tlan

erosion.

mountainous form. Moreover, the crystalline texture and deformed foliation of the rocks prove that the surface now seen was once buried deep beneath the surface of an earlier time, for only at great Lauren- depths can such texture and foliation be acquired. Both these lines of evidence lead to the conclusion that the bighlands. moderate relief prevalent over the existing Laurentian region is the work of persevering erosion during a long continuance of dry land conditions, and hence that the region must be regarded as a worn-down mountain system. The worn-down old land is gently overlapped, chiefly around the south and west, and south of Hudson Bay, by very early Palaeozoic strata which rest upon the eroded surface of the crystallines, thus proving that the destruction of the ancient mountains had already been accomplished before some of the oldest fossiliferous formations of the world had been deposited. All the evidence goes to prove that from then to now the Laurentian region has been relatively quiescent. In all subsequent time there have been here only moderate oscillations of level, one of which allowed the transgression of the ancient sea in which the overlapping strata were deposited, while another of much more modern date gave the region its present highland altitude (1000 to 2000 ft.; mountains near the Labrador coast; 8000 ft.), again offering it to the forces of It is this ancient Laurentian area that the earlier geologists named the "Continental Nucleus," as if it had been the first part of North America to rise from the primeval waters of an assumed universal ocean. The "Archean V, formed by the two arms of the Laurentian oldland stretching from Labrador to the Arctic, between which Hudson Bay is included, has been repeatedly described as the oldest area of the continent, the beginning around which many later additions have built the existing outlines; and as such it has been adduced in favour of the theory of the permanence of continents. But when thus stated, the half of the story in favour of this theory is not told. Hudson Bay is not due to a primitive failure of elevation between the arms of the "Archean V"; it is not a deep basin whose floor has never emerged from the primeval ocean, but an ancient and comparatively shallow depression in a pre-existent land, over which the sea flowed as the surface sank below sea-level. South and west from the" Archean Nucleus," the Cambrian strata of the medial plains of North America are found to lie, wherever their base is discovered, on a foundation that possesses all the essential features of the Laurentian oldland. This relation is found all around the Adirondack mountains in New York, along the Appalachians southward to Georgia, through the Mississippi basin in Wisconsin and Missouri, and beyond in Texas, and farther west in the Black Hills, as well as certain points in the Rocky Mountains region. Hence the pre-Cambrian land surface of the continent must have had not only a vastly greater area than was formerly attributed to it, but also an earlier origin; for at the time when it was thought by the older geologists to be first rising from the primeval ocean, it is now proved to have been slowly sinking after a prolonged land existence. The crystalline Archaean rocks in the Laurentian region and its scattered fellows cannot possibly be explained as a primitive sea bottom, rising above sea-level to make the beginning of a continent and receiving Cambrian strata upon its still submerged borders, but only as portions of an already old and deeply-denuded land area, which was in pre-Cambrian time much larger than the visible Laurentian area of to-day, and which was reduced to perhaps half its primeval dimensions by a gradual submergence beneath the transgressing sea in which the Cambrian sediments were laid down. We are thus led to believe that much of the continent of to-day was a continent in the earliest geological times, and that the seas which partly covered it in Palaeozoic and Mesozoic_time were due to partial submergence, not to partial emergence. Furthermore, all the marine strata that now stretch over a large part of what is believed to have been the ancient continental surface are of relatively shallow water origin; none of them bears any close resemblance to the deposits of the deep oceans that have been so well studied of late years. Hence the Palaeozoic and Mesozoic seas of North America were not deep occans, and as far as this continent is concerned it is by no means admissible to assume, as some of the earlier geologists did, that the position of continents and occans have repeatedly changed places. The testimony of the rocks is decidedly in favour of Dana's view that continental masses are relatively permanent. The early history of the Laurentian region has been dwelt upon because of its great importance in the history of the continent, and because its history has so generally been misunderstood. To these reasons may be added a third: through Palacozoic and Mesozoic time the history of the Laurentian region is for the most part a blank. Records are wanting from the early Palaeozoic to the Pleistocene, when the Laurentian uplands became the centres from which the ice sheets of the Glacial period spread out on all sides. As a result of this late chapter in the history of the region, the weathered soils of earlier periods were swept away along with an unknown amount of firm rock, leaving bare ledges, scattered boulders and gravelly drift to-day upon a rugged upland without mountains (except in north-east Labrador), but diversified by innumerable knobs and hollows. The drainage of the region has thus been thrown into disorder; large and small lakes and marshy hollows abound; the streams are repeatedly interrupted by rapids, and frequently split into two or more channels, enclosing islands many miles

