^ .^^ V] %. '/ IMAGE EVALUATION TEST TARGET (MT-3) 1.0 1.1 11.25 1^128 |25 ■^ 1^ 12.2 L£ 12.0 US lit i: 6" - Photographic Sciences Corporation -► 23 WIST MAIN STRilT WIBSTIR,N.Y. MSM (716)t72-4S03 '^ CIHM/ICMH Microfiche Series. CIHM/ICMH Collection de microfiches. Canadian Institute for Historical Microreproductions / Institut Canadian de microreproductions historiques Technical and Bibliographic Notes/Notes techniques et bibliographiques The Institute has attempted to obtain the best original copy available for filming. Features of this copy which may be bibliographically unique, which may alter any of the images in the reproduction, or which may significantly change the usual method of filming, are checked below. 0" D D D D D Coloured covers/ Couverture de couleur Covers damaged/ Couverture endommag^e Covers restored and/or laminated/ Couverture restaur6e et/ou pellicul6e Cover title missing/ Le titre de couverture manque Coloured maps/ Cartes gdographiques en couleur Coloured ink (i.e. other than blue or black)/ Encre de couleur (i.e. autre que bleue ou noire) r~> Coloured plates and/or illustrations/ D D D Planches et/ou illustrations en couleur Bound with other material/ Relid avec d'autres documents Tight binding may cause shadows or distortion along interior margin/ La reliure serr^e peut causer de I'ombre ou de la distortion le long de la marge intdrieure Blank leaves added during restoration may appear within the text. Whenever possible, these have been omitted from filming/ II se peut que certaines pages blanches ajout^es lors d'une restauration apparaissent dans le texte, mais, lorsque cela dtait possible, ces pages n'ont pas M film4es. Additional comments:/ Commentaires suppldmentaires: L'Institut a microfilm^ le meilleur exemplaire qu'il lui a 6ti possible de se procurer. Les details de cet exemplaire qui sont peut-dtre uniques du point de vue bibliographique, qui peuvent modifier une image reproduite, ou qui peuvent exiger une modification dans la m6thode normals de filmage sont indiqu6s ci-dessous. □ Coloured pages/ Pages de couleur □ Pages damaged/ Pages endommag^es □ Pages restored and/or laminated/ Pages restauries et/ou pellicul6es Pages discoloured, stained or foxed/ Pages ddcolordes, tachetdes ou piqu6es I I Pages detached/ D D This item is filmed at the reduction ratio checked below/ Ce document est filmi au taux de reduction indiqu^ ci-dessous. Pages d^tachdes Showthrough/ Transparence Quality of prir Qualit^ indgale de I'impression Includes supplementary materif Comprend du materiel supplementaire I I Showthrough/ I I Quality of print varies/ I I Includes supplementary material/ Only edition available/ Seule Edition disponible Pages wholly or partially obscured by errata slips, tissues, etc., have been refilmed to ensure the best possible image/ Les pages totalement ou partiellement obscurcies par un feuillet d'errata, une pelure, etc., ont 6t6 filmdes d nouveau de faqon d obtenir la meilleure image possible. 10X 14X 18X 22X 26X 30X ■ I ■ • . ■ ■- ; . „■ 12X 16X 20X MX 28X 32X '■-■'-': The copy filmed here has been reproduced thanks to the generosity of: Thomas Fiihor Rare Boole Library, University of Toronto Library L'exemplaire film4 fut reproduit grAce i la g6n6rosit4 de: Thomas Fisher Rare Book Library, University of Toronto Library The images appearing here are the best quality possible considering the condition and legibility of the original copy and in keeping with the filming contract specifications. Les images suivantes ont 6t6 reproduites avec le plus grand som. compte tenu de la condition et de la nettetd de l'exemplaire film6, et en conformit6 avec les conditions du contrat de filmage. Original copies in printed paper covers are filmed beginning with the front cover and ending on the last page with a printed or illustrated impres- sion, or the back cover when appropriate. All other original copies are filmed beginning on the first page with a printed or illustrated impres- sion, and ending on the last page with a printed or illustrated impression. Les exemplaires originaux dont la couverture en papier est imprimie sont film^s en commenpant par le premier plat et en terminant soit par la dernidre page qui comporte une empreinte d'impression ou d'illustration, soit par le second plat, selon le cas. Tous les autres exemplaires originaux sont film6s en commenpant par la premiere page qui comporte une empreinte d'impression ou d'illustration et en terminant par la dernidre page qui comporte une telle empreinte. The last recorded frame on each microfiche shall contain the symbol —^•(meaning "CON- TINUED"), or the symbol V (meaning "END"), whichever applies. Un des symboles suivants apparattra sur la dernidre image de cheque microfiche, selon le cas: le symbols — ► signifie "A SUIVRE", le symbols V signifie "FIN". Maps, plates, charts, etc., may be filmed at different reduction ratios. Those too large to be entirely included in one exposure are filmed beginning in the upper left hand corner, left to right and top to bottom, as many frames as required. The following diagrams Illustrate the method: Les cartes, planches, tableaux, etc., peuvent Atre filmAs d des taux de reduction diffArents. Lorsque le document est trop grand pour Atre reproduit en un seul clichA, 11 est filmA d partir de Tangle supArieur gauche, de gauche A droite, et de haut en bas, en prenant le nombre d'images nAcessaire. Les diagrammes suivants illustrent la mAthode. 1 2 3 r * 2 3 A 5 6 1' '•' 1 ' /.//^^^^ u LATKK rilYSIOGKArillCAL CKoLOCV OF THE IKilKV :\I0LXTA1N KKdioN IN CANADA, WlTfl SPECIAL KEFEIiENCE To CHANGES IN E!.EVATli»X AND THE HISTOKV OF THE (iLACFAL I'EliloD. By GEORGE M. DAWSON, LL.D,, F.G.S. Assistiint l)iroi:tor (jioluLjicul .Survey of Canada. l-KOM TIIK TltANSACTIONS 01' TIIK ROYAL S0( IKTY OF CANADA, YO!.. \ Hi. SKr. lY. 181)0. Section IV, 18J»0. [ 3 1 TiiANs. KoY. S(K!. Canada. I. — On the later Phi/si'ogniphica/ OtoJoi/t/ of tlit Uorky Afonntnin Jirginn in C'anofh. with Hper'ud rrfcfevcr to C'hnntjrs in I'JIcvotlon and to the IJlxtort/ n/ tJn- QUtcial PcrimJ ; hciny the Presidential Address /or the year. By Gei»K(1E M. Dawson, D.Sc, A.R.S.M., F.G.S., AssistiiHf Dimlor Geotn« classotl togothor as ihe Gold Ranges, (3) Th«> Coast Kaiigt's, (4) An irn'gular inountain-systcm which in its nnsubraorgod parts ronstitutt's VanioiiviT Island and tho Quoon Charlotto Islands, and which may bo di'signattMl thf Vamouvfr systi-m. Ki'twoon the socond aJid third of those mountain- systi'ms is a region without important mountain ranges, whioh is rel'errod to as tho Interior Plateau of Hritish Columbia. To siniplily our ption olthe main features of this part of the Cordillera for our present jjurpose, we may, however, regard it l>roadly us being outlined on the north- east and south-west sides liy the Koeky Mountains proper and by the Coast Ranges, as dominant sy.-tems. This view is justiKed by the remarkable eonstancy of these two systems and their relative iniitortance. The intervening region may then be described as comprising the Interior Plateau together with the various ranges which have been grouped togetli'-r under the name of the Gold Ranges, as well as other detached moun- tains and irregular mountainous tracts. The Cordillera belt, taken in the aggregate, constitutes an elevated region, of which the broken Vancouver system of mountains forms the south-western border, based upon the edge of the continental plateau, whic^h, at a short distance beyond it, shelves rapidly down from moderate soundings to the iJjyssal depths of the Pacific basin. On the other side, the north-eastern base of the Itocky Mountains is bordered by a narrow zone of foot- hills, beyond which, in easti'rly and north-easterly directions, the Great Plains slope gradually down to the base of the old Laurentian plateau. The elevation of these plains at the eastern ba.se of the Rfjcky Mountains, in the vicinity of the forty-uinth parallel of latitude, is al)out 4,0'tO feet, about .-J.OOO feet only near the iifty-sixth parallel, and is even less still further north. The average ]>n'adth of the Rocky Mountain Range proper in the southern part of British Columbia is about sixty miles, but the range decreases near the Pcaca River to twenty miles or less in width, and apjiarently loses its imiportance and regularity locally where cut through bv the Liard River, though recovering both still further to the north- westward. Near the forty-ninth parallel, several summits occur in this range which exceed 10,000 feet in height, but northward, few attain to this height till the head-waters of the Bow River are reached. About the sources of the North Saskatchewan and Athabasca, the range appears to culminate, and Mounts Rrown and Murchison occur, with reputed heights of 1(!,000 and 13,500 feet respectively. Near the Peace, few sum- mits ex(!oed tl.OOO feet, so far as known.' The Coast system of mountains runs without any important break or change in char- acter from the viA. S and other rimges, lie betwet'ii the Interior Plateau and the south-western base of the Rot;ky Mountains proper, from which thoy are separated by a long and straight valley only. Many peaks in these ranges are scarcely inferior in altitude to those of the Rocky Mountains themselves, and the subsidiary position accorded to them in this very general description of the country, is justified only by their more irregular character and absence of orographic continuity. Some of the rocks entering into the composition of this mountain-system are regarded as Archiean, and these rocks are more continuous along the line of their strike than the mountains themselves. It is i ^bable indeed that this mountain axis is of greater age, as a physical feature, than either the liocky Mountains or the Coast Ranges. The Interior Plateau of British Columbia will be found to possess particular import- ance in the present enquiry. It has an average width of aboxit one hundred miles, and I have described it in previous publications as having a mean elevation of about y,500 feet. It does not, however, maintain the same continuity iu a north-west and south-east bearing as do the Rocky Mountains and Coast Ranges, being practically closed to the north about latitude 55 30' by the ends of several rather high mountain ranges, which are here interpolated between the Rocky Mountains and Coast Ranges. Nearly coincident with the forty-ninth parallel, it is cut off in the same manner by a second transverse moun- tainous region, formed in a similar way to the first, which separates it from the great lava- plain of the Columbia River. It has thus a length of about live hundred miles. Its elevation decreases gradually from south-east to north-west, and it is now traversed in various directions by systems of trough-like valleys of erosion, which are deepest where the general level of the plateau is greatest. Water standing at an elevation of 3,000 feet above the present sealevel would Hood most of these valleys in its southern part, while in its northern portion, a large tract of country about the lifty-third and fifty-fourth parallels of latitude would be completely submerged by it. The mean elevation above stated must therefore be accepted merely iu a general way, and is an approximation to the average height which this region might have if its irregularities were levelled down. It is in fact, as a rule, only when broadly viewed, and in contrast with its bordering moun- tain ranges, that its true character as a plateau becomes apparent. North of the Cariboo Mountains on the fifty-fourth and fifty-fifth parallels of latitude, the plateau region interrupts the line of the Gold Ranges and abuts directly on the inner slope of tht^ Rocky Mountains proper. Southward from the Cariboo Mountains, it is irregularly bounded to the eastward ])y the several members of the Gold itanges. Beyond the northern end of the Interior Plateau, the interval between the Rocky Mountains and Coast Ranges appears to be occupied by an irregular mountainous country about which little is yet known, for a distance of about 250 miles, till, in the vicinity of the fifty-eighth parallel, this country again begins to assume a plateau-like character, having at first a height of about 2,500 feet, but sloping down gradually north-westward, and constituting the upper drainage basin of the Yukon. Having thus very briefly characterized the main features of this Avestern portion of the continent as they now exist, we are iu a position to follow iu greater detail the steps by which they have been produced. Omitting then from consideration the imperfectly known progress of events in the earlier stages of the geological history of the region, we may endeavour to picture to ourselves its condition in the Triassic or first stage in the 6 (;. M. KAWSON ON TIIK IMIYSIOCUAI'IIFCJAL (iF,()Ti()(iY I. MoNozoic cliviNinn ol' }?foIoiiii'iil time. Tho central n'h sliallow, mcilitcrrancan soa, whicli was citln'r fiiliicly rut (iHTroin tin' oifini or had only oiciisional and hrit'l' lonuct-tion with it. and in whirh rt'd l>i'ds willi ocnisional layers ofijypsnni and sail were lieinj? deposited.' Kfurks whieh represent a portion of the lied of this inland s-a enter into the ((nnjiosition of tho lloeky Mountain itanife near the I'orty-ninlh parallol. hut aro not known to oectir to tho north of that parallel lor a distame of nior<' than thirty or forty miles. To the west, they are not found in the Selkirk or I'ureell Mountains. We nppoar in fact to discover in ihis vicinity tho northern end of the inland Tria.ssic sea. To the west of the (rold Ranely of materials of volcanic origin, which give tn'idenco of contemporaneous volcanic activity on a great seale. To the north, in tho Poaco River country, and to tli' east of I lie present position of the Kocky Mountains, rooks holding the same marine forms are found, and they iiave (juite recently again been discovered by Mr. Mt<':)nnell in a similar position, still further north, on the Liard U'iver. Il would thus appear that in Triassic times the eastern border of the Pacific washi'd the western slopes of the (J old Ranges, and that where this mountain-system became in- terrupted, in its northern part, the sea was continued across its line and covered a large tract of country to the east of the present position of the whole Cordillera belt. I'recisely how far to tho east the shore of this northern expansion of the Pacific was situated, has not yet boon determined. The region between it and the northern end of the inland sea pre vicmsly referred to must have been a land area, which separated the open ocean of tho north from the mediterraiuian on the south. The Rocky Mountains proper had not yet boon formed, nor is there any evidence of mountain ranges in the region of tho Coast and Vancouver systems of to-day, though tho volcanic action tluTi^ in progress may have produced insular volcanic peaks. The deposits of tho inland Triassic sea, in- cluding as they do beds of salt and gypsum, ajjpoar to prove tho existence of a very dry climate in tho area occupied by it, and as the land barrier separating it from the moisture-bearing westerly winds of tho Pacific cau not have been wide, it must have been high. It is thus probable that the mountains of tho Gold system formed at this time a lofty sierra, which was continued to the south of the forty-ninth parallel by the Cabinet, Cour D'Alaine, Bitter Root and other mountains at least as far as tho Wahsatch Range in Utah. Tho Triassic period was closed by one of those epochs of folding and dislocation of strata which aro found to be recurrent in geological time, and which are generally attri- butot<>ly iiirosstlioprt'Sfnt iwRilionofthe Cordilltira hfll, wliilf lii('(iolH, and |inil)iil>ly also miiiiy ollirr iiisultir urotiH, lunliiined to t'liHl UN dry land, lii tliJH ras(>, as in tluit ol' I lit; TriaHKic, il h»H not yelbci-n i'uiind puisisi))lt> looullini' fxmliy Ih*' ftwlfrn liniil oi' llie soa, in consi'tiuenci' ol" tlu' want ol' Hcctions cnUiiijr down lo tin- Imsc ol lln' (In-tan'ous in tho iinni oi" the (in-al I'lains. There uro, however, reusonw lor hejicvinf? that it did not extend far beyond the line of the present foot-hillH of the Rocky Mountains.' In one important partitular. the conditions in this earlier Cretaceous period dillered from those of the Triassic There was at this time no isolated inland sea, and waters in connection with the main ocean slrct.hed southward to the east of the (jold lians^es as far as the forty-nintli parallel, and beyond it to a further distance which is as yet undeter- mined. This exttnision of the opei\ sea thus actually overlapped, to a considerable extent, the area formerly occupied by the Triassic mediterranean. No important