It is now generally conceded by those who have given the subject much attention that the greater portion of North America above the latitude of 39° north to the shores of the Arctic Ocean has been furrowed and scoured by the action of ice.

Vast traces of ancient glaciers are also found in Europe; for it is reported that ice-sheets have left unmistakable marks of having overrun the greater part of the lands lying between the arctic seas and the latitude of the Pyrenees.

In Asia evidences of glacial action have been noticed from Northern Siberia to the mountains of Syria.

The great glaciers of Himalaya have in times past attained gigantic proportions. In Northern China huge bowlders are found scattered over the valleys, and a long distance from the mountains.

The southern hemisphere, in proportion to the extent of its land surface, shows ample traces of former ice action. From the latitude of 38° south to the southern extremity of the western continent there is said to be the clearest evidence of former glacial action in numerous bowlders scattered over the land.

On the shores of the South Pacific, from the Island of Chiloe to Cape Horn, the coast is fringed with deep fiords, which appear to be channelled out by ice, like the fiords of Norway and Greenland. And at this date the mountains of that southern region are covered with snow, and the glaciers which flow down the valleys are said to reach the tide-water as far north as the latitude of 47° south. The glaciers of New Zealand, now of Alpine proportions, during the ice age descended to the sea, and channelled the deep fiords on its south-western coast; and certain traces of glacial action have been observed in Southern Australia, and also in the province of Natal, South Africa.

Kerguelen Land is pierced with deep, narrow fiords, which have the appearance of having been the work of ancient glaciers.

The lands south of the antarctic circle are to-day supposed to be covered by an ice-sheet, of which the great ice barrier surrounding that region furnishes ample proof.

While impressed with the above reports of the work of ancient glaciers, in connection with my own observations along the shores of the several oceans, I have been led to seek for the physical causes which brought about the great climatic changes of past geological ages. And, while having the subject under consideration, I have had my attention directed to the manner in which the great prevailing winds in connection with continental lands are able to move the heated surface waters of the tropical oceans into the colder zones, and also transfer the cold waters of the higher latitudes into the tropical zones.

And it is through this grand movement of the ocean waters that we are enabled to account for the difference in the temperature of places now lying in the same parallels of latitude.

The natural methods for conveying tropical heat into the higher latitudes, and also for excluding it therefrom, are so simple and efficient that on due consideration we are able to conceive how epochs possessing mild climates have been succeeded by periods of frigidity.

It has been admitted by several writers on climatic changes that, should the tropical surface waters of the ocean be moved into the high latitudes in large volume, thus adding their warmth to the heat imparted by the sun, such combined heat would cause a mild climate. And it has been estimated that the amount of equatorial heat moved into the temperate and polar regions of the northern hemisphere by the Gulf Stream alone is equal to one-fourth of all the heat received from the sun by the North Atlantic from the tropic of Cancer to the arctic circle. Still, it appears to me, while viewing the subject from a marine standpoint, that the explainers of climatic changes have never fully comprehended the manner in which the surface waters of the ocean are moved from the tropics into the high latitudes, and returned from the high latitudes to the tropics. Consequently, they have neglected necessary and efficient natural agents in their explanatory theories, and with much learning and ingenuity have laboriously sought to show how great changes of climate could be brought about through other causes.

But when we notice the simple methods employed by nature to-day for transferring the heat of the tropics into the higher latitudes, and also the manner of excluding such heat therefrom, they appear to afford an explanation for the great changes of climate which have taken place during past ages; for it appears that the natural manner of proceeding by which heat is moved from the torrid zone into the high latitudes sufficient to cause a mild climate is through the ocean currents which are constantly set in motion by the great prevailing winds of the globe. These winds, as is well known, blow mostly from the east toward the west in the tropics, and from the west toward the east in the high latitudes.

This counter-movement of the winds, in connection with a continent extending both northward and southward from the equator over many degrees of latitude, such as obtains on the western continent, is abundantly able to create extensive depressions and elevations on the ocean’s surface, and thus cause vast streams of water to move by gravity from the high sea-levels to the low sea-levels; and in this way the tropical waters have been moved during past ages, and to a considerable extent are now moved far into the northern and southern seas.

This transfer of the ocean waters is the main cause of a temperate climate being enjoyed by countries situated in the high latitudes at this age.

But, in order that the tropical currents should be able to flow into the high latitudes, in quantities sufficient to cause all lands and seas situated in such latitudes to enjoy a mild climate, it would be necessary that the land should extend unbroken, or nearly so, from the arctic to the antarctic circles. Thus, with a continent of such vast extent, the westerly winds would blow the surface waters of the ocean away from the eastern shores in the high latitudes, and so cause extensive low sea-levels; while the easterly winds of the torrid zone would heap the surface waters of the ocean against the eastern tropical shores of the continent. Consequently, the warm waters of the tropical high sea-level would be moved by gravity to the low sea-levels of the high latitudes, even to the arctic and antarctic regions, and thus afford them a mild climate. In this way we account for the mild climate enjoyed on lands and seas within the high latitudes during the warm epochs anterior to the glacial periods.

