GLACIAL PERIOD

glacial period, in geology, the name usually given, by english and american writers, to that comparatively recent time when all parts of the world suffered a marked lowering of temperature, accompanied in northern europe and north america by glacial conditions, not unlike those which now characterize the polar regions. this period, which is also known as the "great ice age" (german _die eiszeit_), is synchronous with the pleistocene period, the earlier of the post-tertiary or quaternary divisions of geological time. although "glacial period" and "pleistocene" (q.v.) are often used synonymously it is convenient to consider them separately, inasmuch as not a few pleistocene formations have no causal relationship with conditions of glaciation. not until the beginning of the 19th century did the deposits now generally recognized as the result of ice action receive serious attention; the tendency was to regard such superficial and irregular material as mere rubbish. early ideas upon the subject usually assigned floods as the formative agency, and this view is still not without its supporters (see sir h. h. howorth, _the glacial nightmare and the flood_). doubtless this attitude was in part due to the comparative rarity of glaciers and ice-fields where the work of ice could be directly observed. it was natural therefore that the first scientific references to glacial action should have been stimulated by the alpine regions of switzerland, which called forth the writings of j. j. scheuchzer, b. f. kuhn, h. b. de saussure, f. g. hugi, and particularly those of j. venetz, j. g. von charpentier and l. aggasiz. canon rendu, j. forbes and others had studied the cause of motion of glaciers, while keen observers, notably sir james hall, a. brongniart and j. playfair, had noted the occurrence of travelled and scratched stones. the result of these efforts was the conception of great ice-sheets flowing over the land, grinding the rock surfaces and transporting rock debris in the manner to be observed in the existing glaciers. however, before this view had become established sir c. lyell evolved the "drift theory" to explain the widely spread phenomenon of transported blocks, boulder clay and the allied deposits; in this he was supported by sir h. de la beche, charles darwin, sir r. i. murchison and many others. according to the drift theory, the transport and distribution of "erratic blocks," &c., had been effected by floating icebergs; this view naturally involved a considerable and widespread submergence of the land, an assumption which appeared to receive support from the occasional presence of marine shells at high levels in the "drift" deposits. so great was the influence of those who favoured the drift theory that even to-day it cannot be said to have lost complete hold; we still speak of "drift" deposits in england and america, and the belief in one or more great submergences during the glacial period is still held more firmly by certain geologists than the evidence would seem to warrant. the case against the drift theory was most clearly expressed by sir a. c. ramsay for england and scotland, and by the swedish scientist otto torell. since then the labours of professor james geikie, sir archibald geikie, professor p. kendall and others in england; von verendt, h. credner, de geer, e. geinitz, a. helland, jentzsch, k. keilhack, a. penck, h. schroder, f. wahnschaffe in scandinavia and germany; t. c. chamberlin, w. upham, g. f. wright in north america, have all tended to confirm the view that it is to the movement of glaciers and ice-sheets that we must look as the predominant agent of transport and abrasion in this period. the three stages through which our knowledge of glacial work has advanced may thus be summarized: (1) the diluvial hypothesis, deposits formed by floods; (2) the drift hypothesis, deposits formed mainly by icebergs and floating ice; (3) the ice-sheet hypothesis, deposits formed directly or indirectly through the agency of flowing ice. _evidences._--the evidence relied upon by geologists for the former existence of the great ice-sheets which traversed the northern regions of europe and america is mainly of two kinds: (1) the peculiar erosion of the older rocks by ice and ice-borne stones, and (2) the nature and disposition of ice-borne rock debris. after having established the criteria by which the work of moving ice is to be recognized in regions of active glaciation, the task of identifying the results of earlier glaciation elsewhere has been carried on with unabated energy. [illustration: glacial period.] 1. _ice erosion._--although there are certain points of difference between the work of glaciers and broad ice-sheets, the former being more or less restricted laterally by the valleys in which they flow, the general results of their passage over the rocky floor are essentially similar. smooth rounded outlines are imparted to the rocks, markedly contrasting with the pinnacled and irregular surfaces produced by ordinary weathering; where these rounded surfaces have been formed on a minor scale the well-known features of _roches moutonnees_ (german _rundhocker_) are created; on a larger scale we have the erosion-form known as "crag and tail," when the ice-sheet has overridden ground with more pronounced contours, the side of the hill facing the advancing ice being rounded and gently curved (german _stossseite_), and the opposite side (_leeseite_) steep, abrupt and much less smooth. such features are never associated with the erosion of water. the rounding of rock surfaces is regularly accompanied by grooving and striation (german _schrammen, schliffe_) caused by the grinding action of stones and boulders embedded in the moving ice. these "glacial striae" are of great value in determining the latest path of the vanished ice-sheets (see map). several other erosion-features are generally associated with ice action; such are the circular-headed valleys, "cirques" or "corries" (german _zirkus_) of mountain districts; the pot-holes, giants' kettles (_strudellocher_, _riesentopfe_), familiarly exemplified in the gletschergarten near lucerne; the "rock-basins" (_felsseebecken_) of mountainous regions are also believed to be assignable to this cause on account of their frequent association with other glacial phenomena, but it is more than probable that the action of running water (waterfalls, &c.)