THE SPREAD OF GLACIERS DURING COLD EPOCHS.  
I have before explained that the conditions are such that the cold periods of the northern and southern hemispheres were concurrent. Through this cause, while the glacial epoch was being perfected, the ice followed down the mountain ranges of both hemispheres; and, while gathering on the lands of the temperate latitudes, it also spread over a portion of the tropical zone. It is reported that traces of ancient glaciers are found in India, and also in Central America and in tropical South America. In fact, the denudation caused by ancient glaciers on the elevated lands of the tropics are too well defined to be attributed to any process of weathering, while Alpine plants of the same species are found near the summits of mountains in the tropics as well as in the high latitudes of both hemispheres.
This fact goes to show that a portion of the lowlands of the tropical zone have experienced a temperature favorable for the55 growth of Alpine plants. And, judging from the tropical islands I have visited, situated in the cold currents which flow down the eastern sides of the oceans from the high latitudes, I think they show strong traces of having during some remote period been subject to the action of glaciers. The island of St. Helena, situated in the southern tropical Atlantic, has the appearance of having been heavily iced during a frigid age. Its steep ravines, which deepen as they approach the sea, recall to the southern voyager the ice-worn islands of the high latitudes. It seems improbable that these deep ravines which penetrate the hard volcanic rock, on their short course to the sea, could have been caused by their scanty brooklets.
The bowlders scattered over the island are not in harmony with the weathering process, while the obliteration of its craters seems to point to a more rapid process of erosion than could be attributed to weathering.
Professor Agassiz, in his “General Sketch of the Expedition of the ‘Albatross,’” states that the Galapagos Islands are of volcanic origin, and that their age does not reach beyond the earliest Tertiary period; and his report seems to favor the impression of their having undergone denudation sufficient to slough off large portions of the rims of the older craters, and also the eastern face of Wenman Island. On Hood’s Island, at the time of my visit, its crater had entirely disappeared.
The highest portion of the island, which was the probable site of its ancient crater, showed no trace of its former existence; yet at the foot of this low mountain, on its southern side, I saw a large collection of loose bowlders, composed of hard volcanic rock, which were mostly free from soil and other débris, and easily moved from their places, while the spaces afforded by the loose piles of dark basaltic rocks afforded a secure retreat for numerous owls and lizards. Beyond the rocky piles to the southward a horizontal area of land was strewn with bowlders to the sea, which was some two miles distant from the higher land. The bowlders which covered the plain were somewhat56 smaller than those at the foot of the mountain, as none of the former were more than three or four feet in their longest measurement.
They seem to have been formed from thin strata of lava, which were broken in pieces from pressure, such as the action of ice could perform. In fact, the crowded and angular and somewhat worn blocks of lava presented a different appearance from stones thrown from the crater of a volcano, while no such bowlders are found among the recent volcanic eruptions on the islands.
The plain so thickly strewn with bowlders, and partly shaded by a tall growth of shrubs, fell off abruptly at the seaside, forming a steep cliff some two hundred feet in height.
The rocky floor at the foot of the cliff received such débris as fell from the sea-washed land; yet it contained few bowlders, they having been washed away by the waves soon after falling.
At one place a steep, dry ravine penetrated the land from the seashore, which was dangerous to cross on account of the loose stones resting on its sides. Two or three miles further west, on the level land bordering the sea, a large rookery of albatross were brooding their eggs and chicklings. The land on the south side of Albemarle, near the sea, consists of débris from the eroded high lands; and, judging from the crumbling cliffs by the sea, it seems that the land at one time extended further seaward.
Besides the excessive denudation which appears to have taken place on portions of these bowlder-strewn lands, we have other unmistakable testimony of their having formerly possessed a frigid temperature. The characteristic Alpine flora of these islands points to a time when they were exposed to a cold climate. Furthermore, rookeries of seal and albatross, which naturally belong to shores situated in cold latitudes, still exist on these equatorial islands; and, when we consider the favorable position of the Galapagos for the reception of cold during a frigid period, we can well account for the lingering signs which point to their former cold climate.
57 During the perfection of an ice period the western shore of South America was covered with an ice-sheet from the summits of its mountain range to the sea, extending northward as far as the latitude of 38° south.
This vast ice-sheet, situated in a region of great snow-fall, was constantly sending icebergs into the sea, where they were borne northward by the cold Humboldt current directly toward the Galapagos Islands; while, on the other hand, in the northern latitudes, in regions of great snow-fall, such as Alaska and British America, numerous icebergs were launched into the ocean, to be currented southward to the Galapagos seas. Thus during the frigid epoch the equatorial waters surrounding the Galapagos group was one of the greatest gathering places for floating ice to be found on the globe.
And here the frigidity stored up in the glaciers of the higher latitudes was set free, thus chilling the waters as well as the atmosphere of that region. The Alpine flora of the American coast mountains was probably carried by floating ice to the Galapagos, while its rookeries of albatross and seal date back to a cold period. And it seems that these cold-weather animals, with the assistance of the cool Humboldt current, may be able to preserve their rookeries at the equator until the advent of another ice period. In connection with the evidences of a cold climate having possessed the Galapagos, there are ample traces of ice-sheets having flowed over a large portion of the high lands of tropical America, and in some places the ice may have flowed down to the sea, especially where the large rivers now empty; and it is said that masses of clay, mixed with sub-angular stones, have been found in Brazil, which goes to prove the glaciation of portions of that tropical land during a remote age. Professor Louis J. R. Agassiz, during his research in the Amazon valley, found bowlders resting near the summits of the low hills of that region, which he attributed to the action of ice. The spread of glaciers on southern continents and islands is shown on map No. 1.
58 In Science, Nov. 17, 1893, Mr. J. Crawford published a summary of his discoveries in Nicaragua, during ten months of nearly continuous exploration since August, 1892.
The author of this report says: “The numerous eroded mountain ridges and lateral terminal moraines of that tropical region give unquestionable evidences of the former existence of a glacial epoch, which covered an area of several thousand square miles in Nicaragua with glacial ice. The ice-sheet covered a large part of the existing narrow divide of land (containing about 48,000 square miles) between the Pacific and Caribbean Sea.” And it is likely that other large areas of tropical America were glaciated at the same time, especially in regions of great precipitation.
The island of Cuba, during a portion of the ice age, probably supported heavy glaciers, and obtained an average temperature as low as South-western New Zealand at this age. According to the description given by J. W. Spencer, of the Cuban land, great valleys have been............