As the western continent is the only land that extends unbroken from the equator to the cold latitudes of both hemispheres, thus affording an opportunity for the prevailing winds to move the tropical waters into the high latitudes, I will call attention to that portion of the continent which extends far southward into the southern ocean, where the winds and ocean currents have the greatest range and power to affect the climate on different parts of the globe. Here we see South America separated from the antarctic continent by a wide channel of deep water, where the westerly winds blow with great force. The space now covered by this interesting channel, owing to its being situated in the high southern latitudes, must have been occupied by a channel of comparatively small capacity, or else an isthmus of low land uniting the southern portion of South America with the antarctic continent during the warm epochs when the beds of the ancient seas of the northern hemisphere contained a considerable portion of the water now swelling the southern ocean.

Therefore, the obstructions which separated the Pacific Ocean from the South Atlantic furnished opportunity for the westerly winds to force the surface waters of the sea away from the leeward side of such obstructions, causing a vast low sea-level, sufficient to attract the tropical waters heaped against Brazil by the trade winds into the southern seas in adequate quantity to cause a mild climate throughout the antarctic regions through long periods of time.

Recent discoveries have proved that these high southern latitudes have been subject to great changes of climate. According to the reports from the Dundee whalers, while searching for seal in the icy seas that surround the South Shetlands, they met with the Norwegian ship “Jason,” Captain Larsen, who had traced the eastern shore of Graham Land to 68° south latitude, noting two active volcanoes.

The same mariner brought from Seymour Island fossil shells and coniferous wood of the Tertiary epoch.

These furnish sufficient evidence to show that a warmer climate once prevailed there.

At the commencement of the glacial age the obstructions which separated the South Pacific from the South Atlantic had become deeply submerged by the sea, which may have been caused by a tendency of the ocean’s waters to move southward or by a comparative small movement in the earth’s crust. But, on account of the stability of the crust of the earth during times so late as the glacial epochs, the submergence of this southern region was probably owing to the movement of the ocean’s waters from the northern hemisphere into the southern hemisphere, which appears to have been brought about mostly through the agency of the great prevailing winds; for it seems to have happened that the prevailing winds on account of the disposition of the lands and seas were able to move more of the ocean waters southward than they moved northward during the age preceding the glacial periods. The waters thus slowly and gradually forced into the high southern latitudes must have deprived the northern hemisphere of their heaviness, and added their weight to the southern hemisphere. Therefore, the waters moved southward could not all be returned to the seas of the northern hemisphere by gravity, for the reason that the earth’s centre of attraction would change in accordance with the weight of water moved from the northern hemisphere into the southern. It will thus be seen that, while the northern seas were drained or became shallow, the augmented southern oceans deeply submerged the region south of Cape Horn, thus widely separating the western continent from the antarctic lands.

Although the south-east trade winds on the eastern sides of the Atlantic and Pacific Oceans extend further northward than the north-east trade winds extend southward, owing to the heated tropical shores north of the equator being more extensive than such lands south of the equator, still, on account of the general weakness of the south-east trade winds at the equator, and also because of the obstructing northern lands, they have during remote times, and at this age, been largely prevented from impelling the surface waters of the sea into the northern latitudes in opposition to the brisk north-east trades. Furthermore, on account of the widening of the oceans as they extend southward, the surface currents setting in the latter direction have more broad and easy passages than the great currents setting northward.

Moreover, the great currents setting southward on the western sides of the oceans south of the equator are also much assisted during the southern summer months by the strong north-east monsoons which prevail along the east coast of equatorial Africa and the east coast of South America as far as the latitude of 30° south.

The South African current is impelled northward by the trade winds down the south-western coast of Africa; but it is debarred from entering the northern latitudes by the Guinea currents, and so turned away into the south equatorial current which flows into the Brazilian stream.

The Gulf Stream is much obstructed in its northern movement by the narrow Florida channel and the opposing arctic currents, and also by the trend of the North American coast eastward; while its return current on the eastern side of the Atlantic has a much less obstructed passage in its southern movement, and, while on its way past the Azores and Madeira Islands, is largely assisted by the prevailing winds.

The Brazil current, with the impelling force of a strong north-east monsoon during the summer season, has no obstruction whatever in its southern passage until it meets with an offshoot from the great drift current of the southern ocean.

And the same favorable conditions are obtained by the great currents setting southward on the western sides of the South Pacific while on their way to the low sea-levels east of Southern Australia and New Zealand. That portion of the equatorial stream of the Pacific which continues west across the Indian Ocean finds no open passage to the northern seas. Consequently, it turns south along the east coast of Africa into the southern seas.

Therefore, this current, in connection with the great currents setting southward east of Australia, offsets the great Humboldt current setting north along the coast of Peru.

In the North Pacific the Japanese current setting northward is obstructed by the narrowing of the ocean; while its return current on the American side has a constantly widening ocean on its passage southward, and also favorable winds to impel the surface waters toward the equator. Still, with all the facilities above mentioned for the movement of the ocean waters into the southern latitudes, it is probable that since the shallow seas of the northern hemisphere were drained, or much diminished, the prevailing winds have not possessed sufficient force to further augment the southern seas, because of the superior weight of the land in the northern hemisphere compared with the lands south of the equator.

It will appear to those who attribute the rotation of the earth as being the main cause of ocean currents that I am too much given over to the wind theory. But I have reason to believe, as Dr. Croll has asserted, that “the winds are the principal cause of the ocean currents, and are not due to the trade winds alone, but to the general impulse of the prevailing winds of the globe.”