bi'ds of pure limt.;tono or other evidences of deep sea conditions are lound in these earlier Cretaceous beds, and there is on the contrary abundant proof of shallow waters, and (XMuvsioiially of the loiial existence of tracts of low land, upon which a luxuriant vegetation existed, producing in some places important beds of coal. As, moreover, these local terrestrial conditions arc recurrent throtighout a great thickness of strata, it is obvious that the subsidence just referred to was continuous, or nearly so, and was followed pari paxsii by sedimentation. At>out the stage in the Cretacuious which is represtMited by the Dakota group, however, a much more rapid downward movement of the land occurred. This is marked by the occurrence of massive conglomerates, which have now been recognized in many places in the southern part of the interior of British Columbia, as well as westward to the Queen Charlotte Islands, northward to the upper waters of the Yukon and Porcupine, and east- ward along the line of the Rocky Mountains. This subsidence was not only more rapid but also more extensive than that which had previoixsly been in progress. It was at this time that the open sea first spread across the entire area of the Great Plains, overlap- ping even the western Hanks of the Laurentian plateau, and llowing southward as a wide strait to the Gulf of Mexico. - At or about this stage of the Cretaceous, important evidences of renewed volcanic ac^tivity are found both in the Queen Charlotte Islands and in the Rocky Mountain Range, and also probably in several places in the vicinity of the Coast Ranges of British Columbia. ' The central part of the Cordillera belt, however, for a great part of its entire length, still formed a more or less continuous laud-barrier between this great strait and the western ocean. It is only in the extreme north-west that the waters of these two areas are known to have been freely in communication across it, and there is next found a con- siderable body of evidence (though for the most part of a negative character), ' to show ' Seo map No. 1. Tliis and tlie otlier maps attempt only to sliow tlie conditions at tlie times to whicli tliey are roferroti, in the most general way, no sufflcient data having yet been obtained, for detailed delineation. '' For further detail see a paper by the writer on the " Earlier Cretaceous rocks of the North-western portion of the Dominion of Canada." ' American Journal of Science,' III, vol. ixxviii, p. 120. ' ' Report of Progress Geol. Surv. Clan.,' 1878-79, p. (iO B. ; ' Annual Report Geol. Surv. Can.,' 1885, p. 160 B. ; ' Amer. Journ. Science,' III, vol. xxxviii, p. 127 ; ' Geological Magazine,' Dec. II, vol. viii, p. 218. ' Consisting chielly in the absence of tlio next succeeding beds of the Cretaceous in the area outlined below. Evidence of folding and erosion at about this time has been found to the south in the Western States. Cf. " 8. F. Emmona," ' Bull Geol. Soc. Am.,' vol. i., pp. 276-279. OF TUK JIOCKY MOUNTAIN IllKilON IN (^ANAIM. 9 that not ve.-y lons^ after the great HTibsidenoe, the barrier above rfforrinl to vvaH still I'urther elcviitod, and inereased in width to the westward, till it included almost the entire area of the southern mainland ol' British Columbia, together with the southern part of Van- couver Island and the Pu get Sound i 'rion. ' It is I'urther elear, that while subsideme continued, or the elfect of the main sulisidence had not yet been reversed in the eastern portion of the area of the Canadian (ireat I'laiiis, that toward the w st and uorth-west of the area now occupied by t' ■ e plains, eitlier some local and irregular el-vation of portions of the sea-bed occurred, or sedimentation went on so rapidly as to more or less completely fill portions of the great Creiacpous strait. This is shown by the occurrence of estuarine, lacustrine and terr'strinl conditions in the HcUy River and Dunvegau series of diiferent parts of that region. Next, ns evidenced by the ubiquitous occurrence to the east of the Cordillera of the marine Pierre shal<«s, there happened a further subsidence, which continued progressively or stage by stag;? to and throughout the Liuamie or closing period of the Cretaceous. Concurrently with these later Cretaceous conditions to the east of the Cordillera, it appears that on what now constitutes the littoral of British Columbia, to the west of the axis of the Coast Ranges, the area of Cretaceous sedimentation was transgressively extending to the southward, the local base of the Cretaceous being found at successively higher stages in the system in that direction, till at a time which is believed to have corresponded with the Laramie of the plains, the sea invaded the Puget Sound region.- Thus, at the closing epoch of the Cretaceous period,' on the western margin of the Cordilleran region, massive accumulations of strata of a delta-like character were in pro- cess of formation in the Puget Sound region, ' and very probably also as far as the northern end of the i^resent Strait of Georgia; the land being along this part of the littoral some- what lower relatively to the sea-level than now, and the source of the detritus brought down by rivers being found in the elevated interior of British Columbia. ' To the north, in aboxit latitude 61°, in the central region of the Cordillera belt, whore the upper waters of the Yukon are now found at an elevation of 2,000 to 3,000 feet, evidence is again obtained of the presence of the sea or of an estuarine body of water at approximately the level of the ocean, "o the east of the Cordillera, in lacustrine basins of great extent, which w^ere in more or less free connection with the sea, and extending nearly to the line of the present Rocky Mouu.'dn Ranges, if not in places overlapping this line, the latest beds of the Laramie were in course of deposition. There is, further, reason to believe that at this time a nascent Rocky Mountain Range was in part marked out by low hills, in which some of the Paloeozoic and Triassic bods were already subject to subaerial denudation, and that rivers rising on the flanks of the Gold Ranges and flowing across the line of the present Rocky Mountains may have supplied a portion of the material of the Laramie of the North Saskatchewan district, while further south a great part of the material came directly from the prototypal representatives of ' Cf. for tlie CO- responding soutliern region, S. F. Emmons, ' Bull. Geol. Soc. Am.,' vol. i, p. 278. -' Cf. ' American Journal of Science,' III., vol. xxxix., p. 182. ' The Laramie is, for ttie purposes of the present discussion, assumed to represent the highest formation of the Cretaceous, as, though ita flora leads to the belief that it comprises passage beds to the earliest Tertiary, it is pliysically attached to the Cretaceous rocks- • Dr. C. A. White, ' Bulletin, United States Geol. Surv.' No- 51. ' ' American Journal of Science,' III, vol. xxxix, p. 183. Sec. IV, 1890. D 2. 10 G. M. DAWSON ON THE PIIYSIOGEAPHICAL GEOLOGY the Rocky Mountains proper. ' An area including the entire inland portion of British Columbia, whit^h is marked ont by the; borders of these regions of deposition, has afforded no eviden(M' of Laramie rocks, and there is every reason to believe that it constituted a somewhat elevated tract of land at this time. - This state of affairs was brought to a close by another of the recurrent epochs of folding and dislocation of the earth's crust, which was one of the greatest of those of which we find the results in the region under discussion, as well as the last of an im|> 'vtant character to which it was subjected. Under the influence of enormous pressure acting from the Pacific side, the strata, till then nearly horizontal, which bordered the Gold Ranges on the north-east, were folded together and thrown up into a dominant ridge ot Alps, which finally outlined the Cordillerau belt on this side. A similar folding and upthrust allVctcd also the western marginal mountains which have been referred to as the Vancouver system, but the action was there probably less violent and certainly affected a narrower zone. A portion of the crumpling to which the rocks of till- Coast llMnge have been subjected was doubtless also produced at or about the same time, and certain granitic extrusions which cut the earlier Cretaceous rocks on its easti'rn flunks, as well as much of the flexure of these Cretaceous rocks, are further attributed to this period of disturbance. There is really no means of ascertaining what effect this dynamic movement pro- da -ed in the region of the Gold Ranges, but it is more than probable that the whole width of the Cordillera then suffered change and d 'formation of such a character that little if any trace of its surface contour of an older date can be found to-day.'' It does not, however, necessarily follow that the general altitude of the Cordillera belt was at this time greatly changed. The greater part oftbe accumulated pressure appears to have been relieved by folding along the lines of its two bordering ranges, and it seems to be not improbable, as a general proposition, that changes in elevation affecting wide areas are due to other caiises than those producing mountain ranges.' "We are warranted in assum- ing, however, that a certain movement in elevation was ( oincident or nearly so with that of the great disturbances above outlined, as no strata representative of the Eocene period proper have yet been Ibund anywhere in the western part of Canada. The entire area of the Great Plains was thus sufficiently elevated to become dry land, as occurred at the same time in the Western States to the south of the international boundary.'" ' Cf. 'Report of Protrress Geol. Snrv. Can.,' 1882-84, p. 113 C; 'Annual Report Geol. Siirv.Can.,' l«s« p. 135 E. '' It niiiy bo noted, that to tlie south of the international boundary, in Wasliington, Dr. C- A. Wiute and Mr. Bailey Willis find reason to suggest tliat certain beds occurring to the east of tlie Cascade Mountains, are of the same nge as those of the Pugct Group.— ' Bulletin, U. S. Geo!. Surv.,' No. 51, p. 84. Bee also S. F. Emmons, ' Bull. Geol. Soc. Am.,' vol. i, p. 282. ''In respect to this great epoch of orographic movement, as evidenced particularly in the more southern part of the Cordillera, which has now been somewhat closely studiel, Mr. S. F. Hinmons may be quoted as follows :— " It is unquestionably one of the most important events in the orographic history of the entire Cordilleran system. With the exception of the great unconformity between the Arclucan and all overlying sediments, which is a phenomenon mi gemris and altogetlier exceptional, no movement has left such definite evidence as that which follows the deposition of ihe coal-bearing rocks, to which the name Laramie has by univereal consent been applied." — ' Bulletin Geol. Soc. Amer.,' vol. i, p. 285. * Cf. LeConte, ' American .Fournal of Science,' III, vol.xxxii, p. 178. » It should be mentioned, however, as exceptio.ns to this general statement, that a small occurrence of beds believed to represent the Green River Eocene, has been noted by Dr. C. A. White near the forty-seventh parallel OF TIIR ROCKY MOUNTAIN RRdlON TN CANADA. 11 In the fortieth parallel ivgion, during the liooone period, many thousand feet of beds holding- characteristic fossils wore laid down in a series of lakes ix^tween the "Roc-ky Mountains and the Sierra Nevada, but no such deposits have been met with in any partol the northern Cordillera. The whole sweep of country from the Laurentian region, possibly quite to the now submerged edge of the Continental Plateau on the Pacific side, thus became, and continued to be throughout thi; earliest Tertiary, an area of denudation, within which, if any small areas of deposition o<'curred, the beds formed in these have either been subsequently removed or have become concealed by later deposits. For thi;- iirst p(>riod of th(> Tertiary proper, intdudcd b(>tween the mountain-forming epoch which closed the Laramie and the oldest beds of the Miocene or middle Tertiary of the West, we must therefore endeavour to discover traces of another kind, viz., those impressed on the land by sub-aerial waste and the erosion of rivers. Evidence of very prolonged and important action of this nature is believed to exist. It is quite probable as a consequence of the greater elevation of the land and the ridging up of the central parts of the Cordillera brought about by the great post-Laramie dynamic move- ment, that the excavation of the remarkable system of valleys began, which now in a partially .submerged condition exist as fiords along the western margin of this part of the continent. No evidence can, howev(>r, yet be advanced on this point. The main result, which there is good reason to refer to this time of denudation, in that part ot British Columbia which lies between th(> Coast and Gold Ranges, is the formation of a first Interim- Plateau or plain,' of whii>h (Extensive, though now more or less disconnected fragments still exist. These r(^mnants of an old Eocene land-surface are most marked in the southern half of the Interior Plateau of to-day, where they now have elevations of from 4,000 to 0,000 feet above the actual sea-lev(d, and stand ;},000 to .5,000 feet above adjacent valleys and low tracts of later origin. The highest portions of this older plateau are found attached to the Hanks of the bordering mountain ranges,- and its general elevation decreases toward the central line of the present Interior Plateau, as well as gradually in the direction of that line to the north-west, till, in its northern part, its average elevation is not more than about 3,000 feet, and the actual valleys are cut much less deeply into its surface. Climbing to the level of this old plateau, or to that of some slightly more elevated point about the fifti(>th or fifty-first parallel of latitude, the deep valleys of modern rivers with other low tracts are lost fight of, and the eye appears to range across an unbroken or but slightly diversified pl;^ 'e-^v«-^«^<^».. 12 (i. M. DAW,SO>^ ON TIIK PIIVSIOGR/VPIirCAL GROLOGV /^ '^*-'\'\^ _ Mosozoic rocks in Iho country about Deaso, Franc(»s and Finlayson Lakes, now drained l)y streams at lower levels which discharije to the Paciiic, or to the Mackenzie and the Yukon.' This has an elevation al)ove the present sea-level of 2,900 to 3,200 feet, and may with considerable probability be assigned to this time. During the same period the great Columbia-Kootanie Valley which separates the Gold and Rocky Mountain Ranges (and to which reference will auain be made) as well as the Flathead Valley, in both of which deposits believed to be Mioc(»ne occur, must haAC betm cut out by riAans flowing to the southward.- It is further probable^ that many more of the larger A'allt^ys of the more mountainous tracts also date from the l<;ocene period, even when th(>se ar(> actually known to contain Miocene beds. ' It must be explained here, in advance of such remarks as aamH subsequently be made r.'specting the Miocene period, and to meet objections which may possibly occur to the read(>r as to the permanence as a tiible-land of a denudation-suri'ace of such great age, that although the beds attributed to the Miocene are oiten found to be inclined at consid area of this old peneplain is now capped by basaltic and other accumulations of the Miocene, such capping is, over at least eight-tenths of its area, too inconsiderable in thickness to be regarded as of importance in its bearing on thi> general question. A plain like that of which the remnants are here found, and based as it is in various places upon granites and folded and distiirbed FaltTiozoic and earlier Mesozoic rocks for the most part highly indui'ated (though varying considerably in this respect), can only haA'e been produced, as I conc(>ive, in tAVO possible Avays, viz.. (1) as a plain of marine denudation, or (2) by a system of streams and riv(>rs cutting down and working over a land-surface during a vast lapse of time, and under conditions of great stability, till that surfiice has become throughout approximately worn doAvn to the base-level of erosion. In this case, the first method is inapplicable, on account of the confined character of the area affected, Avhich because of its bordering mountain ranges could have been in no direction open to the ocean, and is further negatived by the absence of strata, whether marine or otherwise, not only in the area itself, but also in the region to the east and along the border of the Pacific to the west of it, referable to the period in which the plateau is believed to have been formed. It is therefore to the second cause that the production of this early Tertiary peneplain must be attributed. The circumstances which resulted in the production of this old nearly level surface I believe to have been as follows :— At the close of the great post-Laramie disturbance, the form of the surface of the Cordillera was such as to give rise to a riA'er system, the limits of whose drainage basin nearly coincided with those of the Interior Plateau of to-day. It would be unsafe, from the existing slopes of the remaining parts of the old surface, to draw any conclusion as to the direction of outflow of this riA'er-system, but the occurrence ' Annual Reiwrt Geol. Surv. Can.,' 1887-88, p. 109 B. ' Annual Report Geol. 8nr\'. Can.,' 1886, p. SO B. OF THK ROCKY MOUNTAIN REGION IN CANADA. 