bergs, and it is wonderful how little it takes to accomplish their destruction. A blow of an ax will at times split them, and the report of a gun, by concussion, will accomplish the same end. They are more apt to break up in warm weather than cold, and whalers and sealers note this before landing on them, when an anchor is to be planted or fresh water to be obtained. On the coast of Labrador in July and August, when it is packed with bergs, the noise of rupture is often deafening, and those experienced in ice give them a wide berth.

When they are frozen the temperature is very low, so that when their surface is exposed to a thawing temperature the tension of the exterior and interior is very different, making them not unlike a Prince Rupert's drop. Then, too, during the day water made by melting finds its way into the crevices, freezes, and hence expands, and, acting like a wedge, forces the berg into fragments. It is the greatly increased surface which the fragments expose to the melting action of the oceanic waters that accounts for the rapid disappearance of the ice after it has reached the northern edge of the warm circulatory drift currents of the North Atlantic Ocean. If these processes of disintegration did not go on and large bergs should remain intact, several years might elapse before they would melt, and they would ever be present in the transoceanic routes. In fact, instances are on record in which masses of ice, escaping the influences of swift destruction or possessing a capability for resisting them, have,

by phenomenal drifts, passed into European waters and been encountered from time to time throughout that portion of the ocean which stretches from the British Isles to the Azores.

Icebergs assume the greatest variety of shapes, from those approximating to some regular geometric figure to others crowned with spires, domes, minarets, and peaks, while others still are pierced by deep indentations or caves. Small cataracts fall from the large bergs, while from many icicles hang in clusters from every pro

jecting ledge. They frequently have outlying spurs under water, which are as dangerous as any other sunken reefs. For this reason it is advisable for vessels to give them a wide berth, for there are cases on record where vessels were seriously damaged by striking when apparently clear of the berg. Among these is that of the British steamship Nessmore, which ran into a berg in latitude 41° 50' N., longitude 52° W., and stove in her bows. On docking her a long score was found extending

broken and the plates were almost cut through. The ship evidently struck a projecting spur after her helm had been put over, as there was clear water between her and the berg after the first collision.

It is generally best to go to windward of an iceberg, because the disintegrated fragments will have a tendency to drift to leeward while open water will be found to windward. Serious injury has occurred to vessels through the breaking up or capsizing of icebergs. Often the bergs are so nicely balanced that the slightest melting of their surfaces causes a shifting of the center of gravity and a consequent turning over of the mass into a new position, and this overturning also frequently takes place when bergs, drifting with the current in a state of delicate equilibrium, touch the ocean bottom.

551. FIELD ICE.-Field ice is formed throughout the region from the Arctic Ocean to the shores of Newfoundland and yearly leaves the shore to find its way into the path of commerce. Starting with the Arctic field ice and coming to the southward, we find this ice growing lighter, both in thickness and in quantity, until it disappears entirely. Ice made in the Arctic is heavier and has lived through a number of seasons. After the short summer in high latitudes ice begins to form on all open water, increasing several feet in thickness each season. Much of this remains north during the following summer, and, though it melts to some extent, it never entirely disappears, so that each succeeding winter adds to its thickness.

This continues from year to year until it reaches 12 or 15 feet in thickness, often more. If it remained perfectly quiet it would be of uniform thickness, increasing with the latitude, but it is in a state of almost continual motion, often a very violent one, which causes it to raft and pile until it becomes full of hummocks and other irregularities. Immense fields are detached from the shore and from other fields, and under the influence of winds, currents, and tides are set in motion and kept continually drifting from place to place; after a snow, thaw, or piling the whole becomes cemented together into solid pieces, when under the influence of a low temperature. The space of open water between the fields becomes frozen, joining smaller fields, and making a solid pack which will remain so until the elements again break it to pieces. Along the shores from headland to headland the bays and inlets often remain solid for years, almost invariably through the Arctic winter, but in Baffin Bay and Davis Strait open water can be found at intervals all the year round.

Ice becomes rafted in a variety of ways. If two fields are adrift the one to windward will drift down on the one to leeward; the one which is rougher on its surface gives the wind a better hold and drifts the faster; fields may be impelled towards each other by winds from contrary directions. Ice that is secure to the shore is rafted on its seaward edge from contact with that which is adrift. Fields in drifting often have a turning motion, which is caused by contrary currents, or one variable in strength at different places, or by the friction of a field coming in contact with another field afloat or one attached to the shore. This rotary motion is especially dangerous when a vessel finds itself between two fields. A heavy gale will break up the strongest fields at times and cause them to raft and form hummocks.