Dr. Croll also declares that “all of the principal currents of the globe are moving in the exact direction which they ought to move, assuming the winds to be the sole impelling cause.”

Those who think that the rotation of the earth is the real cause of the movement of the great surface currents of the sea should explain in some reasonable way why the Agulhas current turns west into the Atlantic from the Mozambique stream, and why the Guinea current turns to the east from the main tropical current of the North Atlantic; for it seems that these two great currents move in direct opposition to the rotation theory, while at the same time many things go to show that they receive their motion from the winds. This view of the question will receive further attention in succeeding pages.

It is the opinion of some writers that a difference of temperature and density between the waters of the polar latitudes and the torrid zone is the principal cause of the movement of the surface waters of the ocean from the equatorial latitudes toward the polar seas, and so returned in under-currents; and this is a favorable factor for assisting the winds on some parts of the sea, especially in aiding the Brazil current in moving the surface waters from the high sea-levels abreast Brazil, and the equatorial calm belt of the Atlantic into the southern ocean, and also for favoring the surface currents setting southward on the western sides of the South Pacific and Indian Oceans.

Yet, whatever gravitating force it may possess for assisting the above-named currents, it would also act against the impelling force of the trade winds, while they were drifting the surface waters northward toward the equator on the eastern sides of the several oceans, and also to retard the returning surface currents, while being drifted by the winds southward on the eastern sides of the North Atlantic and North Pacific. Therefore, while it would seem to favor the winds in their work on the one hand, it would act as an opposing agent on other parts of the ocean. Still, the difference of temperature between the tropical and antarctic seas probably does act in opposition to the wide and brisk trade winds on the eastern sides of the great oceans south of the equator, and so prevents their impelling the surface waters northward to a great extent; and this seems to be one great cause of there being less surface water moved northward than southward over the greatest oceans of the globe.

The theory that the difference of density caused by the difference of temperature between the polar seas and the equatorial oceans made under-currents to flow from the polar latitudes, and meet in the equatorial seas, can only be carried on in the Atlantic Ocean, and in a comparatively less perfect way in the Pacific Ocean, and not at all in the Indian Ocean.

The North Atlantic being open to the Arctic Ocean, a portion of the Gulf Stream waters that enter it from the north-west of Europe do sink and return southward in under-currents; and the cold waters which pass down the east and west coast of Greenland also sink under the Gulf Stream while on their southern movement. The meeting of these arctic currents with the cold under-currents from the antarctic seas in the tropical zone is probably one cause of their cold waters rising near the surface of the sea in the torrid latitudes of the Atlantic; and the same conditions probably obtain in a somewhat less degree in the Pacific Ocean.

Yet it appears that the cold waters of the Antarctic occupy the largest space in the tropical zone, even in the North Atlantic. Dr. Carpenter, in his lectures on Ocean Currents, speaks of meeting with antarctic water so far north as the latitudes of the West India Islands; and he also says that all of the Pacific Ocean at its depths is supplied from the Antarctic Ocean, as are the cold under-waters of the tropical Indian Ocean, which extend over twenty degrees north of the equator.

Thus, from what we can learn of the antarctic under-currents, they seem to show that they are not wholly attracted northward on account of the difference of temperature between the antarctic and the tropical oceans, but partly because of more surface water being moved southward by the prevailing winds than they are able to move northward.

And it appears that, if through the winds, combined with the difference of temperature between the antarctic seas and the equatorial waters, and also because of the oceans widening toward the south, more surface water is being carried southward than northward, the waters of the under-currents so caused must rise toward the surface in the latitudes from which they were first removed. Having called attention to the fact that the prevailing winds are not able at this date to augment the southern ocean waters from the scanty northern seas, because of the preponderance of northern lands, still there is reason to believe that even now, owing to the form of continents and oceans, and the attraction of the tropical surface waters into the Antarctic Ocean because of the difference of density between the warm and cold seas, the prevailing winds of this age are able to force more of the surface waters of the sea southward than they force northward; but, owing to the superior weight of the land in the northern hemisphere, the surplus surface water forced into the southern seas is returned by gravity after being cooled by the antarctic ice, and so adding to the deep under-currents which flow with a sluggish movement over the bottom of the sea into the tropical and northern temperate latitudes. And in this way the northern oceans are maintained at their present sea-level.

The cold under-currents are probably assisted in their northern movement by whatever difference there may be in the density of the antarctic waters over the bottom waters of the equatorial seas. But, as such currents extend into the northern tropical latitudes of the northern hemisphere, it seems that the winds are the main cause of the under-currents which carry so much antarctic cold into the northern tropical seas, because the winds have forced an undue proportion of ocean surface water southward, to be attracted northward in under-currents by the preponderating northern lands.

Yet, notwithstanding the superior weight of land in the northern hemisphere, it appears that there have been periods when there was somewhat more water in the oceans of the southern hemisphere than now; for it is reported that a portion of the low lands of Australia show traces of having been submerged during late geological times.

This may have happened through an increased weight in the antarctic glaciers, which have in past ages, and probably may in future epochs, cause more of the ocean waters to be attracted southward than now obtains. But it is probable that an increase of southern ice would be largely counterbalanced by the accumulation of ice on northern lands.

Yet it appears certain that since the Tertiary epoch the waters of vast shallow seas have been moved from the northern hemisphere into the southern. The dry beds of the ancient northern seas encourage this opinion, while the comparatively small area of southern lands serves to support such views.