13 of a high mountain barrier near the forty-ninth parallel tends to show that it discharged in a northerly direction. Whether its drainage issued eventually to Ihe west, across the line of the Coast llanges, or to the east across that of the Rocky Mountains, there is no available evidence to show. Little by little this river and its tributary streams, aided by other stibai'rial agencies, cut down almost its entire drainage-basin, till this became a nearly uniform plain, with some slight slope in the main direction of the river's How ; but of which the lowest part approximately coincided with th(> sea-level of the time. The lower part of this plain having to-day an elevation of about 3,000 feet above the sea, in- dicates that at the time at which this long continned action came to an end, the level of the central zone of the Cordillera stood lower by about that amount than it now does. After reaching this base-level of erosion, the rivers would of course be unable to do more than serve as channels for the conveyance of material brought into them from the sur- rounding country, which, wherever it stood above the general level, was still subject to waste. The valleys b:>came wide and shallow, and the surface as a whole assumed per- manent characters. Round the borders of the region in which a peneplain was thus formed, higher parts of the original surface remained, though in much reduced form, as rugged mountain- partings between this river-system and others on the east, south and west. It is possible that some movement in elevation still continued along these mountain ax(>s, which did not alTect the Interior Plateau region, or affected it but little. This, however, is merely a conjecture which may help to explain the dominance which these mountains evidently continued to maintain. Having reached the relatively stable condition of a peneplain, approximately at the base-level of erosion, the surface of what is now the Interior Plateau region may have con- tinued to exist for a long time without notable farther change ; but this state of afTairs was eventually brought to a close by some further orographic movement. According to Mr. King,' several such movements took place in the Cordilleran region of the fortieth parallel during the Eocene, and one of particular importance is attributed to the close of this period. There is, in British Columbia, some reason to suspect that a considerable amount of erosion of the surface of the I'Jocene peneplain occurred before the initiation of the deposi- tion of the Miocene. This might be explained by a moderate elevation effected ^ fore the close of the lilocehe; but it was probably by a more important movement, which took the form of a re-elevation of the mountain axes, and which may have been synchronous or nearly so with King's post-Eocene disturbance, that the Miocene lacustrine conditions were brought about. By an interruption of the drainage produced in some such way, the great Miocene lakes of that portion of British Columbia between the Coast and Gold Ranges were first formed. As to whether the disturbance referred to was accompanied by faulting along the bases of these bordering' ranges, or by any considerable increase in the elevation of the Cordillera as a whole, no means has yet been found of determining ; but the character of the Miocene flora seems to indicate that the land did not stand at this time at any very great height, and that the climate was temperate. In whatever manner produced, there is ample and good evidence respecting the condition of the entire region here treated of in the earlier part of the Miocene. The >■-' U .^. ' " Geological Exploration of the Fortieth Parallel," vol. i, p. 755. 14 G. M. DAWSON ON TIIK PirVSlOfillAPlTICAL GEOLOOY 7;.' ^'y • ' Interior riiitoau benimc tiio sito of a groat lako, or more probably of a sories of lakes of Cfrt'utt'r or less diint'ii.sions, Rome of which wt-ro of oarli«r date than others. That tho n-irion was not siinultanconsly covorotl l)y a sini-li' lake seems prol)able from the diverse lithologieal rharacters of the sediments met with, as well as from Ihe ditleront facies of the contained fossils, whir vegetable accumulations, now found as lignites and coals, were produced ; insects of species long since extinct, sported in the sun ; and land animals doubtless roamed through the forests, though in this particular part of the Cordillera no relics of these have yet been di.scovered. On the littoral, to the west of the Coast Ranges, bods are found in a number of places which have been referred, on evidenci^ more or les.s complete, to the same period ; some being marine and some of lacustrine origin. In the Western States, tho Pah-ute Lake of King was about this time formed to the east of the Sierra Nevada, while to tho east of the Rocky Mountains the Clreat Plains sullered a depression, deepest along tho Rocky Mouiitain foot-hills, which produced the Sioux Lake of the same author. The depression which gave rise to this last-miMitioned lake probably extended northw^ard into the Canadim plains, and was there greatest along a line running north-westward, parallel to the Cordillera, but at a distance of nearly 200 miles from its base. In the Cypress Hills and Hand Hills, outliers of Miocene rocks are found, upon which tho statement hero made is l)ased, and these contain mammalian remains in some abundance, which are according to Prof. Cope referable to tlie White River period.' It is important to note that the beds here met with consist largely of conglomerates, the well-rounded pebbles of which have been derived from the harder beds of the Roi"ky Mountains, as tho occixrrence of these coarse materials at such a distance from their source implies the existence at the timo of a considerable gradient from the base of the mountains to this line of greatest depression, a gradient such as to produce rapid rivers with great powers of transport, and involves besides tho recognition of the fact that the western margin of the plains stood at a relatively high elevation. Had the orographic conditions remained permanent for a prolonged period, tho lacustrine phase of the Interior Plateau of British Columbia would have boon in the end terminated by tho filling up of some lake-basins, and tho drainage, by tho gradual cutting ! ' ' Annual Report Qeol. Sarv. Can.,' 1886, pp. 68 C, 79 C ; ' Annual Report Geol. Surv. Can.,' 18SC, p. 138 £ ; 'Ameriran Naturalist,' vol. xix, p. 103, OP THE ROCKY MOUNTAIN REGION IN CANADA. 18 down of the beds of the effluent rivers, of others. Before this had taken place, however, a new factor appeared. Volcanic action was rocoramencod ou a great sv;au> ; fragmental volcanic ejectamonta, supplied from numerous vents, wore discharged and mingled them- selves with the ordinary detrital deposits of the lakes, while basaltic and other lavas flowed out in great vohime over large parts of the plateau country. The principal centres of this volcanic action appear to have been aligned near the eastern or inner base of the Coast Ranges, where the more or less degraded bases of some of the old volcanic vents may still be traced." Isolated patches of volcanic material, which are probably due to the same period, are found, however, to occur far north, on the Stikine and in the Upper Yukon basin, as well as along the littoral of what is now the Province of British Columbia. It is probable that we may refer the volcanic activity of which evidence is afforded in one place near the southern boundary of the Canadian Great Plains to the same period. The Three Buttes, or Sweet Grass Hills, there rise as isolated monuments of such action, standing at the present day, by reason of the superior induration of their rocks, high above the level Cretaceous plain. These peaks are evidently now but the remnants of the necks of volcanic mountains, of which the history has not yet been traced in detail.-' "We shall have occasion to allude to them again in a later part of this paper (pp. 60, Ql.) Prof. LeCoute believes that the great lava-flow of Middle California occurred near the end of the Pliocene, while similar irruptions to the north, in Oregon, took place somewhat earlier, at the close of the Miocene or early in the Pliocene.' The last named region closely corresponds with the Interior Plateau of British Columbia, but in British Columbia the evidences of the blending of the Miocene lake deposits with those due to volcanic action, are in some places so distinct, that we are amply justified in attributing at least the first eruptions to the Miocene period itself. That volcanic action may have continued into the Plioceiie period can not, however, be denied, particularly in view of the impossibility, in the absence of physical changes, of drawing any perfectly distinct line between the various subdivisions of the Tertiary. All that can be said at present is, that there is no known proof of Pliocene volcanic activity in British Columbia.^ Taken as a whole therefore, the Miocene period, in regard to the Interior Plateau of British Columbia, may be described as one of deposition and accumulation of material. The then lower parts of the old Eocene peneplain were in the first place partially filled with lacustrine sediments, in the second more or less uniformly overspread by lava-flows or other volcanic accumulations. This levelling-up process was evidently so complete as entirely to obliterate the valleys of the Eocene and Miocene river-systems, and it is doubtful whether any one of these valleys has since been re-excavated along its old course for any considerable distance. This, however, applies only to the area of the Interior Plateau. Valleys formed in the more mountainous parts of the Cordillera have doubtless, in many ' ' Reportof Progress Geol. Surv. Can,' 1876-77, p. 75. Further, and as yet nnpublisheJ, evidence, on tliis point has since been found in the soutliern part of British Columbia. '' Cf. ' Report of Progress Geol. Surv. Can.,' 1882-84, p. 45 C. ■' 'American Journal of Science,' III, vol. xix, p. 189 ; vol. xxxii, p. 177. * Evidence is found, however, in some places, of whioii Pavilion Mountain and the upper part of IlatCre-k may be specially mentioned, such ns to show the existence of lakes or ponds of limited size in intervals of the period of volcanic eruptions. Some of these may yet afford organic remains of later date than the Pliocene. The recurrence of volcanic phenomena during a considerable length of time is also shown by the cutting out of valleys in the basalts and tlie refilling of these by later basalts in the Stikine region. See ' Annual Report Geol. Surv. Can.,' 1887- 88, p. 72 B. -'.-/ /.''' 16 G. M. l)A\V.S()N ()X THH I'HYSIOGRAPIIICAL GEOLOGY ciist's, been siuco porpi'tuati'd, aud tho drainage marked owt at this time has in such instances been maintained. In the southern extension of the Cordilleran region, important orographic changes oi various kinds, which were probaljly synchronous or nearly so with the epoch of volcanic eruption al)ove referred to, took place at the close of the Miocene period. At this time the Coast Ranges of California were produced, the beds of the I'ah-ute Miocene Lake were thrown into gentle folds, the area of the Great Basin sank, and became the site of an extensive lake. Th.' plateau region of Utah was further elevated by an amount of from 2,000 to 3,000 feet, and a renewed gradual subsidence affected the Great Plains to the east of the Cordillera.' These movements appear to have been but faintly rellected in the region to the north of the forty-ninth parallel, but it is in all probability to this time that we must attribute the local iolding of the Miocene rocks, which has been referred to on a previous page. There is, however, no valid evidence of any considerable change either by elevation or depression of the region of the Interior Plateau, where, at or shortly after the time at whitdi the volcanic forces had exhausted themselves, the drainage began to outline a new system of stream- and river-valleys. Apart from probable local changes in level, which may alone have been suflicient to prevent the drainage from resuming its old direction, the filling up of the former valleys had been so complete, that the new streams, following the inclinations of the actual surface, began to cut their beds in courses entirely ditferent from those of the old. Local accumulations of volcanic material, together with the slight folding to which the Miocene beds, whether volcanic or otherwise, had been subjected, must have caxised the area of the old Eocene peneplain to be less uniform than before ; and denudation a( ting on its projecting parts began again to reduce these toward the general level. This action was probably long continued, and had important efTects in bringing the Interior Plateait region again down nearly to the base-level of erosion. The streams, which evidently flowed at this time with low gradients, cut out wide, shallow, trough-like valleys, which may yet be found in many places, and are at the present time in some instances still occupied by existing streams, though in others interrupted or abandoned owing to later changes, some of which will shortly be alluded to. These early post-Miocene valleys, where they may yet be studied, present all the characters of a drainage system which had been long maintained under stable conditions of the surface. They are found at the present day pi;rsuing sinuous courses over the surface of the old table-land, and may also be recognized in many oases even in the mountain regions, where they appear as wide U-shaped valleys, in the bottoms of which winding streams pursue a comparatively tranquil course, and in which lakes, produced in various ways, are not uncommon. At a later date in the Pliocene or last period of the Tertiary, on which w^e have now entered, it is evident that a very considerable and general elevation of this part of the Cordilleran region occurred. During the first stage of the Pliocene, the Cordillera must have been lower relatively to the sea-level than at present; when this elevation took place, it became considerably higher than it is now. The gradients of all the rivers were thereby increased, and being thus armed with new powders of erosion, the streams began to cut deep and, at first, narrow channels. To this time the cutting out of the greater part of ' King, "Geological Survey of the Fortieth Parallel," vol. i, p. 756; leConte, ' American Journal of Science,' III, vol. xxxii, p. 177 ; Dutton, " Tertiary History of the Grand Caaon District," p. 226. 1 ] OF TUE liUCKY MOUNTAIN JiliGION IN CANADA. 