--influenced no doubt by the disposition of the ice--has had much to do with these forms of erosion. as regards rock-basins, geologists are still divided in opinion: sir a. c. ramsay, j. geikie, tyndall, helland, h. hess, a. penck, and others have expressed themselves in favour of a glacial origin; while a. heim, f. stapff, t. kjerulf, l. rutimeyer and many others have strongly opposed this view. 2. glacial deposits may be roughly classified in two groups: those that have been formed directly by the action of the ice, and those formed through the agency of water flowing under, upon, and from the ice-sheets, or in streams and lakes modified by the presence of the ice. to differentiate in practice between the results of these two agencies is a matter of some difficulty in the case of unstratified deposits; but the boulder clay may be taken as the typical formation of the glacier or ice-sheet, whether it has been left as a _terminal moraine_ at the limit of glaciation or as a _ground moraine_ beneath the ice. a stratified form of boulder clay, which not infrequently rests upon, and is therefore younger than, the more typical variety, is usually regarded as a deposit formed by water from the material (_englacial_, _innenmoran_) held in suspension within the ice, and set free during the process of melting. besides the innumerable boulders, large and small, embedded in the boulder clay, isolated masses of rock, often of enormous size, have been borne by ice-sheets far from their original home and stranded when the ice melted. these "erratic blocks," "perched blocks" (german _findlinge_) are familiar objects in the alpine glacier districts, where they have frequently received individual names, but they are just as easily recognized in regions from which the glaciers that brought them there have long since been banished. not only did the ice transport blocks of hard rock, granite and the like, but huge masses of stratified rock were torn from their bed by the same agency; the masses of chalk in the cliffs near cromer are well known; near berlin, at firkenwald, there is a transported mass of chalk estimated to be at least 2,000,000 cubic metres in bulk, which has travelled probably 15 kilometres from its original site; a block of lincolnshire oolite is recorded by c. fox-strangways near melton in leicestershire, which is 300 yds. long and 100 yds. broad if no more; and instances of a similar kind might be multiplied. when we turn to the "fluvio-glacial" deposits we find a bewildering variety of stratified and partially bedded deposits of gravel, sand and clay, occurring separately or in every conceivable condition of association. some of these deposits have received distinctive names; such are the "kames" of scotland, which are represented in ireland by "eskers," and in scandinavia by "asar." another type of hillocky deposit is exemplified by the "drums" or "drumlins." everywhere beyond the margin of the advancing or retreating ice-sheets these deposits were being formed; streams bore away coarse and fine materials and spread them out upon alluvial plains or upon the floors of innumerable lakes, many of which were directly caused by the damming of the ordinary water-courses by the ice. as the level of such lakes was changed new beach-lines were produced, such as are still evident in the great lake region of north america, in the parallel roads of glen roy, and the "strandlinien" of many parts of northern europe. viewed in relation to man's position on the earth, no geological changes have had a more profound importance than those of the glacial period. the whole of the glaciated region bears evidence of remarkable modification of topographic features; in parts of scotland or norway or canada the old rocks are bared of soil, rounded and smoothed as far as the eye can see. the old soil and subsoil, the product of ages of ordinary weathering, were removed from vast areas to be deposited and concentrated in others. old valleys were filled--often to a great depth, 300-400 ft.; rivers were diverted from their old courses, never to return; lakes of vast size were caused by the damming of old outlets (lake lahontan, lake agassiz, &c., in north america), while an infinite number of shifting lakelets--with their deposits--played an important part along the ice-front at all stages of its career. the influence of this period upon the present distribution of plant and animal life in northern latitudes can hardly be overestimated. much stress has been laid upon supposed great changes in the level of the land in northern regions during the glacial period. the occurrence of marine shells at an elevation of 1350 ft. at moel tryfaen in north wales, and at 1200 ft. near macclesfield in cheshire, has been cited as evidence of profound submergence by some geologists, though others see in these and similar occurrences only the transporting action of ice-sheets that have traversed the floor of the adjoining seas. marine shells in stratified materials have been found on the coast of scotland at 100 ft. and over, in s. scandinavia at 600 to 800 ft., and in the "champlain" deposits of north america at various heights. the dead shells of the "yoldia clay" cover wide areas at the bottom of the north atlantic at depths from 500 to 1300 fathoms, though the same mollusc is now found living in arctic seas at the depth of 5 to 15 fathoms. this has been looked upon as a proof that in the n.w. european region the lithosphere stood about 2600 ft. higher than it does now (brogger, nansen, &c.), and it has been suggested that a union of the mainland of europe with that of north america--forming a northern continental mass, "prosarctis"--may have been achieved by way of iceland, jan mayen land and greenland. the pre-glacial valleys and fjords of norway and scotland, with their deeply submerged seaward ends, are regarded as proofs of former elevation. the great depth of alluvium in some places (236 metres at bremen) points in the same direction. evidences of changes of level occur in early, middle and late pleistocene formations, and the nature of the evidence is such that it is on the whole safer to assume the existence only of the more moderate degree of change. _the cause of the glacial period._--many attempts have been made to formulate a satisfactory hypothesis that shall conform with the known facts and explain the great change in climatic conditions which set in towards the close of the tertiary era, and culminated during the glacial period. some of the more prominent hypotheses may be mentioned, but space will not permit of a detailed analysis of theories, most of which rest upon somewhat unsubstantial ground. the principal facts to be taken into consideration are (1) the great lowering of temperature over the whole earth; (2) the localization of extreme glaciation in north-west europe and north-east america; and (3) the local retrogression of the ice-sheets, once or more times repeated. some have suggested the simple solution of a change in the earth's axis, and have indicated that the pole may have travelled through some 15 deg. to 20 deg. of latitude; thus, the polar glaciation, as it now exists, might have been in this way transferred to include north-west europe and north america; but modern views on the rigidity of the earth's body, together with the lack of any evidence of the correlative movement of climatic zones in other parts of the world, render this hypothesis quite untenable. on similar grounds a change in the earth's centre of gravity is unthinkable. theories based upon the variations in the obliquity of the ecliptic or eccentricity of the earth's orbit, or on the passage of the solar system through cold regions of space, or upon the known variations in the heat emitted by the sun, are all insecure and unsatisfactory. the hypothesis elaborated by james croll (_phil. mag._, 1864, 28, p. 121; _climate and time_, 1875; and _discussion on climate and cosmology_, 1889) was founded upon the assumption that with the earth's eccentricity at its maximum and winter in the north at aphelion, there would be a tendency in northern latitudes for the accumulation of snow and ice, which would be accentuated indirectly by the formation of fogs and a modification of the trade winds. the shifting of the thermal equator, and with it the direction of the trade winds, would divert some of the