Still, during the ages prior to the glacial periods, while the low lands of the northern hemisphere were covered by the sea, the wide shoal channels which submerged the lower portion of North America afforded convenient passages for the surface waters of the ocean in their northern movement, and so prevented the oceans of the southern hemisphere from gaining undue preponderance.

Hence long geological ages passed away before the winds were able to force more of the ocean waters southward than they could move northward, and thus augment the southern ocean from the waters of the northern seas. But the slow growth of such immense marine deposits in the shallow seas as are found in the Florida Peninsula and other portions of that region was at length sufficient to greatly obstruct the passage of the Gulf currents in their northern movement, and thus cause conditions which enabled the winds to force more of the ocean waters southward than they could move northward after the close of the Tertiary epoch.

Mr. Alfred R. Wallace says in “Island Life” that the seas in the northern hemisphere during the Tertiary period covered a much larger area than now, and extended across Central Europe and portions of Western Asia, and the Arctic Ocean was enlarged.

As it is not likely that any portion of the waters of the sea have been absorbed by the earth during the late epochs in the world’s history, therefore the ocean waters have not diminished except during cold periods, when the water evaporated from the sea was converted into ice, and, eventually, again returned to the sea.

Thus it necessarily follows that, when the seas of the northern hemisphere contained a much larger portion of the waters of the globe than at this age, the seas of the southern hemisphere must have contained proportionally less. Consequently, during such times a portion of the shoal seas of the high southern latitudes must have been dry land. Therefore, this must have been the condition of the shallow sea basins in the region of Cape Horn.

Mr. Wallace also says that “many peculiarities in the distribution of plants and some groups of animals in the southern hemisphere render it almost certain that there has sometimes been a greater extension of antarctic lands during Tertiary times.”

And he also asserts that the great ocean basins have not changed, and that the form of continents has been permanent. It will thus be seen that it was through the movement of the ocean’s waters southward that the low lands south of Cape Horn were covered with water previous to the frigid periods, and so caused the wide separation between the western continent and the antarctic lands.

The Cape Horn channel thus enlarged, the continuous mildness of the high southern latitudes which possessed the earlier ages came to an end, and gave place to alternate epochs of frigid and mild weather. For it appears that it is owing to the creation or enlargement of the Cape Horn channel that it is possible for frigid periods to be brought about, for the reason that its enlarged space of water prevents the westerly winds from maintaining a great low sea-level in the higher latitudes of the southern ocean; for, whenever the capacity of the Cape Horn channel is enlarged, the westerly winds, instead of maintaining a low sea-level on the South Atlantic, employ their force in impelling the surface water of the southern seas around the globe. And this work the strong westerly winds of the high southern latitudes have always accomplished whenever the Cape Horn channel was widely open, and this is what the winds are doing at this date.

Therefore, such waters of the torrid zone as are moved southward from their high sea-level, caused by the trade winds abreast the Brazilian coast, are largely turned away from the high southern latitudes. It is true, even with an enlarged Cape Horn channel, they can always flow along the South American coast to an inferior low sea-level, caused by the westerly winds blowing the surface waters of the sea away from the coast of Argentine and Patagonia; but on gaining that region they meet the cold ice-bearing currents which turn away east of Cape Horn from the great southern drift current to gain the same low sea-level which attracts the Brazil water. Consequently, the ice-bearing currents from the south, which branch off from the great southern drift current, are able to largely turn away the warm Brazil current from the higher southern latitudes; and, furthermore, the great southern drift current which passes through the Cape Horn channel, and so onward around the globe, also partly turns away the Mozambique current as well as the East Australian current, and so largely prevents their waters from warming the southern seas.

Therefore, it is evident that, whenever the Cape Horn channel obtains sufficient capacity to give an independent circulation to the southern ocean, the conditions are favorable for the increase of cold in the southern latitudes. For it is because of the large exclusion of the tropical waters from the southern seas that ice-sheets have been able to form in early periods and in later epochs on the antarctic lands, and store away the annual frosts for thousands of years, and at the same time furnish icebergs sufficient to chill the waters of the southern temperate oceans, and consequently make cold such of the surface waters of the sea as are forced into the southern latitudes by the winds in surface currents, and so returned to warmer seas in cold under-currents, and thus with such frigid combinations bring about cold periods.

Thus it appears, as I have previously shown, that it is owing partly to there being more of the surface waters of the sea forced southward by the prevailing winds than they impel northward that the cold under-currents are maintained; but it also requires an independent circulation of the southern ocean, such as I have pointed out, to cool its surface waters before they can sink and form cold under-currents.

And there is reason to believe that such cold under-currents are more efficient in lowering the temperature of the temperate and tropical oceans than even the icebergs which such under-currents move into the temperate seas. And, when it is considered that the cold antarctic under-currents fill the depths of the Pacific and Indian Oceans in the northern hemisphere, and also largely the tropical depths of the North Atlantic, I am led to believe that the frigid conditions of the ice age were concurrent in the northern and southern hemispheres. The main reasons for such belief I will explain in the following chapter. After the foregoing explanations, showing how frigid periods are brought about through the independent circulation of the southern ocean surface waters, it is evident that, whenever through a slow natural process the Cape Horn channel is closed, a great change is wrought in the circulation of the southern ocean.