17 the deep valleys, which now oxist in a submerged statt; as the fiords of thi^ coast, is with all probability assigned. While admitting that these fiords may have been to some extent shaped and enlarged locally by ice during the Glacial period which ibllowed, their depth is such as I believe to show that at the main period oi' their formation, in the I'lioceue, the land stood relatively to the Pacific about 900 feet higher than it now does. This later Pliocene elevation doubtless produced many important changes in the drainage system of the country, and particularly in that of the Interior Plateau region. Certain rivers which were fed from the perennial snows of extensive mountainous districts were then enabled, by the increased gradients given to them, to cut down so rapidly, compared to other streams which were less copiously supplied, as to achieve and maintain a dominant position, and even to completely rol) some of the old valleys of their waters. It was doubtless at this time that the Fraser and its great tributary the Thompson managed to extend their drainage areas over so large a part of British Columbia. The Fraser may be described as running round the south end of the Coast liange of British Columbia, or, if regarded more strictly as cutting through the southern end of this mountain-system, it does so at a point of decreased elevation and width. Having in consequence been enabled in a compav I lively short time to greatly deepen the lower part of its present course, this river next aicw to itself the drainage of a great part of the northern and eastern portions of the Interior Plateau. Between the north and south part of the present Fraser valley and the inner borders of the Coast Ranges, from latitude 50° to latitude 54°, a broad belt of country seems at an earlier time to have been drained by several smaller rivers, each of which had cut for itself a valley completely through the Coast Ranges ; but these rivers appear to have been gradually deprived of their upper tributary waters by the iixtension of the western tributaries of the Fraser, till, in the case of the Homathco, and to a greater or less extent in that of other rivers, we lind at the present day a very small stream making boldly toward the mountain range, and still carrying its waters through it in deep canons to the sea, though apparently quite incapable of itself excavating a gorge such as that which it now occupies. To the long continiied denudation and erosion which wen; in progress during the Pliocene, the greater part of the deep pre-glacial auriferous gravels, of valleys such as those of the Cariboo Mountains, must be assigned, though it is at the same time possible tliat much of the excavation of these deep mountain valleys may have been more or less con- tinuously in progress since the date of th- post-Laramie disturbance. In the Interior Plateau region no deep auriferous gravels have yet been found, though it is more than probable that if the now concealed beds of the Eocene drainage system ot this district could be discovered and ibllowed, they also would be found to be characterized by important deposits of the precious metal. It would be beyond the scope of the present sketch to endeavour to catalogue and describe the various older Pliocene river valleys which have afforded the evid(>nce upon which the statements just made are largely based; but as the observations on these have not yet been published, it may be well to enumerate a few of them, and to point out the influence of the later Pliocene elevation of the land upon them. This influence is apparent chiefly in the formation of deep narrow gorges and canons iii the parts of these older and wider valleys which debouch upon the larger and lower river valleys, such as that of the Fraser, the Thompson and the Columbia. To keep pace with the rapidly Sec. IV, 1890. 3. fl. ^'U-v 18 (i. M. KAWSdN on TIIK PlIYSlOGHAl'llICAL (iKOLOtiY /- «l(!cpoiiini^ valleys of these maiii rivers, the streams of the tributary valleys were foroed to cut down very fast toward their mouths, ami this futtini^" down has gradually retro- gressed (ill mnny cases I'or a eoiisiderable distance) along the line of the old valley. A few instances only of observed facts in this connection, derived i'rom the southern part of the Interior Plateau region in the vicinity of the Thompson and Fraser Ilivers, will here be cited. The heights placed after the names of streams in the subjoined (al)ular statement are those of the lower remaining parts of the older Pliocene valleys, below whieh, in each case, the present stream enters a narrow, steep, gorge-like valley, by which it descends to one of the large rivers or to an adjacent deep tributary valley of one of the.se rivers. While the main features of difference above noted as existing between the older and later Pliocene valleys are well marked, it is of course difficult in most I'ases to state the exa. t point at which the older valley should be considered to give place to the newer, and impossible to eliminate completely the inlluence of still later changes ol' greater or less importance.' The heights given must therefore be con- sidered as approximate only. They are, however, it is believed, sulhcieut to illustrate and cojilirm the point in question : — (Meadow Creek),' (iiipp(ir valley) 3,350 foet (I'likaist C're«k), " " 3,750 " (Witches' P.rook) " " 3,G50 " Three-Lake Valley (soutli part) 2,750 " Kelly Lake Creek (upper valley) H,300 " (Yicho Creek " '• 2,650 " (La-loo-wissin Creek), (upper valley) 3,450 " Pavilion Creek 2,300 " The difhculty above alluded to of eliminating the inlluence of other causes, both of earlier and later date than those of the Pliocene, renders it impossible to take the actually lowest instance of these or other similar wide low-grade valleys, as a definite point from which to measure the depth of the later Pliocene erosion. Ev striking an averao'e of the height of the lower existing parts of a number of these old valleys, however, we may arrive at a rough approximation to the depth of this erosion, in that part of the region here particularly referred to. Such an average gives a height of about 3,200 feet above the present sea- level, while the mean level of adjacent parts of the Fraser and Thompson Rivers is about 130 feet. As it will be shown in the second part of this paper that these large rivers have not materially, if at all, reduced their general level since the .;lose of the Pliocene, the dilference between (he two figures above given may be taken as roughly indicating the amount ol' cutting-down whi.ii they accomplished subsequent to, and as the result of, the later Pliocene movement of elevation, this dilFereuce being about 2,470 feet. The elevation of the land in this district, as compared with that in the earlier Pliocene lime, must therefore be assumed to have considerably exceeded that amount. Good illustration of the point here specially referred to, though in a more mountainous region, may be seen on the line of the Canadian Pacific Kailway, in the Rocky Mountains ' Among which may be nieutioued the partial infilling' of many of these valleys with drift during the Glacial period. - The names enclosed in parentheses have not yet apjKjared on jjublished maps of the region. As the puriwse here in view is nierc(ly that of ilhistration, it is considered unnecessary to explain in detail the positions of these streams. ■■i ,. OF TJIE llOCKY MOUNTAIN REGION IN CANADA. 16 and Solkirk Rango.' In most of thoso valleys the wide and relatively level upp'^r /'J^.u — ^ portions contrast markedly with their lower canons and gorges. The Kickiiig-IIorse, Upper Kootanie and Klk Valleys, while still wide, and above the narrow steep portions by which their rivers reach the adjacent part of the Columl)ia, have elevations of almut 3,200 feet above the actual sea-level, and etlei-t therefrom a rapid u.-scent through newer and narrower valleys of from 700 to 000 feet to the Columbia. These dillerences are much less than previously quoted, but it must be remembered that the localities referred to are much further inland, particularly if their distanci> from the coast be measured along the sinuous line of the Columbia River and not in a direct line from the sea. This, however, need cause no surprise, as even on the assumption ol an equal elevation of the whole breadth of the Cordillera in the later I'liocene, the iulluem;e of such elevation on the upper parts of the larger rivers must have made itself manifest only after a long time, and must moreover, even if it was allowed time to produce its full result, have been much less etiective at such a distanct; from the coast. The best instance which has yet been studi«'d of the relation now obtaining between the early Pliocene low-grade valleys and the more modern and deeper erosions which occurred during the later Pliocene period of greater elevation, is that afforded by what wo may call the " Old Cache Creek Valley." As I am enabled to present a map of this valley and its vicinity, a few words in explanation of its charactin- may be admissible, as a conclusion to this first part of the general enquiry here made. I must, howan^er, before going further, note that a detailed contoured map would much more adequately represent the actual conditions than that which is here given, on whiih the modi' of delineation adopted does not fully expr<'ss the general plateau-like charact(>r of the region depicted, but makes it appear more moun- tainous than it really is. | The main line of the Old Cache Creek Valley in question, I have been able to trace out and to travel along for a total distance of about thirty eight miles. Its general course across the now broken surftice of this part of the Interior Plateau is nearly east and west, an approximate parallelism being maintained to the direction of the corresponding part of the Thompson Valley. The total dilference of level between the upper or eastern end of the remaining part of the valley and its western determinable end is approximately .'i-SO feet, giving a slope of nearly fourteen feet to the mile in the direction in which its water originally flowed. The amount of the fall thus indicated, being nearly that which might be expected from the character of the old valley, in itself atlords a valuable indica- tion of the absence of any considerable differential elevation in this part of the Int(>rior Plateau since early Pliocene times. Standing on some prominent point of the plateau, one may trace the wide trough-like form of this old valley for many miles, tr^-nohing th(> surface of the plateau and pursuing a nearly direct westward coursi>. The stream by which this notable valley was formed has, however, been unable to perpetuate itself. The history of its decline and interruption I conceive to be as follows : — " I am aware that the orourrence of sncli U-8hai>e(l and wide valleys in mountain resiions lias often been attributed totlie action of glaniera in such valleys. Tlii.i explanation docs not, however, re.iniro to he resorlod to here, nor am 1 prepared to admit tliat it in a valid one. 20 (1. M. DAWSOX ON TIIK I'll VSKMlKAnilCAL OKOLOdV Th.'Htr.'iim l)y \vhi pink line. It will be noted that the small modern stream kuowu as Cilcho Creek has, following the liiu' of the old valley, cut back for a certain distance by a narrow recent gorge, l)ut long before it could deepen the entire length of the old valley, other streams had largely deprivi'd it of its erosive power by robbing it of its upper waters. Of these Eight-mile Creek, Deadmau Iliver and the Tranquille River and its tributaries are the chief oll'enders. It will further be observed that the branch of the Deadmau known as Criss Creek is supposed to coincide for a considerable part of its length with a principal northern tributary of Old Cache Creek, which it has now for a number of miles cut down into the ibrm of a nearly V«baped and later valley. The course of the Trauquille is also identical for some miles with that of the old valley. With reference to the hver, the old valley has been largely choked by drift deposits, duo . to a still later period, from which it has since been only partially cleared by denudation, the various heights given can be accepted only as approximately indicating the actual level of the several parts of the old valley-bottom. From what has been stated in the foregoing pages, it will be evident that through- out the whole Cordilleran region of British Columbia, the prolonged era of denudation of the Pliocene produced very important results. The rivers and streams acted in the first jilace for a long time during a relatively low stage of at least the greater part of the land- Burface, and subsequently for a similar period at a high stage. During this latter many of the first-formed valleys wore greatly deepened in their courses as originally outlined, while others were robbed of their drainage and now valleys wen; produced with other 01' TIIK KOCKY MOUNTAIN UKUION IN CANADA. 21 directions. Those rcsiilts of »h« riioci'uo iTosion arc of special iiiiportiimo from ii gt'ographical point of view, as tho whole oircct of 8ul)si'(|Ut'nt cvfiitH can he .sliown to have been relatively in.sigiiilicant, and the main features of tlu- «riod. The prim ipal rivers and streams still, with small exi-eptions, occupy the same valleys, and since the Glacial epoch (as subsecjuently noted) many of them have been unable to ss llowed round the south end of Vancouver Island, receiving important tributaries from the valley now occupied by the Gulf of Georgia and from the region of Puget Sound.' A second river of some importance, carrying the drainage of the Coast Ranges to the north of Bute Inlet, probably discharged its waters to the north of Vancouver Island, while between the Queen Charlotte Islands and the mainland a wide plain must have been formed. The Great Plains, to the east of the Cordilleran belt, are in the main plains of deposition, based for the most part on unaltered and undisturbed Cretaceous and Laramie strata, and only in a very secondary way owe any of their present regularity of surface to planing down by denudation. It has already been stated that, during the earlier portion at least of the Miocene period, deposition was going on over some part of the plains to the north of the forty-ninth parallel, as well as over a very extensive area to the south of that parallel in the Western States. Denudation must, however, have been in progress over other unsubmerged parts of the plains, and, to the north of the forty-ninth parallel, par- ticularly in that steeply-sloping western portion which it has been shown must have existed between the central area of deposit and the corresponding part of the base of the Uocky Mountains. (See p. 14.) In the Western States Mr. King describes the occurrence of a further general sub- sidence of much of the extent of the plains at the close of the Miocene, which resulted in the formation of a vast Pliocene lake, named by him the Cheyenne Lake, th»>. sedi- / z, CJ. Newlwrry in ' Annals of the New Yorli Academy of Science,' vol. iii, p. 265. 22 A\VS()\ ON TIIK I'HVSlOdltAI'lIICAIi (iKOFAXiY iiit'iilN (»r vvhi.h enormously ovi-rlappi'd those of tl»e previous Mioot-no lake iu evt'ry direclioii,' Imt no tniee ol' the extenniim of this lake over the (!ai»ii heeii removed I'rom thejreneral surhiee ol" the plains since the earlier part ol' the ^1 iMie period, and in all prohahility the greater jiart of this denudation occurred duriiijr the I'lioi d.' It would appear, from the facts here referred to, that the later or mid-Pliocene elevatory movement of the Cordilleran region in British Columbia cither did not produce any ell'ect on the adjacent area of the Great Plains, or that if it did so it was that of a correlative subsidence, or was shortly followed by an independent subsidence of a great part of these plains Further south in the Cordillera, Prof. Le Conte finds reason to believe that, at the close of the Tertiary and inaugurating the Pleistocene, ' "V. a (Sool. I'^xp]. Fortiotli Parallel," vol. i, pp. 542, 75(i. ■' ' Anniuil Ileport Gool. Snrv. Can.,' 1885, p. (i!M'. ; KS.SO, p. 78 K. ' Kor details nwiKictinn tlioao liUest I'liocomi or oarly (ilacial DeiMisita, aeo 'Report of Pro<;res.s Geol. S«rv. Can.,' l.S8'.>-8-t, p. HOC. ' Annual Rei>ort (ieol. Surv. Can.,' 18.85, p. 70 C. ; 'Annual Uo|>ort Gool. Snrv. Can.,' 1880, p. i;ttt K OK Til K IIOCKY .MOI'N'I'AIN UKdION IN CANAKA. 