warm ocean currents from the cold regions, and this effect was greatly enhanced, he considered, by the configuration of the atlantic ocean. croll's hypothesis was supported by sir r. ball (_the cause of the great ice age_, 1893), and it met with very general acceptance; but it has been destructively criticized by professor s. newcomb (_phil. mag._, 1876, 1883, 1884) and by e. p. culverwell (_phil. mag._, 1894, p. 541, and _geol. mag._, 1895, pp. 3 and 55). the difficulties in the way of croll's theory are: (1) the fundamental assumption, that midwinter and midsummer temperatures are directly proportional to the sun's heat at those periods, is not in accordance with observed facts; (2) the glacial periods would be limited in duration to an appropriate fraction of the precessional period (21,000 years), which appears to be too short a time for the work that was actually done by ice agency; and (3) croll's glacial periods would alternate between the northern and southern hemispheres, affecting first one then the other. sir c. lyell and others have advocated the view that great elevation of the land in polar regions would be conducive to glacial conditions; this is doubtless true, but the evidence that the glacial period was primarily due to this cause is not well established. other writers have endeavoured to support the elevation theory by combining with it various astronomical and meteorological agencies. more recently several hypotheses have been advanced to explain the glacial period as the result of changes in the atmosphere; f. w. harmer ("the influence of winds upon the climate during the pleistocene epoch," _q.j.g.s._, 1901, 57, p. 405) has shown the importance of the influence of winds in certain circumstances; marsden manson ("the evolution of climate," _american geologist_, 1899, 24, p. 93) has laid stress upon the influence of clouds; but neither of these theories grapples successfully with the fundamental difficulties. others again have requisitioned the variability in the amount of the carbon dioxide in the atmosphere--hypotheses which depend upon the efficiency of this gas as a thermal absorbent. the supply of carbon dioxide may be increased from time to time, as by the emanations from volcanoes (s. arrhenius and a. g. hogbom), or it may be decreased by absorption into sea-water, and by the carbonation of rocks. professor t. c. chamberlin based a theory of glaciation on the depletion of the carbon dioxide of the air ("an attempt to frame a working hypothesis of the cause of glacial periods on an atmospheric basis," _jl. geol._, 1899, vii. 752-771; see also chamberlin and salisbury, _geology_, 1906, ii. 674 and iii. 432). the outline of this hypothesis is as follows: the general conditions for glaciation were (1) that the oceanic circulation was interrupted by the existence of land; (2) that vertical circulation of the atmosphere was accelerated by continental and other influences; (3) that the thermal blanketing of the earth was reduced by a depletion of the moisture and carbon dioxide in the atmosphere, and that hence the average temperature of the surface of the earth and of the body of the ocean was reduced, and diversity in the distribution of heat and moisture introduced. the localization of glaciation is assignable to the two great areas of permanent atmospheric depression that have their present centres near greenland and the aleutian islands respectively. the periodicity of glacial advances and retreats, demanded by those who believe in the validity of so-called "interglacial" epochs, is explained by a series of complicated processes involving the alternate depletion and completion of the normal charge of carbon dioxide in the air. whatever may be the ultimate verdict upon this difficult subject, it is tolerably clear that no simple cause of glacial conditions is likely to be discovered, but rather it will appear that these conditions resulted from the interaction of a complicated series of factors; and further, until a greater degree of unanimity can be approached in the interpretation of observed facts, particularly as regards the substantiality of interglacial epochs, the very foundations of a sound working hypothesis are wanting. _classification of glacial deposits--interglacial epochs._--had the deposits of glaciated regions consisted solely of boulder clay little difficulty might have been experienced in dealing with their classification. but there are intercalated in the boulder clays those irregular stratified and partially stratified masses of sand, gravel and loam, frequently containing marine or freshwater shells and layers of peat with plant remains, which have given rise to the conception of "interglacial epochs"--pauses in the rigorous conditions of glaciation, when the ice-sheets dwindled almost entirely away, while plants and animals re-established themselves on the newly exposed soil. glacialists may be ranged in two schools: those who believe that one or more phases of milder climatic conditions broke up the whole glacial period into alternating epochs of glaciation and "deglaciation"; and those who believe that the intercalated deposits represent rather the _localized_ recessional movements of the ice-sheets within one single period of glaciation. in addition to the stratified deposits and their contents, important evidence in favour of interglacial epochs occurs in the presence of weathered surfaces on the top of older boulder clays, which are themselves covered by younger glacial deposits. the cause of the interglacial hypothesis has been most ardently championed in england by professor james geikie; who has endeavoured to show that there were in europe six distinct glacial epochs within the glacial period, separated by five epochs of more moderate temperature. these are enumerated below: 6th glacial epoch, upper turbarian, indicated by the deposits of peat which underlie the lower raised beaches. 5th _interglacial epoch, upper forestian_. 5th glacial epoch, lower turbarian, indicated by peat deposits overlying the lower forest-bed, by the raised beaches and carse-clays of scotland, and in part by the _littorina_-clays of scandinavia. 4th _interglacial epoch, lower forestian_, the lower forests under peat beds, the _ancylus_-beds of the great freshwater baltic lake and the _littorina_-clays of scandinavia. 4th glacial epoch, mecklenburgian, represented by the moraines of the last great baltic glacier, which reach their southern limit in mecklenburg; the 100-ft. terrace of scotland and the _yoldia_-beds of scandinavia. 3rd _interglacial epoch, neudeckian_, intercalations of marine and freshwater deposits in the boulder clays of the southern baltic coasts. 3rd glacial epoch, polandian, glacial and fluvio-glacial formations of the minor scandinavian ice-sheet; and the "upper boulder clay" of northern and western europe. 2nd _interglacial epoch, helvetian_, interglacial beds of britain and lignites of switzerland. 2nd glacial epoch, saxonian, deposits of the period of maximum glaciation when the northern ice-sheet reached the low ground of saxony, and the alpine glaciers formed the outermost moraines. 1st _interglacial epoch, norfolkian_, the forest-bed series of norfolk. 