For instead of the westerly winds blowing the surface waters of the southern seas constantly around the globe, and so turning away and preventing the entrance of the tropical currents into the high southern latitudes, the strong westerly winds, whenever the Cape Horn channel is closed or greatly obstructed, would blow the surface waters away from the Atlantic side of the closed channel, and so cause a great low sea-level, sufficient to attract the ocean waters of the tropical high sea-level abreast Brazil well into the southern seas. Therefore, it is important to trace nature’s slow methods of closing the wide Cape Horn channel at the perfection of an ice age.

In my previous explanations on the subject I have thought that, should the southern seas have remained at or near the same sea-level as now, through an ice period brought about in the manner I have described, ice-sheets would accumulate on the antarctic continent, and also on the southern lands of South America, sufficient to flow out into the sea and close the Cape Horn channel.

But further consideration shows the impossibility of the southern seas having maintained their present sea-level during the growth of frigid epochs which have left such ample traces of glaciers having extended widely over the lands of the high latitudes of both the northern and southern hemispheres. For it appears that the larger areas of land in the northern latitudes, embracing wide continents and large islands, must, during the growth of a frigid age, have increased the spread of glaciers many times greater in extent than could be obtained on the smaller lands of the high latitudes of the southern hemisphere. For it is evident that the water evaporated from the sea and deposited in snow on the large continents and islands of the high northern latitudes during the growth of an ice period would, while thus diminishing the ocean waters, greatly increase the weight of northern lands. Therefore, the waters of the diminishing seas of the southern latitudes would be attracted into the northern oceans in opposition to the prevailing winds.

Thus it appears that the Cape Horn channel would be too much reduced at the perfection of an ice age to afford an independent circulation for the southern ocean, even without being filled by glaciers to the extent I have pointed out in previous essays. Still, to whatever dimensions the Cape Horn channel might be reduced at the perfection of a frigid period, the enlarged shores bordering its diminished waters would be covered by heavy glaciers that would flow into the shrunken strait, and so close it effectually. Thus the reduction of the Cape Horn channel during the advance of an ice age seems, on close consideration, to be a simple operation of nature, which in the normal course of events must have taken place.

As the closing of the Cape Horn channel has been considered by reviewers the weak and questionable point in preventing my views from gaining acceptance, it becomes necessary to be explicit concerning the manner in which the Cape Horn channel has in past ages been obstructed.

According to the charts prepared by John James Wild, the middle portion of the strait is represented as being over a thousand fathoms in depth; but, as far as I know, its true soundings have never been determined. The deep portion of the mid-channel is described as being narrow when compared with its whole breadth from Cape Horn to the antarctic continent.

And, when it is considered, with the growth of an ice age, how much of the ocean waters would be stored in the vast ice-sheets of the northern hemisphere, and consequently because of their weight a large portion of the diminished southern oceans would be attracted into the northern seas, it seems that the bottom of the shoaler waters of the Cape Horn channel, which now comprise so large a portion of its breadth, would be raised above the surface of the sea.

The one-hundred-fathom depth south of Cape Horn, now supposed to extend from longitude 70° west to 55° west, and southward to the latitude of 57°, would be a land supporting heavy glaciers for six hundred miles along the north side of the reduced channel during the advanced growth of a frigid age; and the same conditions would be obtained in the vicinity of the South Shetland. And when, in addition, we contemplate the great snow-fall of that region, and the consequent gathering of glaciers which would occur on the widened shores of the lessened channel, and the certainty of their flowing into the diminished strait, together with the immense icebergs of such an age grounding in the shoaled waters, it seems that the complete obstruction of the reduced channel would be accomplished.

While contemplating the conditions that would obtain while the Cape Horn channel was being reduced, it will be seen that the independent circulation of the icy southern ocean would be carried on to a considerable extent even after the narrowing strait was no longer able to afford space for wide drift currents, for the reason of the strong current that would be caused on account of the high ocean-level maintained by the westerly winds on the Pacific side of the diminishing channel, and the great low sea-level that would take place on its Atlantic side.

Still, as previously shown, it seems that during an advanced stage of the frigid epoch, the heavy glaciers from the enlarged northern and southern shores of the shrunken channel, together with the ponderous icebergs, blocking its waters, the closing process would at last be speedy and effective.

And on further consideration it might be said that a channel of much less width and depth would not have been of sufficient capacity to have caused ice periods so wide-spread as those that have left their traces on the continents and islands of the globe, for the reason that the independent circulation of the southern ocean would not have been sufficiently complete and long continued to have brought such world-wide cold periods to perfection.

With the Cape Horn channel closed, as above explained, there would be, as I have asserted, a great change wrought in the circulation of the southern ocean; for instead of the westerly winds blowing its surface waters constantly around the globe, and so turning away and preventing the entrance of tropical currents into the higher latitudes, the strong prevailing westerly winds would blow the surface waters of the sea from the Atlantic side of the closed Cape Horn channel, and so cause a great low sea-level, sufficient to attract the ocean waters of the tropical high sea-level abreast Brazil well into the southern seas.

The winds of the southern westerly wind-belt being stronger in that region than on any other portion of the globe, consequently they are able to do nearly as much work while drifting surface water as the belt of westerly wind of greater width on other parts of the southern seas. Thus a person who has had a long experience with the forcible westerly winds of the southern ocean can well understand their ability for disturbing the ocean waters in the latitudes of the Cape Horn channel.