in length. They are the only highways of this thinly inhabited region. The Appalachian province is a generally hilly and mountainous belt, stretching from Newfoundland to Alabama. It seems for the most part to have belonged in the earliest times to the great pre-Cambrian land area, of which the Laurentian Appa highland is the more manifest representative; for where- lachlan ever the basal members of the Palaeozoic sedimentary highlands. series are found in the Appalachians, they rest upon a floor of denuded Archean rocks, and the lowest layers are largely composed of Archaean detritus. This province must, however, be set aside from the undisturbed Laurentian region because of the repeated movements of depression, deformation and elevation that it has suffered, generally along a north-cast south-west trend, causing the successive alternations of heavy deposition, and almost equally heavy denudation that have prevailed with varying intensity during the whole stretch of geological time covered by the fossiliferous record. The earliest important mountain-making disturbances interrupted the conditions of deposition in Cambrian time, and produced what has been called the Green Mountain system. A later, and probably greater, disturbance, with its climax at the close of Carboniferous time, established the Appalachian Mountain system; but, as understood to-day, the "Appalachian revolution" of the older geologists should be regarded as a long-lasting process, perhaps intermittently enduring as long as the whole of Carboniferous time. A subordinate period of deposition and deformation occurred early in Mesozoic time, marked by the accumulation and disturbance of several basins of the Newark formation, roughly corresponding to the Triassic of Europe. The Appalachian mountains of to-day were formerly regarded as the unconsumed remnants of the chief Appalachian uplift; but it is now generally agreed that Mesozoic erosion reduced the greater part of the range to a lowland of moderate or small relief, leaving only isolated groups of subdued mountains in the areas of the most resistant rocks, and that the altitude and form of the mountains of to-day are chiefly the result of the Tertiary elevation and dissection of the previously worn-down mass-the additional height thus given in Tertiary time to the pre-existent subdued mountain groups making them now the loftiest areas of the range, as in the White Mountains of New Hampshire (Mount Washington, 6293 ft.) and the Black Mountains of North Carolina (Mount Mitchell, 6711 ft.). It is interesting to note that the axis of Tertiary elevation is nearly parallel to and closely associated with the axes of the carlier disturbances, but it lies somewhat to the north-west of its predecessors, and therefore involves considerable areas of flat-lying Palaeozoic strata on the inner side of the previously disturbed belt from New York to Alabama, thus producing what is known as the Alleghany plateau (altitudes, 2000 to 4000 ft.). It should be added that the Ozark plateau of Missouri and the Ouachita mountains on the south, in Arkansas and farther west, are related to one another in much the same way as the Alleghany plateau and the middle ranges of the Appalachians-the two pairs corresponding to a remarkable degree in regard to conditions of ancient accumulation, medieval deformation and denudation, and more modern uplift and dissection; it is, therefore, admissible to classify this western group of uplifts as an annex to the normal Appalachians. Numerous and extensive coal seams occur in the worn-down Appalachians of Nova Scotia, Pennsylvania and Alabama, as well as in the Alleghany plateau from Pennsylvania to Alabama, and in the extension of the same strata through the Ohio and middle Mississippi basins. The eastern coast of the continent has a rocky and ragged shore line from Maine to Greenland, with numerous submerged lowlands and valleys forming bays, and as many uplands and ridges outstretching in promontories and islands; this being the result of the summation of many movements of the land, whose total gives an increasing measure of depression to the north, where an archipelago at last replaces what was probably once a corner of the continent; but the measure of the depression is uncertain, because of the doubt regarding the depth beneath sea-level to which the Pleistocene glaciers may have worn the pre-Glacial valleys. South of New England, along the Atlantic coast, and around the border of the gulf into Mexico, the dominating movement of the land in late geological periods has been upward with respect to sea-level, whereby a former sea bottom, on which the land waste of Cretaceous and Tertiary times had been outspread, was revealed as a coastal plain, across which the rivers of the former land area now extend their courses, from the old shore line to the new. Part of the same plain, still submerged, forms the "continental shelf" of the mid-Atlantic border. Florida seems to be a projecting swell of this shelf, around whose extremity coral reefs have been added, but whose greater mass is still under a shallow sea cover. Along the ragged coast in the north a moderate and very modern movement of elevation has laid bare clay-floored lowlands that were lately beneath the sea, as in the plain of the lower St Lawrence valley, while along the coastal plain of the south a slight movement of depression has drowned a number of low valley floors, producing shallow arms of the sea, as Chesapeake Bay, Albemarle and Pamlico Sound and Mobile Bay. All the coast south of New York is low, and a great part of it is fringed with wave-built sand-reefs.