1st glacial epoch, scanian, represented only in the south of sweden, which was overridden by a large baltic glacier. the chillesford clay and weybourne crag of norfolk and the oldest moraines and fluvio-glacial gravels of the arctic lands may belong to this epoch. in a similar manner professor chamberlin and other american geologists have recognized the following stages in the glaciation of north america: the champlain, marine substage. the glacio-lacustrine substage. the later wisconsin (6th glacial). _the fifth interglacial._ the earlier wisconsin (5th glacial) _the peorian (4th interglacial)._ the iowan (4th glacial). _the sangamon (3rd interglacial)._ the illinoian (3rd glacial). _the yarmouth or buchanan (2nd interglacial)._ the kansan (2nd glacial). _the aftonian (1st interglacial)._ the sub-aftonian or jerseyan (1st glacial). although it is admitted that no strict correlation of the european and north american stages is possible, it has been suggested that the aftonian may be the equivalent of the helvetian; the kansan may represent the saxonian; the iowan, the polandian; the jerseyan, the scanian; the early wisconsin, the mecklenburgian. but considering how fragmentary is much of the evidence in favour of these stages both in europe and america, the value of such attempts at correlation must be infinitesimal. this is the more evident when it is observed that there are other geologists of equal eminence who are unable to accept so large a number of epochs after a close study of the local circumstances; thus, in the subjoined scheme for north germany, after h. w. munthe, there are three glacial and two interglacial epochs. / the _mya_ time = beech-time. post-glacial epoch < the _littorina_ time = oak-time. \ the _ancylus_ time = pine- and birch-time. / including the upper boulder clay, | "younger baltic moraine" with the 3rd glacial " < _yoldia_ or _dryas_ phase in the \ retrogressive stage. 2nd _interglacial_ epoch including the _cyprina_-clay. 2nd glacial epoch, the maximum glaciation. 1st _interglacial epoch_. 1st glacial epoch, "older boulder clay." again, in the alps four interglacial epochs have been recognized; while in england there are many who are willing to concede one such epoch, though even for this the evidence is not enough to satisfy all glacialists (g. w. lamplugh, address, section c, _brit. assoc._, york, 1906). this great diversity of opinion is eloquent of the difficulties of the subject; it is impossible not to see that the discovery of interglacial epochs bears a close relationship to the origin of certain hypotheses of the cause of glaciation; while it is significant that those who have had to do the actual mapping of glacial deposits have usually greater difficulty in finding good evidence of such definite ameliorations of climate, than those who have founded their views upon the examination of numerous but isolated areas. _extent of glacial deposits._--from evidence of the kind cited above, it appears that during the glacial period a series of great ice-sheets covered enormous areas in north america and north-west europe. the area covered during the maximum extension of the ice has been reckoned at 20 million square kilometres (nearly 8 million sq. m.) in north america and 6-1/2 million square kilometres (about 2-1/2 million sq. m.) in europe. in europe three great centres existed from which the ice-streams radiated; foremost in importance was the region of fennoscandia (the name for scandinavia with finland as a single geological region); from this centre the ice spread out far into germany and russia and westward, across the north sea, to the shores of britain. the southern boundary of the ice extended from the estuary of the rhine in an irregular series of lobes along the schiefergebirge, harz, thuringerwald, erzgebirge and riesengebirge, and the northern flanks of the carpathians towards cracow. down the valley of the dnieper a lobe of the ice-sheet projected as far as 40 deg. 50' n.; another lobe extended down the don valley as far as 48 deg. n.; thence the boundary runs north-easterly towards the urals and the kara sea. the british islands constituted the centre second in importance; scotland, ireland and all but the southern part of england were covered by a moving ice-cap. on the west the ice-sheets reached out to sea; on the east they were conterminous with those from scandinavia. the third european centre was the alpine region; it is abundantly clear from the masses of morainic detritus and perched blocks that here, in the time of maximum glaciation, the ice-covered area was enormously in excess of the shrivelled remnants, which still remain in the existing glaciers. all the valleys were filled with moving ice; thus the rhone glacier at its maximum filled lake geneva and the plain between the bernese oberland and the jura; it even overrode the latter and advanced towards besancon. extensive glaciation was not limited to the aforesaid regions, for all the areas of high ground had their independent glaciers strongly developed; the pyrenees, the central highlands of france, the vosges, black forest, apennines and caucasus were centres of minor but still important glaciation. the greatest expansion of ice-sheets was located on the north american continent; here, too, there were three principal centres of outflow: the "cordilleran" ice-sheet in the n.w., the "keewatin" sheet, radiating from the central canadian plains, and the eastern "labrador" or "laurentide" sheet. from each of these centres the ice poured outwards in every direction, but the principal flow in each case was towards the south-west. the southern boundary of the glaciated area runs as an irregular line along the 49 deg. parallel in the western part of the continent, thence it follows the mississippi valley down to its junction with the ohio (southern limit 37 deg. 30' n.), eastward it follows the direction of that river and turns north-eastward in the direction of new jersey. as in europe, the mountainous regions of north america produced their own local glaciers; in the rockies, the olympics and sierras, the bighorn mountains of wyoming, the uinta mountains of utah, &c. although it was in the northern hemisphere that the most extensive glaciation took place, the effects of a general lowering of temperature seem to have been felt in the mountainous regions of all parts; thus in south america, new zealand, australia and tasmania glaciers reached down the valleys far below the existing limits, and even where none are now to be found. in asia the evidences of a former extension of glaciation are traceable in the himalayas, and northward in the high ranges of china and eastern siberia. the same is true of parts of turkestan and lebanon. in africa also, in british east africa moraines are discovered 5400 ft. below their modern limit. in iceland and greenland, and even in the antarctic, there appears to be evidence of a former greater extension of the ice. it is of interest to note that alaska seems to be free from excessive glaciation, and that a remarkable "driftless" area lies in wisconsin. the maximum glaciation of the glacial period was clearly centred around the north atlantic. _glacial epochs in the older geological periods._--since ramsay drew attention to the subject in 1855 ("on the occurrence of angular, subangular, polished and striated fragments and boulders in the permian breccia of shropshire, worcestershire, &c., and on the probable existence of glaciers and icebergs in the permian epoch," _q.j.g.s._, 1855, pp. 185-205), a good deal of attention has been paid to such formations. it is now generally acknowledged that the permo-carboniferous conglomerates with striated boulders and polished rock surfaces, such as are found in the karoo formation of south africa, the talkir conglomerate of the salt range in india, and the corresponding formations in australia, represent undeniable glacial conditions at that period on the great indo-australian continent. a glacial origin has been suggested for numerous other conglomeratic formations, such as the pre-cambrian torridonian of scotland, and "geisaschichten" of norway; the basal carboniferous conglomerate of parts of england; the permian breccias of england and parts of europe; the trias of devonshire; the coarse conglomerates in the tertiary flysch in central europe; and the miocene conglomerates of the ligurian apennines. in regard to the glacial nature of all these formations there is, however, great divergence of opinion (see a. heim, "zur frage der exotischen blocke in flysch," _eclogae geologicae helvetiae_, vol. ix. no. 3, 1907, pp. 413-424). authorities.