The drift currents of this region are moved by the winds and waves from one to four miles an hour. Therefore, with the Cape Horn channel closed, there is nothing more certain than that the westerly winds would be able to cause a vast low sea-level on the Atlantic side of the closed Cape Horn strait, and that the waters of the high tropical sea-level abreast Brazil would be attracted to its wide depression, as shown on mapNo.1.

The tropical waters thus attracted far southward would be cooler than the tropical waters of to-day, owing to the great amount of cold imparted to the ocean by the numerous icebergs of a frigid age. Still, they would begin the slow process of raising the temperature of the southern ocean, and would in time carry sufficient heat into the southern regions to melt the ice from all southern lands; for, in addition to the Brazil currents, the waters of the high sea-level of the tropical Indian Ocean which pass southward down the Mozambique channel would reach a much higher latitude than during periods when the Cape Horn channel was open.

The ice periods of the northern and southern hemispheres being concurrent, a condition which I shall explain in another chapter, makes it obvious that during the melting of the glaciers from the antarctic continent and other southern lands the depleted Cape Horn channel could not gain sufficient capacity to give an independent circulation to the southern ocean during the melting of the southern ice-sheets, on account of the diminishing heaviness of the antarctic ice and the greater weight of the extensive glaciers and augmented seas of the northern latitudes. Consequently, it seems that the southern seas would continue in a lessened state while the glaciers were being melted from the northern hemisphere, as was the case during the melting of the ice from the southern hemisphere; and, furthermore, during such times the glaciers which overrun all the low lands and shoal waters of the Cape Horn region would, on account of their position being to the windward of the tropical currents, be the last great mass of ice to melt from the southern hemisphere.

Therefore, it seems that the Cape Horn channel would continue closed or greatly obstructed while the glaciers were being melted from the lands of both hemispheres. Thus at length a mild climate would extend over the globe, and so remain until the prevailing winds slowly forced the surface waters of the sea into the southern ocean in the manner explained in previous pages, thus filling the Cape Horn channel to its present capacity, and again restoring the independent circulation of the southern ocean. While contemplating the conditions that would obtain during the melting of the ice from the antarctic lands, it will be seen that the tropical waters attracted to the great low sea-level to the leeward of the closed Cape Horn channel would eventually enter the great bight of the antarctic continent to the eastward of Graham Land, where Captain Weddell sailed to the latitude of 74° south. This deep gulf, owing to its situation, would receive the full impact of the southern movement of the tropical currents; and, as the warm waters spread over the wide sea-level, the westerly winds would convert them into a drift current, and under such conditions would be driven along the shores of the antarctic continent, past the South Indian and Pacific Oceans, and eventually, after undergoing a cooling process from the long icy passage, be forced against the Pacific side of the closed Cape Horn channel and the western Patagonian coast.

While regarding the circulation of the sea during an ice age, it may be said that the ocean’s being composed of brine was the cause of its waters being able to circulate in frigid latitudes where fresh water would congeal. Consequently, this is one of the reasons why successive periods of frigidity and mildness have been brought about; for with an ocean of fresh water, repeated epochs of cold and warmth could not have occurred, because a sea composed of fresh water would have congealed while circulating in the high latitudes during a frigid age. Therefore, it required a sea of brine to maintain a liquid state during the low temperature of an ice period.

For, while the cold of a glacial age increased, the saltness of the sea increased also, because of the great amount of fresh water evaporated from the ocean, and stored in ice-sheets on the great continents and islands of the globe. Thus the briny sea was maintained in a liquid state, while washing vast ice-fields and glaciated shores and floating the numerous icebergs of a freezing age. The cold which radiated from such ice-bound seas must have been severe; but meanwhile the evaporation from the ocean was much reduced, while the saltness and coldness of the sea increased, and so prevented the ice of a glacial period from gaining invincible proportions before the independent circulation of the southern ocean was arrested. Therefore, the remaining warmth of the tropical waters after gaining free access to the antarctic latitudes was able to overcome the accumulated cold of that frigid region.

At this date the observant navigators who have visited the antarctic seas report that the surface currents above the latitude of Cape Horn, while being drifted eastward by the prevailing westerly winds, also set toward the antarctic ice cliffs, as shown on mapNo.2.

The reason why this southerly set of the surface currents becomes noticeable above the latitude of 55° south is because the tropical currents which set southward from the torrid latitudes on the western sides of the Atlantic, Indian, and Pacific Oceans, although largely turned away from the high latitudes by the westerly winds and drift currents, are also able to send sufficient water into the great belt of westerly winds to furnish water for the deep under-currents setting northward from the antarctic shores. Thus the surface waters moving from the north in order to gain the higher latitudes, after entering the westerly wind-belt, are moved in drift currents by the impelling winds easterly over many degrees of longitude, and also at the same time slowly southward among the cooling icebergs, because of the attraction caused by the difference of temperature and density between the northern drift waters and the icy seas of the antarctic ice barrier. Consequently, the gradual movement of the surface waters of the westerly wind-belt southward before entering the higher latitudes is not generally apparent; for it is after they enter latitudes where the globe becomes much reduced in circumference that their southern movement in the contracted seas becomes more noticeable. The impact of this southerly current, which finds its outlet in deep under-currents, and retards somewhat the increase of ice on the southern continent at this date, also largely prevents the small icebergs and field-ice from floating northward, away from the antarctic ice barrier; for it is such large icebergs as penetrate the deep under-currents that are the best able to move into the more temperate latitudes.