23 lilt' I'Icviilioii of thi' Sinrm Nt'vada was toiiHidonildy iiu rcuNcd with llu' produftioii of greater hIoim-s on the west and ol' limiting along the east widti ol' thiw range. The evideni't' ( onsistN ehielly in the existenre ol' gfi'iit river erosion, rel'ern'd l»y him to this date, hut he (|U()teH I'urtlierthe olisorvationH ol' Untton, (Hlliert, Howell and UiisNell to mHow that other important orographic niovenuMitM were Nynehronous, or nearly ho, with the ehivation ol' the Sierra. Thewe compriKe a genoral increase of I he heiglit ol' the (Ireat I'lateau of Utah ol" :!,000 to 4,000 feet and normal faulting in the region of the Great llasin and in southern Oregon, which lias, in the last-named instance, all'ected the later Tertiary lava-beds. ' The (question now occurs as to wliellier these changes can he assumed to have been simultiiiieouH with the iriid-l'liocene elevation in the llritish ('oluinhian portion of the length of the Cordillera. To this it may he replied that the liTeat amount of erosion juet with in Hritish Columbia and referred to the latter half of the I'lioceue period, implying as it does a long interval between tlio mid-Pliocene ehivation there and tbe initiation ol' tlu^ glacial conditions, appears deiinitely to negative such correlation. In British (/oluinbia no facts have yet been met with to prove a renewal of elevation at the close of the I'liooene, but it is at the same time jmssible that such elevation may have occurred, and, in view of the evidence brought forward by LeConte, it may I'ven be said that it is li the Meso-coic and Tertiary eras, we are brought iu contact with phenomena of a uew kind, or at least of a kind of which uo marked evidence had heretofore been found, those, namely, of the period or epoch of glaciation. More than any other changes which the northern hemisphere has undergone, these have iu late years become a battle-grouud >- f rival theories, in consequence of which the most extreme views have bei-u niaiutaine... While it is admitted on all hands that the Glacial epoch presents many problems of great diHiculty, it would appear that much of the dilFereuce of opinion which has been developed has resulted from a disinclination to admit the application to this period of arguments and explanations similar to those which are freely allowed iu other fields of geological reasoniug. It would further appear that the complex character of the evidence met with, and the resulting difficulty found in embracing all the facts under a satisfactory hypo- thesis, seem to have driven many writers to take refuge in extreme theories of a simple character, under which very diverse observations have been included, even though some of these require to be led to a common centre by very devious routes. It is true that in meeting at this stage with evidence of ice action on a great scale, we are called upon to deal with a new agent capable of producing uew effects ; but it is also true that the con- current action of various changes in relative elevation affecting oceanic and atmospheric currents, may have had much to do with the inauguration of the new state of allairs. Whether any still wider changes of a cosmic character co-operated with those above alluded to is a further question having little connection with our present enquiry, and one upon which nothing need be said here. Regarded as the latest in the series of great changes which have been effective in bringing about the actual condition of the region under diccussion, and being in conse- quence one of which the results are still everywhere apparent, the phenomena of this particular time might justify a very detailed mode of treatment and be made the subject of an extensivi aouograph. The object in view is, however, rather that of connecting the phenomena of the Glacial epoch with those of antecedent periods, referring for detail iu respect to observed facts to previous publications. Many of the observations made by the writer in different parts of the Cordillera and in the region to the eastward extending Sec. IV, 1890. 4. 26 (i. M. J)A\VSON ON TllK niYSlOlillAPUlCAL (jEOLOGY to Jvakc Superior, have, been separately published in a number oi' diHereui papers aud reports. Tliis sketch *in;iy, however, allbrd a means of drawing these together aud correlating them in ii manner not heretolbre attempted. In the pul)lications above referred to, it has been made a rule to detail the I'aets wiiich have eonie under my notice in eaih partieular district, and to suggest or discuss t lie deductions and Miore or less theoretical conclusions which these facts appeared to warrant. In the i)reseiit instance this inductive method is necessarily reversed, in order to present the matter under consideraiion with clearness and brevity; this being the only nmde available for a sul)ject which depends on so great an accumulation of observations. With the ol)ject, however, of maintaining as close a connection as possible with the actual l)asis of our knowledge, numeroixs references are given to ("ormer publications, while certain ol)servalions wliich have recently l)een made, but \\ i^ich have not yet beeu published, are accorded a somewhat fuller treatment. 'I'houuh the i)resent paper deals primarily with the Cordilleran region of (_*auada, it has been found essential to refer also to adjiicent regions in tlie foregoing pages ; and in connection with the present stage of the emiuiry it is more than ever necessary to keep in view the conditions at the same time ad'ecting the (Ireat Plains. As, however, the glacia- tion of theCordillera presents in itself a sulliciently complex nroblom, it will be best to consider this alone in the first instance, and isubsequeitly, with the knowledge so obtained as a basis, to treat briefly of the contemporaueor.s phenomena to the eastward. Tliis method is, moreover, rendered appropriate by the circumstance that while the glaciation of these two regious is without doubt correlative, the known evidences of the Ci facial epoch in these two provinces barely overlap. In a region with such pronounced physical featur.s as that of the Cordillera, the solution of the problems offered by the traces remaining to u*. of the Glacial epoch are by no means so simple as in less rugged districts, and it is more than elsewhere necessary to keep clearly in view the chief outlines of the orogi'aphy, as sketched on a previous page. It is thus not possible, without greatly exceeding the limits appropriate iu this paper, to fully i)resent the various local conditions which appeal to the eye iu the licld aud at the time of observation. Tliese circumstances render it difhcult to do full justice to many of tliese ol)servations, a difficulty which is increased by the non-existence of any really accurate detailed maps of large parts of the region. All that is known of the Glacial period goes to show that, relatively to that occupied by the movements and periods of rest which we bavo previously examined, the whole time embraced l)y this epoch was short ; while the results produced were such that, though striking enough becaxise of their recency, they might almost, if not altogether, have beeu overlooked had they occurred at some long previous geological time. For the same reason the character of th(> lata available for the history of this epoch differs considerably from that on which the geologist depends in the case of older deposits. Here we must appeal very largely to the nature and arrangement of incoherent deposits which still occupy the surface, while with regard to water-levels we may often directly consult beaches and terraces which still exist in a condition little changed from that in which they were produced. The evidence brought forward in former pages presents to us at the close of the Plioceue the Cordilleran region at an elevation of at least nine hundred feet above that which it now ■ dal T'"* *M ^'f OF Till) R0(;KY MOl'NTAIN UIOarON IX CAN'ADA. 27 has, and leaves it probable that by a further uplift at tho oiul of the riiocouo period (and marking the close of that period in this region) this amount of clrvalion 'lad been still furthi^r increased. The Great Plains were at the same time at a relatively low level, and had not had impr(>ssed on them that lono- gvntle slope from the l)ase of the Rocky Moun- tains to thj east and north-east which they possess to-day. Asnolhiuii-has been found to show that these two »rt»at areas of uplift and depression were separated bya lin(> of fault- ing- in the vicinity of the eastern margin of the Cordillera, it may hi' ;issumed that a hinge- like Ilexure 0(>curred along- this margin or not far to the (>astward of it.' The Cordilleran region, in consequence of its high elevation, and probably also in part as a result of other nt causes by which the northern heniisph(M-e was affected at th(< inception of tho period of glaciation, appears to have beconn' at this time pn^-emiuently the condenser of the North Tacific. Pro(-ipitation occurred upon it chielly in the form of snow, which was so much in excess of the inlluence of the summer heat as to accumulate from year to year, (ireat glaciers formed in the higher mountains, pr()])al)ly in the lirst instance among those situated nearest to the coast; but eventually tlie greater part of the region becanu^ covered and buried either in ncvc or beneath glacier-'ce. Th(? directions of motion of the glaciers at first produced were doubtless in conformity with that of the valleys of mountain streams, but at a later dat<>, when the Cordillera bertain proportion ol'the ice, howt^ver, during the maximum phase of this great glacier, liowed throvigh passes in the Coast Ranges, and uniting there witli ice derived from the western slopes of these ranges, filled the great valley between Vam-ouver Island and the mainland, impinged upon the shores of the Queen Charlotte Islands, and still farther north reached the ocean across the coast archipelago of the south-eastern coast-strip of Alaska. Though at first in doubt as to the probable origin of the traces met with of this first and most general epoch of the glaciation of the Cordillera,' much additional information gained in later years convinced me that it clearly indicated the former existence of a great glacier-mass such as that here described.' Still more recent observations havt^ proved the north-western movement of the northern part of the great glacier, and ' It is wortli noting liere that Mr. JlcConnoll's carefully elaborated section through the Uocky :Mouutains on the line of tlm Buw and Kickiii;,;-Hoi-.so Rivers allorda niui-ii <,'rouud fur tho belief tliat the central line of this ranije constituted a fieoloK'ical hinge even in PaliiDozoic times. Tlii.s evidence consists chielly in tlio diirorence in character met with in the formations of the western and eastern parts of tlie ran^'e.— ' Annual Ue|xjrt (ieol. Surv. Can.,' 188(), Part D. '' Such general movement probably affected only the central portion of the ico-raa.ss by whicli tho Cordillera was covered, and there is no reason to sui)posc that it wasotherwiso than shigj;ish. ■' 'Quart. .lourn. Geol. Soc.,' vol. xxxiv, p. 118. * ' Quart .Tourn. Geol. Soc.,' vol. xxxvii, p. ^83. 28 G. M. DAWSON OV TITK PIIYSrOGRAPHICAL GEOLOGY iiltor hiiviiig thu8 asoertaiued its area, I ventured to designate it as the QsjulJilfiXaJi The i'orniation luul movomont of the Cordilh'ran glacier naturally resulted in the coiiipl.'t.' oblitcraiion ol' the traces of the earlier and smaller glaciers which must have prec.'dcd it, as well as pvav 'i-- illy the complete removal of all unconsolidated pre-glacial gravels iuid sands. Of these last the only known remnants are the deep gravels of certain old streams in the more mountainous regions, which are in some places distinctly < apped by boulder-.lay, and have proved in the Cariboo District and elsewhere to be highly auriferous. The limits in latitude above assigned to the great neve or gathering-grouud and point of dispersion of the Cordilleran glacier are maximum limits, depending merely on the localities of observed traces of its action, and it is probable that detailed examina- tion of the intermediate region will eventually admit of a much more precise localiza- tion of this area. It is further probable, arguing from the existing conditions of pre- cipitation n'latively to the Pacific, that the point of greatest accumulation was not more than 20U miles inland, while it is quite possible that it may not have been situated over 100 mih's from the coast." On the assumption that a slope of ten feet to a mile is neces- sary in order to produce motion in such a glacier,' and taking into consideration the known length of the south-eastward moving portion of this great glacier-mass, its iiighest c(nitral part witliin the limits above assigned must have had an elevation of at least 7,000 feet ' above the mean elevation of the Interior Plateau, which would be (>quivalent to an elevation of about 10,000 feet above the present sea-level, or probably 11,000 feet above the sea-level of the time. Nothing has yet been ascertained which might throw light on the question as to whether th(> great height of this miue w'as due simply to a local accumulation of ice, or whether its existence and the required degree of slope from it may in part be attributed to a greater amount of uplift w^hich allected the particular region upon which it rested. A brief statement of the known limits of the Cordilleran glacier may now be given. Tn 187S the writer was able to state that if the general glaciation of the interior of lUitish Columbia was attributable to the action of a continent glacier (a point as to which he •WAS at that time uncertain), tht; ice of its southern extremity must have poured southward through the gaps on the forty-ninth parallel.' Beyond this parallel, which constitutes the international boundary, his investigations were not carried. This inference has since been conlirmed, and the southern tongues or lobes of the Cordilleran ghuier have in part bt>cn traced out by Prof. F. C. Chamberlin and Mr. Hailey Willis of the U. S. Geological Survey, th(i furthest southward extension being, as determined by Prof. C;haraberlin, near the south end of Pend D'Oreille Lake in or about latitude 48", 20'." ' ' Gwlodical Magazine,' l>ec., Ill, vol. v., p. 'MS. j^^Sr ■ III iin •• ., le ill the ' Aiimrican (ioolo);i8t,' vol. iv, p. 215, Mr. Warroii Upliam j;ives somewhat different limits for tlii.s awa ol disiiorsion, but as liis slatomonts are basoil n|K)n my observations only, I am at a loss to under- Rtaiul Ids jiroumls for so doing. ' ('/'. Dana. ' Ainorioan .loiirnal Scionro,' III, vol. v, p. L'05. ' Uy iisHiiniing a ininiiiuiin slope of Olio dogriw, in areordanco with observations by Hopkins, this amount would lio iiu'roasiMl to about lil.OOO foet. ' ' (iuart .Idurn. (ieol. Soc.,' vol. xxxiv, p. lli>. " Cj\ • itullwin No. 4(1 II. S. (ioi)l. Survey,' 1887 ; " Seventh Annual Reiwrt U. S. Geol. Survey, 1888," p. 178. ii T OF THH ROCKY MOUNTAIN liHUION IX CANADA. 29 T On the coast, the extension of the Cordilleran glaeier which occupied the wide valley between the highlands of Vancouver Island and those of the adjacent Coast Ranges, divided at a point a few miles north of Si>yniour Narrows, to form two broad glacier-streams llowing in opposite directions, or south-east and north-west respectively. These have been designated the Strait of Georgia and (.^ueen Charlotte Sound Crlaciers.' Of these streams of ice, the first-mentioned proba])ly did not extend far beyond the south- eastern extremity of the island,- and it appe.n-s to be doubtful that it ever pushed south- ward so as to cover any considerable portio'i of the I'uget Sound basin. The (iueen Charlotte Sound glacier similarly appears to have terminated in the vicinity of the north point of Vancouver Island-Cape Commerell.' The extension of the l)order of the Cordilleran glacier to the Queen Charlotte Islands is believed to be indicated by certain striic found near the northern extremity of these islands, the direction of which would show that the ice here advanced southward, in conformity with the main direction of the long liords and channels between Observatory Inlet and Duke of Clarence Strait, whi Porcupine River, nor l)y Mr. Uussell further down the main vall(>y of the Yukon, the appearances there being on the contrary those of a country which had long been subjected to subaerial decay, and ' 'Quart .Toiirn. Geol. Soc.,' vol. xxxvii, p. '-'78. '-' Tliis iiiforeiK* is drawn i'rom the diaracterof tlie ieo action siiown l)y tlio roclvS in tlie vicinity of Victoria.— ' (Jnnrt. Jonrn. Geol. Woe.,' vol. xxxiv, p. 9."). ' ' Annual Roport Oiool. Surv. Can.,' 1880, p. 10:! 15. ' ' Quart .loiirn. Geol. Soc,' vol. xxxvii, ]). 282 ; ' Report of I'rogre.ss Gool. Surv. Can.,' 1878-79, p. !);i 15. ■' ' American Naturalist,' March, 1887; ' Geolo^ncal Magazine,' 1>pc., HI, vol. v, p. ?AS. '• ' Annual Report Geol. Surv. Can.,' 1880, p. hu U. ' ' Bulletin Geol. Soc. Am.,' vol. i, pp. .■)40-()i>. "' Geobftical Ma!,'azine,' Dec, III, vol. v, p. :!-IS; ' Annual Report Geol. Surv. Can.,' 18S7-S8, p. 40 1$. 