The great complex of mountains in the Western highlands.

The Cor

North

America.

sometimes styled the Cordilleras of North America (the Rocky | Mountains being the castern members of the system in the United States and Canada), differ from the Laurentian and Appalachian regions in having suffered numerous dilleras of disorderly movements at dates so recent that the existing relief of the region bears a significant relation to its irregular uplifts; a relation that doubtless once obtained in the older mountain areas of the east, where it has now been obliterated by erosion. It is not, however, only in modern geological periods that mountain-making disturbances have prevailed in the regions of the Western highlands; their geological history is one of repeated and long-continued movement-the ruins of the more ancient upheavals supplying materials for the strata of newer ranges. For example, in Canada an axial belt of ancient rocks is bordered on the east and west by stratified formations of enormous thickness (40,000 to 60,000 ft.), those on the west including a large share of contemporaneous volcanic materials; all three belts having been deformed and upheaved, as well as deeply dissected in the later chapters of geological time. It is, however, important to note that the interval between Palaeozoic and Mesozoic time, in which mountain-making disturb. ances were so general in western Europe and eastern North America that the older geologists thought them to be of world-wide extent, was here generally passed over in relative quiet, so that continuous sedimentation produced in certain districts a conformable series of deposits from Silurian to Cretaceous time. Furthermore, the Carboniferous period, which gained its name from the extensive coal deposits that were then formed in western Europe and eastern North America, was a marine limestone-making period in the Cordilleran region. "There is here exemplified, as might be expected in a region extending over 3000 m. from Alaska to southern Mexico, and measuring over 1000 m. in breadth at its middle, a great variety of plateau and mountain structures. The broad upheaval of adjacent blocks of earth-crust without significant tilting or disturbance has produced the plateaus of Arizona and Utah. Some of the simplest and youngest mountain ridges in the world are to be found in the broken and tilted lava blocks of southern Oregon. Tilted blocks on a larger scale, much more affected by processes of sculpture, are found in the lofty St Elias Alps of Alaska, the site of some of the greatest glaciers in the world. The wall of a huge fracture, now elaborately carved, constitutes the western slope of the Wasatch range, facing the desert basin of Utah. Ranges of a relatively simple structure are seen in the Uinta mountains of Wyoming and Utah. Arched upheavals also characterize the front range of the Rocky Mountains proper in Colorado and Wyorning and in the Black Hills of South Dakota, bending up the strata of the adjacent plains in the simplest fashion, and producing dome-like mountains, now deeply dissected by outflowing consequent streams. A remarkable change in the structure of the Rocky Mountains occurs north of the Missouri river in Montana and northward into Canada, where the front range is of synclinal or trough structure, with the youngest instead of the oldest rocks along the axis, while the strata of the plains are bent down and overridden in the most abnormal manner. Indeed, mountain structure occurs of so great diversity in various parts of the Cordilleran region as to elude general description. The disturbances extend directly to the western coast line, including not only the coast range of Calífornia, but the peninsular area of Lower California (belonging to Mexico) and the detached mountainous islands of British Columbia and Alaska.

McKinley, 20,300 ft.; Mount Logan, 19,540 ft.; Mount St Elias, 18,000 ft.); and the great Mexican volcanoes rise nearly as high (Orizaba, 18,250 ft.). Widespread plateaus maintain upland altitudes of more than a mile over vast areas.