From the above explanations it will be seen that the impact of surface water against the antarctic ice barrier when the Cape Horn channel was closed would greatly assist the tropical waters attracted to the great low sea-level to the leeward of the obstructed strait to wash the antarctic shores while being drifted eastward by the westerly winds over the southern ocean against the Patagonian coast and the Pacific side of the closed channel, and there causing a high sea-level. This movement of the winds and currents encircling the antarctic continent is shown on mapNo.1.

The vast, high sea-level caused by the westerly winds drifting the surface waters against the Patagonian coast would obtain a much higher plain, were it not that so much of the water of the great drift current was required to feed the antarctic under-current which constantly sets northward from the antarctic shores; yet it would be sufficient to greatly increase the volume of the Humboldt current, which would flow in the same direction it now flows, down the South American coast to the equatorial latitudes, where it would become the main source of the great equatorial stream, and thus offset the increased southward flow of the equatorial waters through the Brazil and Mozambique streams.

The equatorial stream, with its increased volume, would also move, as it moves to-day, across the Pacific; and, on gaining the western side, after sending off large streams to the northern and southern latitudes, it would pass through the East India passages into the Indian Ocean, where it would be drifted westward by the trade winds and cause a high sea-level abreast the east coast of Africa, and so become the source of the great Mozambique current, which would flow southward along the east coast of Africa, and, with the Cape Horn channel closed, would gain a much higher latitude than it would with the channel open. At this age, when the continuation of this great equatorial stream gains the latitude of the Cape of Good Hope, its waters are largely turned eastward by the great drift current of the southern ocean.

Still, a considerable portion of its waters turns toward the west, forming the Agulhas current, which flows around the Cape of Good Hope into the Atlantic, where it mingles with the cooler currents which branch off from the great southern drift current; and so, in connection with the latter, it is attracted to the low sea-level caused by the south-east trade winds abreast the south-western coast of Africa, and from thence moved as a drift current by the trade winds to the equatorial Atlantic and coast of Brazil. Thus it will be seen that the Agulhas current, even with the Cape Horn channel in possession of its present wide capacity, serves to retard somewhat the advance of a cold period.

The Agulhas current at this date also partly serves to replenish the water which is forced from the South Atlantic by strong westerly winds into the Southern Indian and Southern Pacific Oceans. For it appears that more water is now removed by such winds from the South Atlantic than enters it from the South Pacific, even through the enlarged Cape Horn channel of this date; and this fact seems to favor an impression that a portion of this enlarged channel existed prior to the glacial periods, but with its waters so much reduced as to be unable to give the southern ocean an independent circulation sufficient to exclude the tropical currents from reaching the high southern latitudes in adequate volume to maintain a mild climate in the southern hemisphere.

For previous to the glacial age, with little or no ice gathered on the antarctic lands, it seems that a strait possessing one-half the capacity of the Cape Horn channel of the present age could not prevent the Brazil current and the Agulhas stream from flowing into the southern ocean in quantities sufficient to make it impossible for glaciers to form on southern lands.

Thus it is probable that a reduced channel separated the western continent from the antarctic lands even in the mild eras previous to the glacial epochs.

The Cape Horn channel, at the present age, with a capacity sufficient to largely maintain an independent circulation for the southern ocean, is still only one-third of the breadth of the westerly wind-belt of the southern seas. Therefore, the drift currents do not all pass through it from the Pacific into the Atlantic. Consequently, a considerable portion of the drifted water turns northward west of Cape Horn, and so forms the Humboldt current.

The Agulhas stream, which even now assists in replenishing the South Atlantic with tropical water, would, during the perfection of a glacial period, with the Cape channel closed, be a much stronger stream than it now obtains with the Cape channel possessing its present enlarged capacity, for the reason that the South Atlantic waters would continue as now to be forced eastward by the westerly winds, while they could not be replenished, as they are to-day, directly from the South Pacific.

Consequently, the waters of the South Atlantic Ocean would be correspondingly reduced.

Such conditions alone would greatly increase the volume of the Agulhas stream at the culmination of a frigid age. Therefore, the work of subduing a frigid period in the southern hemisphere after the Cape Horn channel was closed would not rest on the Brazil current alone, but also on the great equatorial stream of the Pacific and Indian Oceans.

Yet during such frigid times the sources of the equatorial stream would be greatly chilled by its two great feeders, the Humboldt current and the returning Japanese current, both of which flow down from the high latitudes and meet in the equatorial latitudes on the eastern side of the Pacific, thus cooling the source of the great equatorial current. But this latter stream, while on its long western passage across the Pacific and Indian Oceans, beneath a torrid sun, with only one cold feeder from the south which approaches it along the west side of Australia, would, on its long tropical journey, be able to obtain considerable warmth, even during an ice period, to supply the Mozambique and Agulhas streams, and so greatly assist the Atlantic waters in bringing about a mild period. Still, the process of subduing the cold of the southern latitudes would be slow, even with the Cape Horn channel closed, because of the vast collection of ice burdening the sea and land.