--the literature dealing directly with the glacial period has reached enormous dimensions; in addition to the works already mentioned the following may be taken as a guide to the general outline of the subject: j. geikie, _the great ice age_ (3rd ed., london, 1904), also _earth sculpture_ (1898); g. f. wright, _the ice age in north america_ (4th ed., new york, 1905) and _man and the glacial period_ (1892); f. e. geinitz, _die eiszeit_ (braunschweig, 1906); a. penck and e. bruckner, _die alpen im eiszeitalter_ (leipzig, 1901-1906, uncompleted). many references to the literature will be found in sir a. geikie's _textbook of geology_, vol. ii. (4th ed., 1903); chamberlin and salisbury, _geology_, vol. iii. (1906). as an example of glacial theories carried beyond the usual limits, see m. gugenhan, _die ergletscherung der erde von pol zu pol_ (berlin, 1906). see also _zeitschrift fur gletscherkunde_ (berlin, 1906 and onwards quarterly); sir h. h. howorth (opposing accepted glacial theories), _the glacial nightmare and the flood_, i., ii. (london, 1893), _ice and water_, i., ii. (london, 1905), _the mammoth and the flood_ (london, 1887). (j. a. h.) glacier (adopted from the french; from _glace_, ice, lat. _glacies_), a mass of compacted ice originating in a snow-field. glaciers are formed on any portion of the earth's surface that is permanently above the snow-line. this line varies locally in the same latitudes, being in some places higher than in others, but in the main it may be described as an elliptical shell surrounding the earth with its longest diameter in the tropics and its shortest in the polar regions, where it touches sea-level. from the extreme regions of the arctic and antarctic circles this cold shell swells upwards into a broad dome, from 15,000 to 18,000 ft. high over the tropics, truncating, as it rises, a number of peaks and mountain ranges whose upper portions like all regions above this thermal shell receive all their moisture in the form of snow. since the temperature above the snow-line is below freezing point evaporation is very slight, and as the snow is solid it tends to accumulate in snow-fields, where the snow of one year is covered by that of the next, and these are wrapped over many deeper layers that have fallen in previous years. if these piles of snow were rigid and immovable they would increase in height until the whole field rose above the zone of ordinary atmospheric precipitation, and the polar ice-caps would add a load to these regions that would produce far-reaching results. the mountain regions also would rise some miles in height, and all their features would be buried in domes of snow some miles in thickness. when, however, there is sufficient weight the mass yields to pressure and flows outwards and downwards. thus a balance of weight and height is established, and the ice-field is disintegrated principally at the edges, the surplus in polar regions being carried off in the form of icebergs, and in mountain regions by streams that flow from the melting ends of the glaciers. _formation._--the formation of glaciers is in all cases due to similar causes, namely, to periodical and intermittent falls of snow. after a snow-fall there is a period of rest during which the snow becomes compacted by pressure and assumes the well-known granular character seen in banks and patches of ordinary snow that lie longest upon the ground when the snow is melting. this is the _firn_ or _neve_. the next fall of snow covers and conceals the neve, but the light fresh crystals of this new snow in turn become compacted to the coarsely crystalline granular form of the underlying layer and become neve in turn. the process goes on continually; the lower layers become subject to greater and greater pressure, and in consequence become gradually compacted into dense clear ice, which, however, retains its granular crystalline texture throughout. the upper layers of neve are usually stratified, owing to some individual peculiarity in the fall, or to the accumulation of dust or debris upon the surface before it is covered by fresh snow. this stratification is often visible on the emerging glacier, though it is to be distinguished from the foliation planes caused by shearing movement in the body of the glacier ice. _types._--the snow-field upon which a glacier depends is always formed when snow-fall is greater than snow-waste. this occurs under varying conditions with a differently resulting type of glacier. there are limited fields of snow in many mountain regions giving rise to long tongues of ice moving slowly down the valleys and therefore called "valley glaciers." the greater part of greenland is covered by an ice-cap extending over nearly 400,000 sq. m., forming a kind of enormous continuous glacier on its lower slopes. the antarctic ice region is believed to extend over more than 3,000,000 sq. m. each of these continental fields, besides producing block as distinguished from tongue glaciers, sends into the sea a great number of icebergs during the summer season. these ice-caps covering great regions are by far the most important types. between these "polar" or "continental glaciers" and the "alpine" type there are many grades. smaller detached ice-caps may rest upon high plateaus as in iceland, or several tongues of ice coming down neighbouring valleys may splay out into convergent lobes on lower ground and form a "piedmont glacier" such as the malaspina glacier in alaska. when the snow-field lies in a small depression the glacier may remain suspended in the hollow and advance no farther than the edge of the snow-field. this is called a "cliff-glacier," and is not uncommon in mountain regions. the end of a larger glacier, or the edge of an ice-sheet, may reach a precipitous cliff, where the ice will break from the edge of the advancing mass and fall in blocks to the lower ground, where a "reconstructed glacier" will be formed from the fragments and advance farther down the slope. when a glacier originates upon a dome-shaped or a level surface the ice will deploy radially in all directions. when a snow-field is formed above steep valleys separated by high ridges the ice will flow downwards in long streams. if the valleys under the snow-fields are wide and shallow the resultant glaciers will broaden out