As in all regions of great altitude, the erosion of valleys has progressed on a magnificent scale in the Cordilleran region, and the actual form of many of its parts is more the result of sculpturing than of uplifting. The plateaus of Arizona are traversed by the deep cañons of the Colorado river and its branches, at places i m. deep, and with elaborately carved walls. Upon the plateaus themselves, long and ragged cliffs of recession attest an even greater work of erosion than the cañons. In all the mountain ranges except those of youngest uplift, valleys have been actively eroded, sometimes producing steep peaks as in Mount Assiniboine (11,500 ft.) in the Canadian Rockies, rivalling the Swiss Matterhorn in sharpness of form; but the greater number of summits have been worn to roughly pyramidal form between wide-flaring valleys, and the mountain flanks have thus come to be extensively covered with rock waste lying on slopes of relatively uniform declivity. Some of the ranges are in a second cycle of dissection, having been once worn down to moderate relief and now being elevated for renewed erosion; the Sierra Nevada of California is believed to be, in part, of this history. having at least in its central and northern parts been well reduced and now again enjoying a mountainous character in virtue of a later slanting uplift en bloc, with rapid descent on its eastern fractured face. Other ranges, almost completely worn down, still remain low, as in south-eastern California, where they are now represented by gently sloping rock floors veneered with gravel and retaining only small remnants of their original mass still unconsumed; thus the end, as well as the beginning, of the cycle of erosion, together with many complications of its progress, are illustrated in different parts of this great and varied mountain system. In the fjorded coast of Alaska, as well as in the higher northern ranges, signs of intense glacial erosion are seen in the cirques at the valley heads and in the discordant junction of the "hanging" lateral valleys and the deep trunk valleys-the floors of the former being cut off on the walls of the latter.

Fitting complements of the deeply-eroded mountains are found in the great accumulations of mountain waste now occupying basins of depression between the various ranges, as in Mexico, Utah, Nevada, Montana and elsewhere. Erosion and transportation here combine to build up the floors of the basins with the waste of the surrounding highlands; a result that is peculiarly beneficial in Mexico where the climate of the plateau basins is rendered relatively temperate by reason of its altitude, and where the surface is easily habitable by reason of its small relief. In the larger depressions, as along the boundary of the United States and Mexico, isolated ranges frequently rise like islands over the plain of waste that has been built up on their flanks. Shallow saline lakes or playas (wet-weather lakes) without outlets lie on the lowest parts of the waste-filled basins; their failure to overflow in rivers discharging to the sea being less the result of enclosure by barriers than of deficiency of rainfall; for it is chiefly in the arid region that the waste-floored basins are best developed. Indeed, the rainfall is often so scanty that the streams from the mountains-where most of the little precipitation occurs-often fail even to form lakes, withering away on the waste plains. In all these cases, the wash of rock waste from the mountains remains on the continent and builds up the basin plains, instead of being carried away from the land to form stratified sediments on the sea floor. The habit of gathering mountain waste interior basins that characterizes so much of the Cordilleran region to-day is only the continuation of an earlier practice, for extensive basin deposits of Tertiary date are found in many parts of the Cordilleran region; some of them are famous for preserving vertebrate fossils, such as those of the many-tocd ancestors of the horse.

Plains.

Between the loftier western highlands and the lower castern highlands (Laurentian and Appalachian) lies a great extension of medial plains, stretching in moderate altitude from the Arctic Ocean to the Gulf of Mexico, and having in their The Medial middle a breadth of 1500 m. They are composed throughout of nearly horizontal strata and mark a region long exempt from strong disturbance. Although for the most part floored by marine formations, their structure and composition indicate, as has already been said, relatively shallow water. The ancient sea that once occupied the middle belt of the continent therefore had little likeness to the abysmal oceans, but resembled rather the shallow ocean margins that to-day overlap various continental massesthe largest example of this kind now existing being between Asia and Australia. The eastern part of the plains is underlaid by Palaeozoic strata, already mentioned as having been laid down upon the sub siding Archaean continent or folded in the making of the Appalachians; coal beds are here included in the Ohio and middle Missis sippi basins. The area of the western plains remained submerged to a later date, preserving a stretch of marine waters to the end of Mesozoic time, and thus resembling the lowland belt of western Asia, which was similarly covered by a broad and a shallow arm of the ocean extending from the Arctic to the European mediter raneans until a late geological date. The surface of the medial plains is not always so even as might be inferred from their name. Both