Yet there were conditions that were naturally brought about to favor the process of returning warmth; for it appears that, when the southern ocean was made shallow because of a considerable portion of its waters having been moved into the northern hemisphere, it will be seen that the conditions were more favorable for the westerly winds to create drift currents than would be the case on deeper seas. Therefore, the high and low sea-levels caused by such winds would be greater on a shallow ocean than would occur on deeper waters. Thus the low sea-levels of the shallow southern sea would have strong attraction for tropical surface waters, and so increase the thickness of its warm drift currents, and at the same time its lessened depths would have less capacity for the storage of cold water to reduce the temperature of the under-waters of the tropical zone.

And, furthermore, when the southern ocean was shallow, New Zealand acquired a longer extension of land to the north and south. Consequently, the enlarged low sea-level on its eastern side attracted more tropical water into the southern latitudes than now.

So, according to the conditions I have pointed out, the ice-sheets would at length melt away, and a long period of mildness would succeed on account of the length of time it would require after the ice disappeared from the earth for the prevailing winds to move the surface waters of the augmented northern seas into the southern ocean, and again restore its independent circulation, and so, after a considerable lapse of time, bring about the geographical and climatic conditions existing at the present date, which can be seen on mapNo.2, which shows that a cold period has already made considerable advance in the southern hemisphere, the southern continent and islands being covered with glaciers, and the prevalence of icebergs as far north as the latitude of 35° south.

Moreover, when we consider that the independent circulation of the southern ocean is caused by the westerly winds blowing its surface waters constantly around the globe through the open Cape Horn channel, and so largely preventing the tropical currents from entering the high southern latitudes, and how, in consequence, the cold is slowly on the increase through the constant accumulation of ice on the lands and in seas of the southern latitudes, it appears that a frigid age is slowly progressing in the southern hemisphere. For it seems that continental ice-sheets should not only be able to retain their freezing temperature, but also the mean of the low temperature in which they were formed, for a considerable length of time, and so impart their extreme coldness in the shape of icebergs into such seas as border on the glaciated lands.

It has been proved at Point Barrow that strata of ice and gravel can maintain a wintry temperature through the summer months. Captain G.B. Borden, keeper of the refuge station in that region, states that Lieutenant Ray, of the Signal Service, excavated through ice and gravel to a depth of forty-one feet, and that the lower portion of the excavation maintains a temperature 15° Fahrenheit above zero the year around. Therefore, with the probability of southern glaciers obtaining a temperature of over 15° Fahrenheit below the freezing point, we can well realize the frigidity imparted to the southern oceans while melting numerous immense icebergs, and consequently will conclude that the temperature of the southern latitudes is gradually lowering. The icebergs of the antarctic seas would not move northward into the temperature latitudes so readily as they now do, were it not that the general southward set of the southern ocean currents were interrupted by the movement of northerly surface currents in the longitudes of the low sea-levels, caused by the westerly winds drifting the surface waters of the sea from the eastern coasts of Southern South America and New Zealand. For it is owing to the low sea-levels thus created, in connection with the deep under-currents which set northward from the ice cliffs of the antarctic lands, that many icebergs are enabled to move into the temperate latitudes, especially to seas north-east of the Falkland Islands.

On other portions of the southern ocean above the latitude of 55° south the surface waters, while being drifted eastward by the strong westerly winds, also set toward the antarctic shores, and so furnish water for the cold under-currents which set northward from that frigid region. Thus from such parts of the coast only the largest bergs, which require a deep sea to float them, are moved by the under-currents into the temperate latitudes. Therefore, it happens that, while an ice period progresses, and the antarctic icebergs increase in size, the more readily the cold, deep under-currents force them into the temperate zone, in opposition to the winds and surface currents.

The icebergs, after gaining the temperate latitudes, are moved more or less eastward by the westerly winds and drift currents, and so are scattered over the southern temperate oceans, where the melting bergs impart whatever coldness they were able to store up while forming in the antarctic regions.

The low sea-levels caused by the westerly winds to the leeward of New Zealand and to the leeward of Argentine, not only cause the ice-bearing currents to set northward, but they also cause the tropical currents to make considerable inroads into the high southern latitudes. This is the reason why the lands are less burdened with ice on the antarctic shores opposite Cape Horn than on other parts of that glaciated continent. The tropical currents which turn southward east of New Zealand largely mingle their waters with the great southern drift current, and so are carried through the Cape Horn channel. Owing to this cause, the antarctic lands abreast Cape Horn are less burdened with ice than other portions of the antarctic shores.

Thus, were it not for this penetration of warm waters southward, the antarctic coasts south of Cape Horn, because of the great snow-fall of that region, would obtain heavier glaciers than other portions of the southern continent. But the time is slowly coming when, with a lower temperature, the ice-sheets on the lands in the vicinity of the South Shetlands will attain greater thickness than the glaciers on other shores of the antarctic continent.

Hence it appears that, when the several agents for producing and distributing cold in the southern latitudes are taken into consideration, the immense and continuous storage of ice on the southern lands, which adds to the wide-spread fleet of icebergs that float the southern temperate seas, and also the vast movement of cold antarctic water into the temperate and tropical oceans in deep under-currents, combined with the increasing coldness of the westerly winds, are now slowly bringing about in the southern hemisphere a period of frigidity.