and partially fill them, and in all cases, since the conditions of glacier formation are similar, the resultant form and the direction of motion will depend upon the amount of ice and the form of the surface over which the glacier flows. a glacier flowing down a narrow gorge to an open valley, or on to a plain, will spread at its foot into a fan-shaped lobe as the ice spreads outwards while moving downwards. an ice-cap is in the main thickest at the centre, and thins out at the edges. a valley glacier is thickest at some point between its source and its end, but nearer to its source than to its termination, but its thickness at various portions will depend upon the contour of the valley floor over which the glacier rides, and may reach many hundreds of feet. at its centre the greenland ice-cap is estimated to be over 5000 ft. thick. in all cases the glacier ends where the waste of ice is greater than the supply, and since the relationship varies in different years, or cycles of years, the end of a glacier may advance or retreat in harmony with greater or less snow-fall or with cooler or hotter summers. there seems to be a cycle of inclusive contraction and expansion of from 35 to 40 or 50 years. at present the ends of the swiss glaciers are cradled in a mass of moraine-stuff due to former extension of the glaciers, and investigations in india show that in some parts of the himalayas the glaciers are retreating as they are in north america and even in the southern hemisphere (_nature_, january 2, 1908, p. 201). _movement._--the fact that a glacier moves is easily demonstrated; the cause of the movement is pressure upon a yielding mass; the nature of the movement is still under discussion. rows of stakes or stones placed in line across a glacier are found to change their position with respect to objects on the bank and also with regard to each other. the posts in the centre of the ice-stream gradually move away from those at the side, proving that the centre moves faster than the sides. it has also been proved that the surface portions move more rapidly than the deeper layers and that the motion is slowest at the sides and bottom where friction is greatest. the rate of motion past the same spot is not uniform. heat accelerates it, cold arrests it, and the pressure of a large amount of water stimulates the flow. the rate of flow under the same conditions varies at different parts of the glacier directly as the thickness of ice, the steepness of slope and the smoothness of rocky floor. generally speaking, the rate of motion depends upon the amount of ice that forms the "head" pressure, the slope of the under surface and of the upper surface, the nature of the floor, the temperature and the amount of water present in the ice. the ordinary rate of motion is very slow. in switzerland it is from 1 or 2 in. to 4 ft. per day, in alaska 7 ft., in greenland 50 to 60 ft., and occasionally 100 ft. per day in the height of summer under exceptional conditions of quantity of ice and of water and slope. measurements of swiss glaciers show that near the ice foot where wastage is great there is very little movement, and observations upon the inland border of greenland ice show that it is almost stationary over long distances. in many aspects the motion of a body of ice resembles that of a body of water, and an alpine glacier is often called an ice-river, since like a river it moves faster in the centre than at the sides and at the top faster than at the bottom. a glacier follows a curve in the same way as a river, and there appear to be ice swirls and eddies as well as an upward creep on shelving curves recalling many features of stream action. the rate of motion of both ice-stream and river is accelerated by quantity and steepness of slope and retarded by roughness of bed, but here the comparison ends, for temperature does not affect the rate of water motion, nor will a liquid crack into crevasses as a glacier does, or move upwards over an adverse slope as a glacier always does when there is sufficient "head" of ice above it. so that although in many respects ice behaves as a viscous fluid the comparison with such a fluid is not perfect. the cause of glacier motion must be based upon some more or less complex considerations. the flakes of snow are gradually transformed into granules because the points and angles of the original flakes melt and evaporate more readily than the more solid central portions, which become aggregated round some master flake that continues to grow in the neve at the expense of its smaller neighbours, and increases in size until finally the glacier ice is composed of a mass of interlocked crystalline granules, some as large as a walnut, closely compacted under pressure with the principal crystalline axes in various directions. in the upper portions of the glacier movement due to pressure probably takes place by the gliding of one granule over another. in this connexion it must be noted that pressure lowers the melting point of ice while tension raises it, and at all points of pressure there is therefore a tendency to momentary melting, and also to some evaporation due to the heat caused by pressure, and at the intermediate tension spaces between the points of pressure this resultant liquid and vapour will be at once re-frozen and become solid. the granular movement is thus greatly facilitated, while the body of ice remains in a crystalline solid condition. in this connexion it is well to remember that the pressure of the glacier upon its floor will have the same result, but the effect here is a mass-effect and facilitates the gliding of the ice over obstacles, since the friction produces heat and the pressure lowers the melting point, so that the two causes tend to liquefy the portion where pressure is greatest and so to "lubricate" the prominences and enable the glacier to slide more easily over them, while the liquid thus produced is re-frozen when the pressure is removed. in polar regions of very low temperature a very considerable amount of pressure must be necessary before the ice granules yield to momentary liquefaction at the points of pressure, and this probably accounts for the extreme thickness of the arctic and antarctic ice-caps where the slopes are moderate, for although equally low temperatures are found in high alpine snow-fields the slopes there are exceedingly steep and motion is therefore more easily produced. observations made upon the greenland glaciers indicate a considerable amount of "shearing" movement in the lower portions of a glacier. where obstacles in the bed of the glacier arrest the movement of the ice immediately above it, or where the lower portion of the glacier is choked by debris, the upper ice glides over the lower in shearing planes that are sometimes strongly marked by debris caught and pushed forwards along these planes of foliation. it must be remembered that there is a solid push from behind upon the lower portion of a glacier, quite different from the pressure of a body of water upon any point, for the pressure of a fluid is equal in all directions, and also that this push will tend to set the crystalline granules in positions in which their crystalline axes are parallel along the gliding planes. the production of gliding planes is in some cases facilitated by the descent into the glacier of water melted during summer, where it