the eastern and the western areas have been extensively denuded, even to the point of being reduced to peneplains. Their present altitude is not so much the result of their original uplift from the sea as of a later elevatory movement. The great river basins, for which North America is famous, have thus been formed between the eastern and western highlands-the Mississippi receiving the drainage of a vast area (about 1,240,000 sq. m.) for discharge to the south, while the Saskatchewan and Mackenzie gather their waters from somewhat less extensive areas in the north. Pleistocene glaciation covered the plains of the Ohio, upper Mississippi and Winnipeg districts with extensive deposits of ice-laid or water-laid drift, furnishing a generally smooth surface and a fertile soil: here are the true prairies-treeless, but richly grassed.

The traditional continuity of the Cordilleras of North and South America has been broken by investigations in the isthmian portion of the northern continent. The structural peculiaritics Central of the western highlands of North America may be traced America only to the east and west belt of great volcanoes by and the which the plateau of central Mexico is terminated on the West south. The ranges of the Andes fail to reach Panama, Indies. from which the nearest one is separated by the valley of the Atrato. The two Cordilleras are out of line with each other, and their ends are some 1200 m. apart. Central America, the West Indies and various submarine ridges by which the islands are connected with one another and with the mainland to the west, as well as certain ranges along the northern margin of South America, all belong together in what has been termed the Antillean mountain system, in which cast and west trends of late geological date predominate, with abundant volcanic additions on the Pacific border of Central America, and along the eastern end of the system in the Windward islands of the Lesser Antilles. The unity of this system has been until recently overlooked partly because the Antillean ranges are for the most part still under water, and yet further because the volcanoes which form the strongest reliefs of the isthmian region are so arranged along the Pacific coast as to suggest the continuity of the Cordilleran systems on the north and south; but these volcanoes are really only superadded to a foundation of quite another kind. Geological studies on the mainland and on the islands have shown that both fundamental structure and surface form are not Cordilleran; and numerous soundings in the adjacent mediterraneans suggest that the islands are best interpreted as the somewhat denuded crests of great crustal ridges. The warm waters that bathe the West Indies come with a high temperature from the equatorial Atlantic, and favour the growth of corals along the shores. Fringing and elevated reefs are known on many of the islands. The Bahamas are the slightly overtopping parts of a broad platform of coral and other calcareous marine deposits, of which the greater area constitutes extensive shallow banks, which descend by a steep slope on the north-east to great depths in the Atlantic. The lowlands of Yucatan resemble Florida in being the emerged part of a much larger mass, of which an equal portion is still under water in the shelf around the Gulf of Mexico. All this region is luxuriantly productive and is advantageously surrounded by waters which would be barren and desert, like the Sahara, if replaced by lowlands. The active volcanoes on the Pacific slope have built many cones and uplands, some of their historic eruptions having been of terrible violence. Thus Lake Nicaragua, once a bay of the Pacific, has been cut off by volcanic deposits, leaving only the Gulf of Fonseca open to the western ocean, raising the level of the lake behind the barrier and turning its discharge eastward to the Caribbean Sea across what was once the inter-oceanic watershed.

The successive crustal movements by which the land area of what we now know as North America has been increased and connected have determined the growth of several great river systems

Rivers. through which the broader part of the continent is drained. The movements that resulted in the emergence of the Plains had the effect of engrafting many ancient rivers of moderate size upon trunks of unusual dimensions. The Mississippi system, some of whose eastern branches probably date from early Mesozoic time, received great reinforcements by the addition of many long western branches in Tertiary time, roughly contemporaneous with the uplift of the Gulf coastal plain by which the lower trunk of the river was extended to the sea. The present headwaters of that river-trunk to which the name of Mississippi is applied, and which for that reason have gained an undue subjective importance, are of relatively modern date, as they are controlled by the abundant glacial deposits of northern Minnesota. The evolution of the Mackenzie resembles that of the Mississippi in a very general way, although some of its eastern branches may be the descendants of ancestors more ancient than those flowing westward from the Appalachians; but the regime of the great northern river is strikingly unlike that of its still greater southern analogue on account of its course being from a warmer to a colder climate: hence ice-úams, obstructed discharge, and overflows recur every spring. The Nelson and the St Lawrence systems, draining eastward to Hudson Bay and St Lawrence Gulf, receive drainage from areas that would belong to the Mackenzie and the Mississippi systems under a simpler plan of continental growth; and there is much reason for thinking that this simpler plan obtained until the occurrence of those changes, in association with the Glacial period, whereby sea waters gained access to the depressions that now