Small fragments of bergs find themselves mingled with Arctic fields and become frozen fast. These, when liberated to the southward, are called growlers, and form low, dark, indigo colored masses, which are just awash and rounded on top like a whale's back. They are very dangerous when in ice fields which have become loose enough to permit the passage of vessels through them, and should always be looked for; they can be seen apparently rising and sinking as the sea breaks over them.

During the spring and summer months the bergs, aided by a rise of temperature, so cut up and weaken the ice fields that much ice is loosened and begins drifting out of the Arctic basin. This is joined by that brought from the waters of Spitzbergen by the East Greenland Current, near the sixty-third parallel, whence it flows down the eastern coast of North America, reaching Cape Chidley about October or November. By this time the remaining ice in the Arctic is being cemented into solid fields, while the ice cap is being daily extended to the southward. As fast as fields are detached the open water freezes, and these masses are forced to the southward and can not rejoin the solid pack. With a westerly wind ice formed in Hudson Strait and adjacent waters is swept out and joins the Arctic ice, differing from it only in being a little

Ice begins to form at Cape Chidley about the middle of October, at Belle Isle about November 1, and by the middle of November or 1st of December, the whole coast is solidly frozen. The dates given are approximate and vary from year to year, with many marked exceptions.

The string of ice along the coast of Labrador extends from headland to headland, including the outlying islands, and starting from the heads of the bays works its way out to seaward, forming by the middle of December an impassable barrier to the shore which will probably not be permanently broken until the latter part of April. This ice varies in thickness from 12 feet at the northern extreme to 3 or 4 feet at the southern. During the entire winter the Arctic drift is finding its way down the coast, and is being continually reinforced by fields broken from the Labrador ice. These continue to the southward in the Labrador Current on an average of about 10 miles a day, reaching Belle Isle between the middle of January and the middle of February.

The best example on record of a continued drift from the Arctic is that of Captain Tyson. On October 14, 1871, he and a party of nineteen others were separated from the United States surveying ship Polaris, in latitude 77° or 78° N., just south of Littleton Island, and, being unable to regain the ship, remained on the floe and accomplished one of the most wonderful journeys. After a drift of over 1,500 miles, fraught with danger from beginning to end, they were picked up about six months later, April 30, 1872, by the Tigress, a sealing steamer from Newfoundland, near the Strait of Belleisle, in latitude 53° 35′ N., and carried safely into port.

Much delay in the southward movement of the drift will be caused by winds from the southward of west, as field ice is affected more by wind than current. The prevailing wind and weather will influence the drift very greatly. Strong northerly or northwest winds will increase its speed, but contrary winds will hold it back. The string of shore ice keeps the northern ice off the coast and in the current. At times westerly winds will also send the Labrador ice off the coast and leave it entirely clear, but this does not happen often. Still the outer Labrador ice is constantly being added to the Arctic flow. Frequently the bays remain frozen over until June; again, they are cleared some years in April, making a large variation. During the drift the wind from northwest to southwest will clear the ice off the coast and leave a line of open water, but the ice will be set on the coast by a northeast wind and be rafted and piled. The appearance of the ice when it reaches Belle Isle and to the southward would be a fair indication of the weather it had encountered on its way down. The rougher the ice the more severe the weather. This floating ice string extends approximately 200 miles offshore in the latitude of Cape Harrison, and spreads more during its drift, though narrower farther north. One small stream finds its way through the Strait of Belleisle, while the greater part continues toward the northern limit of the Gulf Stream. By the middle of January the shores of Newfoundland and Gulf of St. Lawrence are full of ice, which has been frozen there and are opened or closed by a favorable or adverse wind. Navigation in the River St. Lawrence is closed about the middle of November and does not open until about May. A wind from northwest to southwest will clear the eastern coast of Newfoundland, while the Gulf of St. Lawrence may remain full of ice until the 1st of May. Even after this date much ice is found in the Gulf until July, and by August or earlier the field ice is replaced in the Strait of Belleisle by bergs.

In the bight from Cape Bauld to Fogo Island a string of ice is often found joining these points, hemming in the shore for weeks at a time.

With each northwest or westerly wind the ice is cleared off the Newfoundland coast, except from some of the deeper bays, and carried out to sea, and frequently before the Arctic and Labrador ice has passed Belle Isle the Newfoundland ice has found its way as far south as latitude 45°. In the same way the Labrador ice sometimes precedes the Arctic ice, while all may arrive at nearly the same time. Ice fields often lose their identity, as coming from any one particular place, by the constant intermingling on its southern journey with ice made in a lower latitude.

With easterly winds the field ice and icebergs may block the harbors on the east coast of Newfoundland until June or even July, but these harbors are usually