30 (J. M, DAWSOX OX TlIK IMIYSKMi liAlMllCAL GKOLOGV whicli had iiol Ixm'h travi'ix'd citluT by glaciers or by lloatiiig ice capable of bearing erratics.' l"'urtlier iilustriitioii ol' the I'act that the extreme iiorth-westerii part ol' th<' coiUinent remained a land siiilaee, upon wliirh no extensive ulaciers were dev probable supposition on account of the absence, which has jiist been referred to, of traces of glaciation in thi' extreme north-west. The north-eastern margin of this great glacier is less easily delinea, l)ut it may (as a whole) be n'garded as having been conterminous with the Rocky Mountains proper, against whicli it must have rested directly in some parts of its Icnu-th, while in others the more or less isolated mountain-groups of the Gold llanges doubtless constituted local gathering-grounds which contributed their quota to swell the main stream. It is certain that the gn>at valley which separates the mountains of these ranges I'rom the llociky Mountains was throughout fiUed with ice. which had, like that of the main glacier, a southeastward direction of movement in the corresponding part of its course (i.e., in that part of the valley whicli is now occupied by the upper parts of the Columbia and Kootanie liivers). While it is possible that some part of this ice discharged laterally by the passes aciross the Uocky Mountains, tliis is rendered improbable by the absence of erratics derived from the (Jold Ranges both in the Rocky Mountains and in the foot-hills to the eastward . rthem. No evidence has oc<"urred to the writer such as to lead him to regard the boulder-clay I ;' *he Cordillerau region as a montiiie profoiuie of the Cordillerau glacier, and if the boulder- clay as a whole be not of this nature, but slight traces of any such bottom moraine are to ))e found. A hard stony material evidently ol' this character has been observed in some j)la<'es near Victoria, wedged into crevices of the glaciated rockf or protected by their overhanging ledges ; and in some of the deep V^haped valleys further inland in this part of Vancouver Island, a very similar deposit sometimes forms a great part of the drift. In similar deep valleys and other sheltered low places in the Interior Plateau region, some much compacted boulder-clays io which this origin may with probability be assigned, and which may represent true till, also occur, but the general covering of boulder-clay app(!ars to have a different and subsequent history.' ' ' Hiilletin Gcol. Sue. Am.,' vol. i., pp. 140, 543. -' " Narrative of Disi-overies on tlie Nortli Coast of America, 1836-39," p. 149. ' Mr. I. C. Ktisaell in his " Notes on tlm Siirfa' e Geoloniy of Alaslia," wliicli lias already been referred to, .seems to assume tliat I regard the bonlder-olay seen al( ig the Lewes, above Fort Sell OF Till*; lUJC'KY MOUJSTAIN KM'IION IN CANADA. 31 ^is* The iulbreucos above drawn rospeotiug the existence and limits ol' the great Cordilleran glacier now rest upon a great body of facts of dill'erent kiuds into the detailed eonsideru- tion of which it is impossible to enttT here. Our knowledge of the direction of motion and thickness of ♦he glacier-mass, however, depends chiclly on observed instances of rock- striation or scoring met with on isolated high points or on the surface of broad plateau- like elevations. In the study of the interior region of British Columbia it became apparent at an early stage of the enquiry, that in addition to striation and shaping of rock surfaces by glacial action referable to the various mountain-systems and diverging in all directions from these, traces of a much more general character and of older date also occur. The instances in which such evidence can be found under quite unt'quivocal circumstances are naturally rather infrequent. Some such were noted and described in my piiper of 1S7H, others have since been published in later papers and reports, and some,. met with during the season of 188!t, have not yet been made public. In order, therefore, to present this important evidenc'e in a concise manner, a number of the principal cases which have been discovered within the area of the south-eastward llowing portion of the Cordilleran glacier are tabulated below, the geni'ral order followed being from north-w^est to south- east. The approxinuite latitude and longitude of the localities is given, as most of these are not shown on the ordinary maps. Several of the mountains, indeed, have been named in the course of topographical and geological work still in progress, and are therefore not to be found on any published map. The names of these last are in parentheses. List of sumk I'lascirAL inst.\ncks oi- St1!iatios i;BKEi!.\r,i.i-; to tiii) SoiTii-EAsrw.Aui) i'outio.n oi- Tiiu CoiiDU.i.uiiA.N GlACIIllI, IN Till) InTE1!I0I: KeCIO.S 01' BuiTISIl COI.UMIIIA. PLACK. 1. Summit of Tsa-whiiz Mountain 2. Summit oCSiutor Knoll 3. Spur of Tsi-tsull Moun^ tain .\ppro.\. Lat. o3° 40' 53° 5-2° 40' 4. Plateau north of Cliil- cotiu liiver 52° 5. Iligli plateau between ) j North Thoini)son and > 51° 02' Bonajiarte Rivers .... J 51° 08' 51° 05' Aji|iro.K. Long. 123° 125° 45' 120° 10' 122° 40' 121° 13' 120° 4(i' 120° 35' ilcight in feet libuvo 3,240 3,550 3,700 3,G50 5,000 4,220 i 5,220 IJirectinn of strijD ((rue beiirinKs). f About \ S. 10" W. S. «■' E. S. 37° W. S. 2° E. S. 35° E. S. 30' E. S. 34° E. llKMAnK.s \si> Kkkkuknck. tVn isolated point 800 feet above plateau between Chilacco ami Fraser llivors. Water-worn bouldor.s and pebbles i found. ' Hep. of I'rog. G.S C.,' 1875-70, p. 21)2 ; 'Q.J.G.S., ' vol. xxxiv, pp. lOd 104. An isolated point 250 feet above sur- rounding country. Wide lower plateau to the north. Erratics on summit. ' Rep. of I'rojr. G. S. V.,' 1870-77, p 80; '(}. ,1. ti. S.,' vol. xxxiv, p. 101. Ice here passed between Tsi-tsutl and inner border of Coast Ranges, crossing a 'col.' Lower country to the north. Direction eomewhat affected by the local circumstances. 'Q.l. G. S.,' vol. xxxiv, p. 102. The direction ia transverse to tlie great gorge of the river. Several localities. 'Rep. o."" Prog., G.S.C.,' 1875-76, p. 201; 'Q.J.G.i!.,' vol. xxxiv, p. 101. These are examples merely of the glaoia- tion met with on prominent )iart8 of this plateau. Erratics are strewn over all parts of the plateau. Of these examples all but No. 5 have already Though ijoculiur in some places, in containing considerable masses of stratified clayey gravels, the boulder-clay of the Upper Yukon basin, where I have seen it, is often a typical boulder-day of the character to be found over thousands of square miles in the interior of British Columbia and in the northern part of the Great Plains. Cf. ' Bulletin Geol. ^' I 50°5!t' 1L'0° L'ij' 10. Tod Moiiiilain 5(1'^ DO' U'J°5. 11. Ilij;li philcaii Ik'Iwi'oii j Ailaiiis^ShiiHwiip I,akff 'il" 1' 15. (Cinder Mountiiin). .. , l(i. (Murray's Mountain).. 20. Iron Jlountain 50° 03' 21. Hinli point on plateau 20 miles south of Nicola Lake 40=50' 22. riatoau near Chain hako 40° 40' 21!, ( loailslone Peak) , 49° 25' 4 Toad .Mountain 49° 25' llli°41' Height in Ulrcctiiin iil fuel iiliovo I dtriii' (true veil. boiiriiigs.) 12. (Clear Mountains) he- twccu Ihil Cnek Valley .-)0° 44' ' 121° 42' ami rra.ser Kiver 50° :Jti' 121° 40' 13. Sunindl ea.st of I'aul's I'eak, near Kaniloops.. . oO" 42' 120° 14' 14. I'lateati 14 niilois south- of Kaniloops ol" 31' 120° 24' , 50° 34' : 121° 08' . 50° 31' 121° 33' 17. (Spaist Mountain) 50° 23' 121°05' 18. Zakwiiski Mountain.... 50° 00' | 121° 10' ' 19. Ilijih Point on ridtre l>e-, 1 tween last r,iid Nicola Rivor .50° 12' 121° 13' 120° 45' 120° 35' 120° 15' 120° 50' 117° 21' 5,450 5,840 0,100 7,070 7,040 3,520 4,190 5,070 0,880 5,780 0,000 5,030 5,280 4,380 4,075 0,370 0,990 S. 37° K. S. 35° E. Ukm uks axu Kkkkhknck. been puhlislied, vol. VI, p. 1)52). 7,260 : S. 44" E. S. 27^ E. None seen u a asrE. S. 31" E. S. 50° E. S. 10° E. ! S. 28° E. S. 13° E r About L S. 29° E, f S. 0° E. \ to I S. 18° E. f S. 20° E. to I S. 28° E, S. 15° E. fS. (i° E. \ to I S. 33° E. Grol. Mag.,' Dec, III, p. ;>&-'). As now given slight changes are made in some of the posi- tions an51. A few small erratics seen, some on sum- mit of point with height of 0,210 feet. This is also glaciated but the dir Jii i is indetorminabU<. 'Gooi. Mag,' I'ec I III, vol vi, p. 352. Small erratics found everywhere strewn with local material up to the highest points of these mountains. Siunmits considerably shattered and weathered. Various parts of the range were examined, the positions and heights given are those of tho nortliern and southern high ixDints, (Chi-pooin) and (Blustry) Mountains. Travelled stones and considerable cover- ing of drift material. Local topography explains divergence of glaciation to eastward. General direction from glaciated surfaces, the striic weathered out. Travelled stones on summit. Neighbouring deep parallel valleys of Fraser and Thompson appear to have inlluenced direction. Boulders and stones of varied origin on summit. Isolated high point on plateau north of Nicola River. Isolated high point at head of Nicoamen River. Glaciated but direction uncer- tain, the trachytic rocks much broken up and weathered. Travelled stones 3 or 4 inches through found on summit. Granite boulders strewn over this and neighbouring ridges, which are com- iwsed of volcanic rocks. Situated near confluence of Nicola and Coldwater Rivers. Glaciation heavy. Travelled stones on summit. ' Rep. of Prog. G. S. C' 1877-78, pp. 136 B, 140 B. ; ' Q. J. G. S ,' vol. xxxvii. p. 272. Striit somewhat obscured by weather- ing. Situated midway between Okanagau Lake and Similkameen Rivor. ' Rep of Prog., G.S.C.,' 1877-78, p. 137 B. ; ' Q J.G S.,' vol. xxxvii, p. 273. Higiiest point in this vicinity to east of Coast Ranges. This is in West Kootanie District and forms the watershed between Kootanie and Salmon Rivers. Glaciation light but distinct. No erratics observed. ' Ann. Rep. G.S.C.,' vol. iv, p. 40 B. ^ -i!^' -T OF TllK ROCKY MOUNTAIN UKGION IN CANADA. 33 •- lu furlluT explanation of the above ta1)le, it may l)c pointed out that llie less tou- siderable height of tiie points in the northern part of the Interior I'lateau upon which evidence of the general glaciatioii has been I'onnd, depends prin( ipally on tiie less elevated character of all that part of the plaleair. I'oints equal in height to those enumerated to the southward scarcely occur, but this circumstaiue, with tlie less bold relief of this part of the plateau, in itself enables equally good evidence to l)e obtained at lower levels. It cannot be accepted as having any bearing on the probabl" tliickness of ice during the maximum epoch of its accumulation in the northern and soulliern parts of the plateau respectively. The much greater number of instances drawn from the southern part of the region is due cliielly to the fact that portions of this part have now been closely and systematically examined, whili! the suivcys carried out lo the north (in 187;">, 187() and IST'.t) were of the nature of reconnaissances, and thus did not require nor even admit of the occupation and examination of all the high points. The same remark applies to the absence of ol)scrvations at great heights in the area of the north-westerly ilowing part of the Cordillerau glacier. It should also be renuirked in this connection, that glacial striation on the summits of mountains such as most of those here cited, can generally be found only by close examination and search for unvveathered rock-surfaces, ami that many cases occur in which no certain indication of direction can be obtained evi'U by such search. Particular interest attaches to the observations on the elevated rough plateau betweeu the North Thompson and Honaparte liivers (numl)ers ij-'.t), because of the fact that no considerable area of equal height occurs to the north-westward (from which direction the ice came) for a distance of about ;350 miles. An in,spection of the table will show that the maiu direction of motion of the part of the Cordillerau glacier represented by it was from north-west to south-east, along the Interior Plateau and parallel to the maiu mountain elevations of the Cordillera. Where least disturbed by local circumstance's of the relief of the surface over which the glacier ilow(>d, and by the occurrence of adjacent ranges which may have deflected the ice, the mean direction lies between S. 30' E. and S. 35° E. The general surfac.'e of the glacier must have stood at one time, in the southern part of the Interior Plateau, at a height somewhat exceeding t,000 feet above the present sea- level, the thickness of the glacier-ice covering even the higher parts of the plateau here being thus at least 2,000 to 3,000 feet, while it attained a thickness of about 6,00t. feet over the river-valleys and other main depressions of the surface. From the light character of the glaciation observed on most of the higher points, it is probable that the glacier did not much exceed the thickness here assigui'd, but that it preserved this thickness, together with its full width, to near the forty-ninth parallel, is indicated by localities 23 and 24 of the table, which are 170 miles distant from each other on an east and west line. The erratics which occur on the summits of most of even the highest mountains on which glaciation has been found, as well as those scattered over high points where uo striae were detected, are as a rule to be classed as pebbles rather than as boulders. They are generally more or less rounded, but are occasionally striated. They are usually found sparingly dispersed among rocky debris, derived from the mountains themselves, and with little or no accompanying earthy drift. It is believed that these foreign stones were carried on the surface or within the mass of the upper parts of the Cordillerau Sec. IV, 1890. 5. 34 (',. M. DAWSON (JN TllK IMIYSIOCIJAI'IIK'AL (IHOLOGY ulacici- wiicii at about its iiiiixiinuiii, uiul that they vvcro h'lt .slmudod whi'io now Ibuiid as il (Iccliiii'd. Many ohscival ions luii^ht 1)1- litod to fsliow that as i\w main i,'hicior dccTcased iu tliirkiii'ss and its Iroul ivln-atod, its dircrtioii ol' niovcnient l)eranit' more and inoro subsidiary to tlw- local ri'liff ol' lh(> surlauo. This subjcl can, liowi .or, only bo alluded to in'io. The gcnt'ial chaiacti'V ol" thi; change may be diasrainunitically expressed as uhown bi'low, when' the vertical lines represent the mountaih ranges bordering the Interior Plateau, the arrows the direction oi' movement ol' ice at successive periods in its deiline : — \/ 0) w W Kvid('nr. It has already been pointed out that the British Columbian portion of the length of the Cordillera, when the glacial epoch supervened, st^-ad probably at least 000 feet higher than it now does. If this be admitted, as there appears to be every reason that it should, it will be found that the recession of the Cordilleran glacier and its accessory glacier- streams was contemporaneous with, if not brought about by, a movement of subsidence ; for when the Strait of Georgia glacier had diminished only so far as to bare the glaciated rock-surfaces of the south-eastern extremity of Vancouver Island, these were at ouco covered by irregularly stratified deposits, comprising sands, clays, gravels and boulder- clay, in some of which marine shells are found.' Similar facts are observable further north in the Queen Charlotte Islands,' and it may thus be inferred that the laud ' ' Quart, Journ. Geol. Soc.,' vol. xxxiv, p. 102. '' ' Quart. .Tourn. Geol. Soc.,' vol. xxxiv, p. 95. ' ' Quart, .Tourn. Gool. Soc.,' vol. xxxiv, pp. 90, 122. ' ' lioport of Progress Geol. Surv. Can.,' 1878-79, p. 99 B. OF Till-: ROCKY MOUNTAIN ItKCJloN IN CANADA. 35 had goiu' down ivt U'list 1,000 f.«ot*froin the stago of its mivxiimita .'lovivtion h.'Ibiv uny very considonil)!!' di'in-asi' in hW.o of tlic u-rcat ^-lacicrs ociirred. \V'' an- nualdf l<> Ibllow tiiis .su}>.sidi'iicf I'nrlhiT in d.'lail, as whalrvfv tni.rs it may have li-l'l aloiiu' tii«« ooast must have boon olilitcratcd by sul)so(infiit a. lion dnvini-' tlio suoc'iMlinu' p.'viod «>l olovation which is rolorrod to on a lator paj^v. Wo iciiow only tlial in tlio Sliiut ot Ciooi'n'ia and in (iuoon Chavloilo Sonnd a oonsidorahlr tiiirkiioss of lino, ivt-iilarly Htratiiiod silly matorial was laid down in a traii(|uil uiannor altovo I ho lirst hi.iildoi-olay aiul its assooiatod deposits, nndor oonditions iniplyinn- tiiat tlio land was at loast Ironi 100 to 200 I'eot lowor than at prosoiii.' How mnoh lowor the land may havo st<)(.d al tins timo wo havo as yot no ovidonoo on tho coast to show. Ill provions publications I havo dassiiiod Iho supi'viii ial deposits of the inloriov oT British Columl)ia duo to tho Cilacial period under tho names inimo'/i/ird ilrifl and mmliiird flnfl,ihi\ first-montionod including tho boulder-clay, Iho second omln-acin^' deposits of various kinds whic^h hav(> IVequontly in larg.