expands in freezing and pushes the adjacent ice away from it, forming a surface along which movement is readily established. if under all circumstances the glacier melted under pressure at the bottom, glacial abrasion would be nearly impossible, since every small stone and fragment of rock would rotate in a liquid shell as the ice moved forward, but since the pressure is not always sufficient to produce melting, the glacier sometimes remains dry at its base; rock fragments are held firmly; and a dry glacier may thus become a graving tool of enormous power. whatever views may be adopted as to the causes of glacier motion, the peculiar character of glacier ice as distinct from homogeneous river or pond ice must be kept in view, as well as the characteristic tendency of water to expand in freezing, the lowering of the melting point of ice under pressure, the raising of the melting point under tension, the production of gliding or shearing planes under pressure from above, the presence in summer of a considerable quantity of water in the lower portions of the glacier which are thus loosened, the cracking of ice (as into crevasses), under sudden strain, and the regelation of ice in contact. a result of this last process is that fissures are not permanent, but having been produced by the passage of ice over an obstruction, they subsequently become healed when the ice proceeds over a flatter bed. finally it must be remembered that although glacier ice behaves in some sense like a viscous fluid its condition is totally different, since "a glacier is a crystalline rock of the purest and simplest type, and it never has other than the crystalline state." _characteristics._--the general appearance of a glacier varies according to its environment of position and temperature. the upper portion is hidden by neve and often by freshly fallen snow, and is smooth and unbroken. during the summer, when little snow falls, the body of the glacier moves away from the snow-field and a gaping crevasse of great depth is usually established called the _bergschrund_, which is sometimes taken as the upper limit of the glacier. the glacier as it moves down the valley may become "loaded" in various ways. rock-falls send periodical showers of stones upon it from the heights, and these are spread out into long lines at the glacier sides as the ice moves downwards carrying the rock fragments with it. these are the "lateral moraines." when two or more glaciers descending adjacent valleys converge into one glacier one or more sides of the higher valleys disappear, and the ice that was contained in several valleys is now carried by one. in the simplest case where two valleys converge into one the two inner lateral moraines meet and continue to stream down the larger valley as one "median moraine." where several valleys meet there are several such parallel median moraines, and so long as the ice remains unbroken these will be carried upon the surface of the glacier and finally tipped over the end. there is, however, differential heating of rock and ice, and if the stones carried are thin they tend to sink into the ice because they absorb heat readily and melt the ice under them. dust has the same effect and produces "dust wells" that honeycomb the upper surface of the ice with holes into which the dust sinks. if the moraine rocks are thick they prevent the ice under them from melting in sunlight, and isolated blocks often remain supported upon ice-pillars in the form of ice tables, which finally collapse, so that such rocks may be scattered out of the line of the moraine. as the glacier descends into the lower valleys it is more strongly heated, and surface streams are established in consequence that flow into channels caused by unequal melting of the ice and finally plunge into crevasses. these crevasses are formed by strains established as the central parts drag away from the sides of the glacier and the upper surface from the lower, and more markedly by the tension due to a sudden bend in the glacier caused by an inequality in its bed which must be over-ridden. these crevasses are developed at right angles to the strain and often produce intersecting fissures in several directions. the morainic material is gradually dispersed by the inequalities produced, and is further distributed by the action of superficial streams until the whole surface is strewn with stones and debris, and presents, as in the lower portions of the mer de glace, an exceedingly dirty appearance. many blocks of stone fall into the gaping crevasses and much loose rock is carried down as "englacial material" in the body of the glacier. some of it reaches the bottom and becomes part of the "ground moraine" which underlies the glacier, at least from the _bergschrund_ to the "snout," where much of it is carried away by the issuing stream and spread finally on to the plains below. it appears that a very considerable amount of degradation is caused under the _bergschrund_ by the mass of ice "plucking" and dragging great blocks of rock from the side of the mountain valley where the great head of ice rests in winter and whence it begins to move in summer. these blocks and many smaller fragments are carried downwards wedged in the ice and cause powerful abrasion upon the rocky floor, rasping and scoring the channel, producing conspicuous striae, polishing and rounding the rock surfaces, and grinding the contained fragments as well as the surface over which it passes into small fragments and fine powder, from which "boulder clay" or "till" is finally produced. emerging, then, from the snow-field as pure granular ice the glacier gradually becomes strewn and filled with foreign material, not only from above but also, as is very evident in some greenland glaciers, occasionally from below by masses of fragments that move upwards along gliding planes, or are forced upwards by slow swirls in the ice itself. as a glacier is a very brittle body any abrupt change in gradient will produce a number of crevasses, and these, together with those produced by dragging strains, will frequently wedge the glacier into a mass of pinnacles or _seracs_ that may be partially healed but are usually evident when the melting end of the glacier emerges suddenly from a steep valley. here the streams widen the weaker portions and the moraine rocks fall from the end to produce the "terminal" moraine, which usually lies in a crescentic heap encircling the glacier snout, whence it can only be moved by a further advance of the glacier or by the ordinary slow process of atmospheric denudation. in cases where no rock falls upon the surface there is a considerable amount of englacial material due to upturning either over accumulated ground debris or over structural inequalities in the rock floor. this is well seen at the steep sides and ends of greenland glaciers, where material frequently comes to the surface of the melting ice and produces median and lateral moraines, besides appearing in enormous "eyes" surrounded in the glacial body by contorted and foliated ice and sometimes producing heaps and embankments as it is pushed out at the end of the melting ice. the environment of temperature requires consideration. at the upper or dorsal portion of the glacier there is a zone of variable (winter and summer) temperature, beneath which, if the ice is thick enough, there is a zone of constant