hold the bays and sounds of the north-eastern coast. In exemplification of the rule that the larger ocean receives the drainage of the smaller continental area, the rivers that flow into the Pacific rank below those belonging to the Atlantic. The greatest is the Yukon, of farther Canada and inner Alaska, one of the great rivers of the world, little known until the active exploration of its basin for goldfields. The Frazer drains much of the mountainous area of southern British Columbia, as the Columbia drains that of the north-western United States; the latter is peculiar in that one of its headwaters rises at the eastern base of the Rocky Mountains in northern Montana and flows westward through the ranges. The Colorado discharges a muddy current into the Gulf of California; but for the aridity of its large drainage area its volume would be much larger. The same is true of the Rio Grande, whose name would be better justified if so much of its basin were not semi-arid.

The most remarkable lacustrine region of the continent, rivalling that of Central Africa, forms a belt around the border of the Laurentian highland; here, in addition to ten large lakes, there Lakes. are hundreds of medium size, and many thousand small lakes. They are peculiar in occupying a region of moderate relief, in which no strong dislocations have taken place in recent geological time (unless in the case of Lake Superior), and thus in contrasting with the great African lakes which occupy rift-valleys or graben of comparatively recent fracture. The Laurentian lakes are further characterized by an intimate association with the ice-sheets of the Glacial period; but while glacial erosion and drift obstruction suffice to account for the smaller lakes, it is very probable that broad crustal warping and drainage reversal have been potent aids to the other processes in producing the great lakes. The northern Cordilleran region contains many beautiful lakes of moderate size in deep valleys among the crowded ranges of the narrowed mountain belt. Their origin has not been closely studied. The basins among the spaced ranges of the middle and southern Cordilleras, in the United States and Mexico, contain many lakes that occupy shallow depressions in desert plains; they are usually without outlet and saline; many of the basins were formerly occupied by lakes of much greater size, some of which overflowed, implying a climate moister than that of to-day, probably correlated with the glacial climate of the regions farther north. Lakes in volcanic craters or behind volcanic barriers occur in Central America, while Florida possesses many small lakes in limestone basins. The following table is taken from Russell's Lakes of North America:

The climatic features of North America are best appreciated when considered as exhibiting modifications of those general climatic conditions which prevail in consequence of the globular Climate. form of the earth as a whole. In January, when the isotherms of 65° to 75° F. stretch almost directly across land and sea in the north torrid zone, a mean temperature of zero or less invades the region north-west of Hudson Bay, which thus resembles north-eastern Asia in departing greatly from the mean prevailing in similar latitudes on the northern oceans, and in bringing upon the northern lands an extension of frigid conditions that have no analogue in the southern or oceanic hemisphere. In July, when the isotherms of 40° and 50° have a tolerably direct course around the latitude circles that border the continent on the north, a great middle area of North America becomes warmer than the seas on the cast and west, having a mean of over 80°, and in part over 90°. In January the Hudson Bay region is 30° colder than the mean of its own latitude, about 60° colder than the mean of the corresponding southern latitude; while in July the Arizona-Mexican region is 203 above the mean of its own latitude, or about 40° above the mean of the corresponding southern latitude. In both winter and summer the isotherms are more closely crowded while crossing the continent than while crossing the adjacent oceans; or, in other words, the poleward temperature gradient is stronger on the land than on the oceans; and all these features should be regarded as inherent characteristics of the climate of North America in virtue of its being a continent chiefly in temperate latitudes. An associated feature of continental climate is found in the strong annual range of temperature of the central land arca. The range between the means of January and July exceeds 40° for the largest part of the lands, and 70° for much of the northern lands; the range of extreme temperatures is much greater. On corresponding oceanic areas in the northern hemisphere the range is little more than 20°, and in the southern hemisphere it is probably less than 10°. It must appear from this that if the largest part of North America is said to be in the north temperate zone, temperate must be taken as having little of the meaning originally given to it in southern Europe, for the winter cold is severe and the summer heat is excessive over much of the North American continent.

The several members of the terrestrial wind system, including