> part been formed l»y the ro-arrangvd materials of th(> 1)oulder-clay. Though tho divisi(m thus made is not in all cases perfectly dormito, it is warrantable from a general standpoint, and convenient for purposes of description. Tho bouldor-olay though not dilfering in any obvious manner from that of the Groat riains, or indeed in any important respect from that generally found in dilferout parts of the Northern llemisphero, presents hero somo minor po(^uliaritios. It consists generally of a paste of hard, sandy clay, containing usually a very considerable proportion of lino sandy matorial, through which stones of all sizes are irregularly scattered. Its colour as a rule vari(>s from light brown to pale yellowish or grayish brown, but in freshly exposed sections is sometimes bluish-gray. It has usually a more earthy appearance than that ot the eastern part of Canada, and very often over extensive regions forms th(> soil in which the trees are rooted, without tho intervention of any modilied drift. An unusually largo number (including in fact much the larger part of tho w^hole) of the stones and boulders are well rounded and water-worn, but a variable proportion showing distinct and some- times heavy glacial striation or polishing is constantly present. As later and more extended observations have served only to bear out the description of its mode of distribution given in a former publication, this may be quoted as originally writ- ten :_" Over considerable areas this material is concealed beneath tho accumulations which ibrm terraces and low-level ilats, in relation to former lake and river-valleys. There is a remarkable uniformity in these boulder-clays in every locality in which I have examined them. In many places they form low rolling and broken hills between the river-troughs above the level of the higher terraces. In this case they appear sometimes to bo spread in a comparatively thin layer over a rocky substratum ; while in others they are of great depth, and by the irregularity of their arrangomi'ut havo themselves produced many of the minor features of the surface. They frequently show a tendency to form more or less well-defined high-level plateaux, and are spread almost universally over the elevated basaltic region of the interior, in most places so uniformly, notwithstanding minor irregularities, as to allow the underlying rock to be very seldom seen." ' ' ' Annual Report Gool. Snrv. Can.,' 1886, p. 105 B. '' See 'Quart .Tourn. Gool. Soc,' vol. xxxiv, p. 103. i 86 (i. M. D.VVVSOX ON' TIIK I'll VHKXlRAI'irif^M/ (iK()r.O(iY Tli.'sc rcmiiiks ivI.t piiiti. uliiily t(. tli.' IiiLciior IMiit.'uu r.'^ioii, in Hk- souIIutii part (if U'lii'h liili-r ohscrviilioMs show tliiil foiisidcnildc (liU'cifnci-.s I'xist us l)ctw<'fii viirioiis iir.'iis ill till- iivni'4'.' (l.'plli mI' iIk- liMuMiT-cIay d.-posil. In iIp' r|)p.'r Yukon Imsiii, a->Mii», it lias l)i'cM round tlml \\v l)ouloulder-clay is i'ound lo pass. Tiiis is, however, ayain rel'ericd lo on a later p;ii^c. riirouiihoul I he Interior i'lalciiu I he upward limil of the Ixmlder-elay is i'ound at a hciifht somewhat iircati'r than .'),(M)0 feet uhovo the present sca-h-vcl, and corrcspondini? in this res|)ecl with the iiinhest level of Well-marked terraces; the hitiher terraces in lacl ^rencrally consisting of material iilentical in character with that oi' the u:oiieral coverinij,- ol honlder-clay, or so closely ulike as to be indistinguishable iVom it. Thonifh, as previously noted, travelled stom-s occur on much higher points, no boulder-clay, ami very little line drill materiiil of any kind, has been lonnd al)ove the hi^'hest terrace-level role .-red lo. These hiiihesl terraces may be state.' to have an averajfe elevation of about 5,2i)0 feet. Lower terraces, ranyini'' between r», 01)0 and '" '' -^ have also beon i'ound in a certain number of widely separated localities, but th« princn.,.. ' vclopment oi' terrai'es is I'ound ladow :!,S()0 or ;!,.'»(I0 I'eet, and esi)ecially when levids ol' ••'.,(10(1 i'eet and under are reached. I'.elow ;i heii-hl oi'aliout .'i.OOO i'eet the whole Interior I'laleau rei>ion may be described as terraced, ;ind although alonj^' tho various river-valleys many terraces occur which have evidently been produced by the streams themselves whil' cutting down through the drift deposits, wiiich at an earlier stage had lilled these valliv.-, (iiese need not be considered here, the point to which it is wished lo draw attenti >u !>eing the exislenci' ot terraces re(juning for their explanation a general Hooding of tlie , ountry. Such terraces are foiind to be not conlined lo the immediate valleysof the rivi-rs, but lo occur in dill'erent situations along the higher slopes, and to fringe at similar elevations the various irregularities of ihe idateiuxx. The existenre of that which has been rel'erred to as tho highest t'rrace-lovol was first asoertainiid in iSTll, on the upj)er slopes of the 1 1-ga-chuz Mountain, in latitude 5-2" 45'. This terrace or beach-line has already boon fully described,' and it ncoil here bo mentioned only that its elevation is a, '270 feet. The <'ircumstances of observation at this place appear to be perfectly unexceptionable, though the torraco I'ound here has remuiiiod for a long time an isolated inslanco. During the progress of geological work in tho .southern, partof the Interior Plateau in 1.SH8 and ISfSli, iiddiiional inl'ormalion ha.s, however, been obtained rospecling thi.s or other similar very high terraces. No publi(!ation of the results of this work having yet occuiYod, it will bo uetvasary to refor briefly to the observi'd facts. Tod Mountain has already been mentioned ill connection with glacial slriation, and its isolati'd position has boon noted (soo p. ;^2). A uarrow and rough, but fairly wm'II defined terrace occurs or. its south side at a height ascortaiued to bo .^.IIB feet. Nearly i i tho same latitude with the last, but twenty-eight miles further west, ou the edge of tho plateau to the south of tho lake in which the Tranquillo River rises, a torraoo occurs at a hoighi of ;'),340 feet. This is perfectly distinct and somewhat (extensive, and was see)i from a distance to \n\ repeated ou tho slope of the plateau some miles to the north. ' ' Qmirt. Jouni. Geol. tioc.,' vol. xxxiv, p. 107 ; ' Keiwrt of I'rogress Geol. Surv, tan.,' 1876-77, p. 38. ol' TIIK l!(M'K^ .MOINTAIN UKUIoN I N CA N.\ l»A. 37 ward I)oyoii(l tlie laki-. WlnTf I'Xiimiiu'd, this Icriiicf is IiuukI In be ((iiiiimscil dl lioulilri- ••liiy or idi'iiliial miilcriiil. Oil Muriiiy's Moiinliiiii, silnnli'd alioiil lil'ly inili's distaiil IVoiii lln'lasl, uii n S(inlli-rii>t hi-ariiiu', in laUtudt! iiO ' ill', and which luiins a suniinil nn the inoiinlaiimiis lidui' l)ri\\('i'n thi' KiasiT and 'riiompson IJivcrs, a Icrran' al a hi'iylit <»r '),:i7'» r<'i'( oirurs. Il is siliiahd in II dt'prt'Nisioii on till' Noulht-rn .side ol' tiir niuunlaiii, and ihiuiuli mmmII and Miinrwhal in-ciiula'- iis pcrtcctly distiint. l''orty-six niilfs norlli o|' Ihi' lasl-mi'iitioiii'd loralil.y, on lln; wi'st sidi' oi' llic Maildc Mountain ranL!;(', '" h'titudi' ')l^S', is ii diistiutt Icimci' iit an tdcvation ol' hctvvt'cii ;>,JtiO and r>,m) f.'.'t. The clciii't'sl and most nn<'(niivo('ai cvidcni'i' of tlic rxislcn I' a wat'i'diin' al tliis throat, clfviil ion was, liowcvcr, I'onnd in Odohcr last on thr casti'in (ii<)|)pnsitc slcip.' nl' ihi' Maihlc Mouiilains. Tliis look.s out upon ihi'sra-likci'xpansi'or (hi- liasallir ( Irccn Tiiidn'r l^latcau, whii'h has an clcfViition of from 3,K()0 to -t.dllO Tcct. The I'litiic i-astcni liasc nj' iln- rant>(', lor a liMi<;th ol' lil'ti't'ii miles, i.s hciipi'd willi drilt dcposils wliicli ar.' iiioir or li'ss distiiit'liy ItM-raccd all aloiiii', the appi'arancc in-csontcd i'rom a distance licinii' minli like that ol' llic ll-^a-chuz Mountains when similarly vicwfd.' Wln'U closrly fXiimiiird tii s accunudation ol' matt'rial is found to consist in i)arl of moraine mounds an(Mi feet, and other similar terraces, one of w^hich was estimated at ahout 4,900 feet, were seen at a distan(!e. On Dease Kiver the highest observed terrace is at an elevation of about 4,(i00 ' See illiistriition ' Quart. .loiirii. Geol. SdC.,' vol. xxxiv, |). Kill. - The only cases noted wlii('li liiivu the ai>i)earaiico ol" indicatiiij; toriaccs at liit;h(M- IfVi'l.s tiiiin Ihotn here (le.srribed occur njOTii Toil Moiiiitain. Two very small terrace-like Hats, lodjieil in ilopn^s.sioTis w\aniin;ht,s ol'r),()."J(l and .5,7-0 feet re.s|KiCtively, wore herefoum'l. These, hisvever, appeared to lie conipiw d of line earthy material, and as respects the marginal ranges of the Cordillera, to the north, and it will l)e noticed that (with the exception of Tatlayoc'o Lake) none of the higher terraces have been found on the eastern Hanks of the Coast Kanges, which, it may be presumed, were covered by glacier-ice at the time of formation of these terraces. It may iurther be stated that, as a whole, the higher tiTraces referred to in the foregoing paragraphs are comparatively rare, and that many of them show the eifect of considerable denudation, while below levels of 3,iS00 or 3,500 feet terraces ar(> extremely abundant almost every- whei'e, and are very frequently wide and iu an excellent state of preservation. The heights oli he terraces above noted have been barometrically determined within small limits of error, the influence of the weather at the time of observation having l)een eliminated by the n.se of station l)arometers. An inv(>stigation of the lower terraces, carried out by more reiined means of measurement, might produce inten'sting results, but for the purposes of this sketch it will be uuu(>cessary to enter into the great mass of observations which has been accumulated resi)ectingthe heights of these in difierent localities. Before leaving this branch of the subject I would, however, mention the iact that water-rounded stones occur on the slopes of the mountains at the summit of the Pine Pass across the Rocky Mountains (lat. 55 20), several hundred feet above the actual summit, aiul that an apparent terrace was noted at 300 to 500 feet above the same summit, or 3,300 to 3,500 feet above the sea-level. Allusion may also he made to the terraces at great heights met with on the eastern slopes of the Rocky Mountains, which are again referred to on a subsequent page. In the paper on the " Superficial G-eology of British Columbia," published in the ' Quarterly Journal of the Geological Society' for ISYS, to which reference has frequently lieen made, it is stated that I had not up to that time met with any distinct indications of moraines which might l)e referred to a great Cordilleran glacier.' Ob.servatious made during the past than elsewhere beset with flilliculties. These arise in great part from the varied and bold orographic features of llie Cordillera, which render it necessary to have in view the local ciriumstan.es of each observed fact in a manner not found so essential in a less broken country. There is necessarily much dilfuulty in assigning a relative order to tht' various phenomena, and it is especially difTicult in some cases to separate those due to the epoch of the Cordilleran glacier from those l>rought about by a later advance of glaciers from the various mountain ranges, whiih without doubt occurred. It has for this reason been considered necessary to review brieJly, in foregoing pages, the phenomena believed to be clearly connected witli the first and most severe epoch of glaciation before attempting any esplanationof the mode in which these have been brought about. It is particularly in a complex and vast region like that here treated of that the accumulation of a great number of observations, someoi whi li appear to be incompatible one with another, embarrasses any process of generaliza- tion. With the equipment of a limited number of observations only, derived from some single portion of the region, no great diiliculty might be found in including these in some logically consistent scheme which might or might not eventually prove to be correct. While therefore some general account of the observed facts has been given, it is impossible to ex- tend this so as to include the description and discussion of the whole number, and the writer, in having to bear in mind a mass of detail with which this paper cannot be encumbered, labours under some disadvantage in drawing conclu'jious which the reader may feel able to ac(^ept, from that part of the evidence which has been broiight to his notice. This dilRcnlty is added to by the circumstance that it appears to be desirable to study the glaciation of the Cordilleran region on its own merits, divesting thi' mind as far as possible from hypothes'^s advocated or received for the eastern half of the continent, wi'h which there is no a jirhri ground for believing that the events of glaciation of the Cordillera were identical, but rather some basis for the belief that they \vere comple- mentary. In previous public ations I have suggested two modes by which the production and arrangement of the boulder-clay and other superficial deposits may be explained,' but subsequent and more extended observations appear to show that neither of these is fully satisfVictory. I will here first mention these siggestions, and then point out the main facts which appear now to stand against them : — ' See illustration ' Quart. Journ. Geol. Soc.,' vol. xxxiv, p. 109. '' 'Quart. .lourn. Geol.Soo.,' vol. xxxiv, p. 113; ' Report of IVogioss (iool. Surv. Can.,' 1880-77, i). 36. ■' ' Quart. Journ. Gaol. Sou.,' vol. xxxiv, p. 119 ; vol. xxxvii, p. 283. i^ 40 fl. M. DAWSOX OX TIIH I'll VSKKiUAl'IHCAL (iHOLOGY (I) Tli«' ufiiiTiil fsubsidcmi' of tlic Cordillt'raii ivgion, wliich \vc round reiison to believe was ill proLircs.s a( tln' liim- at wliicli the s'lin-'iers began to diminish, may have continued, lill, at a later date, wlu'ii the main Cordilleran glaa\;n! that this hypothesis requires an almost inconceivable co-ordination of glacier -dams, besides which there is really no valid (>viden(M< to show that glaciers are capable of holding in ' It slimild he borne in mind, howevur, tliat in timos pteologioally recent, very great dinnges in relative level ef hiiiil !iMil sea can be slimvn tolinve talien place, and tliat it is tlierefore un.safe on any c priori gronnds to exclude sni^li (.'rent local cliaiijies IVeiii consideration. Air. TJpliani, collcctiiij; a number of authenticated cases, wliicb need not hero be enumerated, notes post-I'lioceno elovati^ ■'■ iopression of 1,000, l.'iOO, 2,0(10, '_',(!( to and ;!,(I00 feet. Wri^bt'.s "Ice Aj;o in America," p. .j8'J. Si - . :i .. ' .r by the Duke of Arjryll, ' Scottish (ioogruphical Ma;,'a/.ine,' vol. vi, p. 177. OF TlllO KOCKY MOUNTAIN KKGION IN CANADA. 41 any very great depth of water, and much reason to believe that they are ineompetent to do so.' lu addition to this, such an hypothesis is even h'ss adapted to a<'count for the body ol' water beneath wliieh a great part of the Yukon district must have been submerg^-d when the northern end of the great glacier left it. 8till further, it leaves us without any apparent cause for the decay of the Cordilleran glacier at this particular period, the subsequent partial resum)ition of glacial conditions, and their final disappearance ; all wliich events would require to be attributed to some general climatic or cosmic series of changes. A third hypothesis which I now venture to suggest combines some of the features of the two first and appears to me to include all the observed fac-ts better than either of these, and to form indeed a reasonably satisfactory explanation of the phenomena with which we have to deal : — It may be supposed that as a consequence of, or correlatively with, the gradual subsidence of the northern part of the Cordillera, the supply of snow producing and maintaining thi' great glacier became smaller, and that more of the winter increment was meUed away during the summers, till at length the glacier itself became nearly stationary. Its decay, still continuing, resulted eventually in the formation of e//^'-/((drt/ lakes. - Tliese might be expectt^d to occur in tlie central parts of the Interior Plateau at a considerable distance Ironi any of the bordering ranges, which still doubtless continued to contribute a certain quantity of ice to the mass. This central part of the Interior Tlateau is, besides, that characterized by least precipitation at the present time, and consecjuently that in whidi the want of a continued re-supply of ice from the main n6v6 would first become apparent. Such lakes, it may further be intern d, would originate in the first instance near projecting mountains or minor mountain ranges of the plateau region, all which conditions are in accordance with the mode of occurrence of the observed highest terraces. The terraces themselves may have been somewhat rapidly built, by the washing down to the water-line of material which had previously accumulated ou the higher slopes of the projecting points, as well as from deposits borne by floating ice. The similarity in elevation of these highest terraces would appear to imply that the bodies of water thus formed within the area of the glacier were in more or less complete connection, and if all the observed terraces at about the level of 5,290 feet in the southern part of the Interior riateau may be supposed to have owed their origin to a single lake, this must have had a length from east to west of about eighty-eight miles, with a north and south width of about fifty miles. This does not include the terrace or beach on Il-ga-chuz Mountain, much further north ; and though it is possible that this also may have been formed in a part of the same lake, it seems more likely that a separate lake opened here, the height of which may have been in relation to the elevation of the general surface of the glacier at the time. Other terraces which have been noted at heights down to 4,000 feet or thereabouts, may be attributed to later stages of the same or similar lakes produced in the central parts of the glacier, and it may be that some or all of these lakes were com- paratively shallow, being floored as well as walled around by the mass of the glacier. At what particular stage ': .