temperature which will be about the mean annual temperature of the region of the snow-field. underlying this there is a more or less constant ventral or ground temperature, depending mainly upon the internal heat of the earth, which is conducted to the under surface of the glacier where it slowly melts the ice, the more readily because the pressure lowers the melting point considerably, so that streams of water run constantly from beneath many glaciers, adding their volume to the springs which issue from the rock. the middle zone of constant temperature is wedge-shaped in "alpine" glaciers, the apex pointing downwards to the zone of waste. the upper zone of variable temperature is thinnest in the snow-field where the mean temperature is lowest, and entirely dominant in the snout end of the glacier where the zone of constant temperature disappears. two temperature wedges are thus superposed base to point, the one being thickest where the other is thinnest, and both these lie upon the basal film of temperature where the escaping earth-heat is strengthened by that due to friction and pressure. the cold wave of winter may pass right through a thin glacier, or the constant temperature may be too low to permit of the ice melting at the base, in which cases the glacier is "dry" and has great eroding power. but in the lower warmer portions water running through crevasses will raise the temperature, and increase the strength of the downward heat wave, while the mean annual temperature being there higher, the combined result will be that the glacier will gradually become "wet" at the base and have little eroding power, and it will become more and more wet as it moves down the lower valley zone of ice-waste, until at last the balance is reached between waste and supply and the glacier finally disappears. if the mean annual temperature be 20 deg. f., and the mean winter temperature be -12 deg. f., as in parts of greenland, all the ice must be considerably below the melting point, since the pressure of ice a mile in depth lowers the melting point only to 30 deg. f., and the earth-heat is only sufficient to melt 1/4 in. of ice in a year. therefore in these regions, and in snow-fields and high glaciers with an equal or lower mean temperature than 20 deg. f., the glacier will be "dry" throughout, which may account for the great eroding power stated to exist near the _bergschrund_ in glaciers of an alpine type, which usually have their origin on precipitous slopes. a considerable amount of ice-waste takes place by water-drainage, though much is the result of constant evaporation from the ice surface. the lower end of a glacier is in summer flooded by streams of water that pour along cracks and plunge into crevasses, often forming "pot-holes" or _moulins_ where stones are swirled round in a glacial "mill" and wear holes in the solid rock below. some of these streams issue in a spout half way up the glacier's end wall, but the majority find their way through it and join the water running along the glacier floor and emerging where the glacier ends in a large glacial stream. _results of glacial action._--a glacier is a degrading and an aggrading agent. much difference of opinion exists as to the potency of a glacier to alter surface features, some maintaining that it is extraordinarily effective, and considering that a valley glacier forms a pronounced _cirque_ at the region of its origin and that the cirque is gradually cut backward until a long and deep valley is formed (which becomes evident, as in the rocky mountains, in an upper valley with "reversed grade" when the glacier disappears), and also that the end of a glacier plunging into a valley or a fjord will gouge a deep basin at its region of impact. the alaskan and norwegian fjords and the rock basins of the scottish lochs are adduced as examples. other writers maintain that a glacier is only a modifying and not a dominant agent in its effects upon the land-surface, considering, for example, that a glacier coming down a lateral valley will preserve the valley from the atmospheric denudation which has produced the main valley over which the lateral valley "hangs," a result which the believers in strong glacial action hold to be due to the more powerful action of the main glacier as contrasted with the weaker action of that in the lateral valley. both the advocates and the opponents of strenuous ice action agree that a v-shaped valley of stream erosion is converted to a u-shaped valley of glacial modification, and that rock surfaces are rounded into _roches moutonnees_, and are grooved and striated by the passage of ice shod with fragments of rock, while the subglacial material is ground into finer and finer fragments until it becomes mud and "rock-flour" as the glacier proceeds. in any case striking results are manifest in any formerly glaciated region. the high peaks rise into pinnacles, and ridges with "house-roof" structure, above the former glacier, while below it the contours are all rounded and typically subdued. a landscape that was formerly completely covered by a moving ice-cap has none but these rounded features of dome-shaped hills and u-shaped valleys that at least bear evidence to the great modifying power that a glacier has upon a landscape. there is no conflict of opinion with regard to glacial aggradation and the distribution of superglacial, englacial and subglacial material, which during the active existence of a glacier is finally distributed by glacial streams that produce very considerable alluviation. in many regions which were covered by the pleistocene ice-sheet the work of the glacier was arrested by melting before it was half done. great deposits of till and boulder clay that lay beneath the glaciers were abandoned _in situ_, and remain as an unsorted mixture of large boulders, pebbles and mingled fragments, embedded in clay or sand. the lateral, median and terminal moraines were stranded where they sank as the ice disappeared, and together with perched blocks (_roches perchees_) remain as a permanent record of former conditions which are now found to have existed temporarily in much earlier geological times. in glaciated north america lateral moraines are found that are 500 to 1000 ft. high and in northern italy 1500 to 2000 ft. high. the surface of the ground in all these places is modified into the characteristic glaciated landscape, and many formerly deep valleys are choked with glacial debris either completely changing the local drainage systems, or compelling the reappearing streams to cut new channels in a superposed drainage system. kames also and eskers (q.v.) are left under certain conditions, with many puzzling deposits that are clearly due to some features of ice-work not thoroughly understood. see l. agassiz, _etudes sur les glaciers_ (neuchatel, 1840) and _nouvelles etudes ..._ (paris, 1847); n. s. shaler and w. m. davis, _glaciers_ (boston, 1881); a. penck, _die begletscherung der deutschen alpen_ (leipzig, 1882); j. tyndall, _the glaciers of the alps_ (london, 1896); t. g. bonney, _ice-work, past and present_ (london, 1896); i. c. russell, _glaciers of north america_ (boston, 1897); e. richter, _neue ergebnisse und probleme der gletscherforschung_ (vienna, 1899); f. forel, _essai sur les variations periodiques des glaciers_ (geneva, 1881 and 1900); h. hess, _die gletscher_ (brunswick, 1904). (e. c. sp.)