\ the decay of the great glacier these englacial lakes were ' Tlimijih it apiHiars to 1)6 frequcMitly tiike'i for gruiited tluit filiuier-ice is capable of liolding in t'reivt inhiml Beau, tliore is really little warrant in nature for sudi a belief, Huch miniature instancca as Morjelen See beinj; Bcaruely oases in point. = This term is ouiployod as a convenient one for eucli lakea developed ou the surface of a great glacier svs those found by Nordonskjold on the inland ice of Green, and. Sac.IV, 1890. 6. 42 fi. M. DAWSON ()X Till'] PIIYSIOGRAPHICAL (JKOLOGY iinally (liainetl it is irnpossil)lc to say, but it appears probable that this must have hap- pened before the glacier ceased to cover the greater part of the Interior Tlateau. It may further be supposed that the gen-'ral subsidence of the Cordillera before alluded to proti-ressed iiari passu with the deca; t +'.i^ Oordillerim glacier, and there is some evidence (which, however, cannot be given her tail), attbrded by terraces formed contempor- aneously with moraines, to show that tii : nbsidence had carried the laud down to a stage al)0ut 3,000 feet below the present sea-level, while tongues of the great glacier still exti-nded as far south as latitude 49° 40'. Further evidence of the same kind favours the ])clief that by the time the end of the glacier had retreated about one hundred miles to the north, or to latitude ST 30', the subsidence had progressed to about 3,800 or possibly to 4,000 feet, which was probably about its maximum. The great glacier must have retreated rapidly toward the close of the first period of glaciation, and have become reduced to syistems of small local glaciers in the mountain regions, unless indeed the central portion of the nave may have retained a coniluent character till after the second period of glaciation. Along the retreating front of the glacier, and subject to a certain amount of re- arrangement by the water which washed its base, the boulder-clay appears to have been laid down, and is, as before stated, indistinguishable in general character from the earlier and higher deposits of the same kind attributed to the englacial lakes. The land can not have long remained at the low level which has been above assigned to it, the movement in subsidence being immediately followed by one in progressive elevation, diuing which all the more obvious and well-preserved terraces of the Interior Plateau and other parts of the entire southern portion of British Columbia were formed. Eefore, however, following this presumed re-elevation further, we may glance for a moment at the condition of the northern part of the Cordilleran glacier during the supposed period of greatest subsidence. It has already been mentioned that the boulder-clays of the Upper Yukon basin present certain peculiarities. "While in many places along the Upper Pelly and in most instances along the Lewes Eiver the boulder- clay is of a typical character, it is often on the former river and sometimes on the latter represented by earthy or clayey, gray or brownish, stratified gravel-beds. These are found to pass horizontally into true boulder-clay, while in other instances they are interbedded with rude layers of boulder-day, or form the lowi'r or upper members of sections showing a considerable thickness of boulder-clay. The evidence I believe to be conclusive that they constitute with the boulder-day a single formation, which represents the first deposit of the retreating northern extension of the Cordilleran glacier.' The stones both of the stratified gravels and the more typical boulder-clays are generally in this region well rounded, and glaciated stones or boulders are comparatively scarce. The stratified earthy g-ravels are moreover most abundant in the higher parts of the country traversed by the Upper Pelly River, at levels between 2,o00 and .'5,000 feet. The character of the deposits representing the boulder-clay period in this northern region are taken to indicate that the total amount of subsidence was there less than to the south, that the material dropped along the front of the retreating ice-foot fell into shallower water, and that in conformity with this circumstance and the less perfectly enclosed character of this northern region and more diffuse arrangement of the mountain f ' ' Annual Roiwrt Geol. Surv. Can.,' 1887-88, pp. 119 B, 120 B. or TIIH EOCKY MOUNTAIN IMXiTON IN CANADA. 43 •f I rauges there, it was subjected to a greater degree to current aetiou.' Still further to the north and north-west, reasons have already been given for the belief that the land was not generally submerged at this time. It appears to ]u- quite probable, however, that a wide strait was opened along the Yukon Valley to Behring Sea.- It should be noted, before going further, that the faet that marine shells have not Ix'en found in the deposits which are supposed to have been laid down along the retreating front of the great glacier in water in communication with the sea and governed in elevation by it, may at first sight appear to constitute an argument equally weighty against the hypothesis above stated as against that of a general submergence to an extent of over 5,000 feet. On consideration, however, it will be found that a submergence of the Interior Plateau to a depth not exceeding 3,800 or 4,000 feet would still leave that region well enclosed by mountain ranges, and bearing in mind the quantity of fresh water which must have been poured into it, the necessary coldness of its water, and the great amount of earthy matter which was being discharged into this water, ind comparing these condi- tions with their nearest known analogues, we are led to suppose that marine life must have been very scarce if at all able to maintain itself. These arguments, liowever, do not apply with equal force to the region of the Yukon basin, and here, if anywhere, we may eventually hope to find evidence of such life. "When the re-elevation of the land succeeding this great subsidence had progressed to a certain extent, the increased height of the mountain ranges of the Cordillera, in com- bination with the more general causes of glaciation of the northern hemisphere, became such as to lead to the increase and re-advance of their local glaciers, and probably also to a resumption of southward and northward movements from the central gathering-ground of the o-reat Cordilleran glacier. Evidence of this second advance of ice is found particu- larly in the Interior Plateau region, where the growing glaciers have pushed out from the mountains along the various river-valleys, ploughing up previously formed terraces and gravel deposits and piling them into moraine ridges. ' On the coast, evidence of like bearing is afforded by the upper boulder-cl'iy of tho northern part of the Strait of Georgia, which may find its representative in the southern part of the same strait in the scattered large erratics, which are everywhere abundant there, in or overlying the highest layers of the drift deposits.' ' In liis notes on the " Surface Geology of Alaska " ' Bulletin Geol. Soc. Am.,' vol. i, p. Un, Mr. I. C. Russell allndea to the houlder-clay seen by him on the Lewas as doubtfully ropresentint: true boulder-clay. I may .say, however, that I feel assured that to anyone acquainted with the bouldcr-day of liritish Columbia and the Great Plains such no doubt would be admissible. 1 may add that Mr. iMcC'onnoll, who is jicrfoctly familiar witli these northern boulder-clays, coincides in my reference of those of the Lowes- ■' Some evidence of this is indeed sufi^ested by Mr. I. C. Kussell (Op. Cit. j). KSS)), who writes :— " It may ho well in this connection to direct attention to certain obscure indications of terrr.cns or sea-cliffs, at iin elevation of fifteen hundred or two thousand feet, on a number of the mountains near .. j Yukon, below Nnlatto. None of these mountains have been closely examined, and it is impossible to st.ato whether the indefmiU! lines which may indicate terraces are b0ri7.0nt.al or wliethor tlioy coincide in elevation. It is not safe to assume that tliny are terraces, as it is possible that they may indicate lines of structure or be due to land slides. Tlio mountains aro so situated that they could not have retained a lake, and if water lines exist on thom tlieir origin must be looked for in a submergence of tlio land." ■' Cf. ' Quart. .lourn. Geol. Soc.,' vol- xxxiv, p. 113. It should bo st.ited that snbsc.(|uent investigations h.avo led nie to doubt the validity of the evidence in the cases llrst mentioned in tli(( puldication here referred to, but h.avo at the same time added many additional facts, which cannot here be detail.id, in ci.nnrmation of the general pro|)osition. * ' Annual Report Geol. Surv. Can.,' 1880, p. 105 B. 44 (i. M. DAWSON ON TIFH PTIYSIOGRAPHICAL GKOLOGY The I'XttMit to whiih tlic land of the CovdilU'raii rogiou may have been elevated at this timi' i.s not known. Judging alone from the size of the glaciers produced, it must have hern much less than during the preceding maximum time of glaciation. In this case the coast region may still have remained submergi'd and the ixpper boulder-clay of that region may have bi-en produced as the ice iinally retreated. If, however, some general change toward amelioration of climate was in progre.ss, an amount of elevation e([ual to that of tlu> lirst great period of glaciation may have been required to produce the smaller glaciers ol'thti second epoch. Some indication of an elevation equal to or greater in anrount than that now held by the land is alforded by the notable absence of boulder-clay in parts of the larger valh-ys to the east of the Coast liauges.' By such an elevation the removal of this deposit in these places maybe accounted for, while the deposition of the upper boiilder-clay of the littoral may have occurred during the partial subsidence which inaugurated the deposition of the white silt formation of th(! interior, about to be referred to. To the latter part of this second and less intense, though possibly more protracted, epoch of glaciation \ve may with confidence assigu the origin of the deposits which I have designated in previous reports and papers as the " ivhile siltst.''' These were lirst examined in some detail in the basins of the Nechacco and Chilacco IMvers, in the northern part ol the Interior Plateau, but subsequent and wider observations appear to show that these silts possess a greater signilicance than originally supposed, and that they serve to mark out pretty delinitely a synchronous period of stability which extended to the entire northern part ol the Cordillera, and which succeeded a partial subsidenci^ of the lately re-elevated land. In thi> regions characterized by them, which are in almost all cases at a less elcA'ation than 2,r)00 feet," these white silts very often rest directly upon th^^ boulder-clay. They are generally line and irniform in texture and are usually well bedded in perfectly horizontal layers of an inch to two or three inches each in thickness. Where occasional sandy or gravelly layers are intercalated, these are attributable to local causes, being most frequently found opposite the mouths of valleys down which streams have flowed. In some places, and particularly in certain sections along the Upper Telly lliver, the layers have sulfered crumpling or disturbance, apparently from the a(;tiou of grounding ice, to which the rare occurrence of stones and boulders of considerable size may also be attributed- The silts have evidently be(>n laid down, as a rule, in tranquil water of cousid(M'abl(> depth, and their material has as obviously been supplied by streams or rivers discharging from glaciers not far removed. In physical characters the silts n^semble the deposits of the Ived Kiver Valley, though usually in theCordilleran rt»gion paler in colour, and seldom so clayey as some parts of those of the lied lliver. They dilFer from loess chiefly in their well-bedded character. It is believed that the general correspondence ^n eh^vation of the various and more or less separated bodies of water in which this white silt formation was formed, in itself constitutes a strong argument in favour of the hypothesis that these bodies of water were in direct commiinicatiou with the sea and were governed in their ' Tlie removal, more or less coniplole, of lionlder-clay from these vallej's can not be referred to tlio time of ixist-jrliu'ial nloviitioii sjiokoii of fiirtlier on, as llio wliito silt fermatinn is found well doveloped in tlmse valleys. •' In this statomiMil I finiit tlio cnnsideralion of a fwv instances of the local occin-reiice of si:nilar silly material in the southern part of the Interior I'lateau, whicii are not cdnnocted witli the main development of sills. These serve, however, to eonneol this main jieriod with an earlier time of greater Hooding of the Int«rior Plateau. / OF TIIK llOCKY MOUNTAIN JJWilON IX CANADA. 4B level by that which it held at the time. No traces of moraiuic^ or other barriers have been found in any ease sullicient to aeeount for the damming baek of water at the requisite levi'l, nor do the loeal eircumstames admit the supposition that such water was held in by glacier- dams. Had the silts been formed merely in lakes produevd in one or oth.-r of the modes last mentioned, they might be expected to occur, -n a region with such strongly marked features as that of the Cordillera, at a variety of very dilierent levels, in correspondence with circumstances varying in each particular cas'. The length of thi' perioil reijuired for the deposition of a great thickness of these line beds also all'ords reason for belief in the long epoch of stable conditions, aiul to some extent.justihes the presumption of the proximate contemporaneity of such a stable period. On following the silt formation town rd the various mouutiiin ranges and sources of local glaciers, it is almost invariably found '.o bo cut off somewhat abruptly before the mere increase in elevation would account fur its disappearance. This circumstance may with very little doul)t be attributed to the fact that during the deposition of the silts these upper parts of valleys weri' occupied by the still considerable local glaciers of the second period, which were engaged in producing the material of the silts. The (evidence is, further, conclusive that these glaciers in the end retreated pretty rapidly, leaving, in many cases, the long trough-like valleys which they had occupied almost ejitirely free from di-bris or detrital matter of any kind, and ready to become the beds of iiord-like lakes.' Assuming the general contemporaneity of the several developments of the silt forma- tion, and in view of the facts last stated, we are furnished with the means of ascertaining the extent of the glaciers, and ice-covered regions during a phase of the second glaeiation, which, however, does not represent the maximum of that period, but a prolonged pause in its decline. lu suppoit ot this belief, it may be stated that the indications thus arrived at agree well with the relative importance which we would on other grounds be justilied in assigning to the glaciers of the dift'erent parts of the Cordillera. For the purpose of illustrating the character of the evidence alforded by the silt for- mation, the principal areas of its occurrence may now be noted, proceeding from south- east to north-west. It is obvious, that in respect to the level at which the water stood at the time the silts were laid down, the important elevations are those marking the maximum heights attained by the deposits in question, in their larger areas, as the lower parts of the deposits may have been produced in water of vi-ry considerable depth. In the great Columbia-Kootanie valley, lying between the Rocky Mountains and Oold lianges and opening widely to the southward, the silts are well represented, with a thick- ness in some cases of fifty to one hundred feet as shown in terraces siuce cut through them, and a maximum observed elevation, near Upper Columbia Lakes, of about 2,700 feet. They are seen at lower levels down to 2,200 feet or perhaps less.- In the southern part of the Interior Plateau region, in consequence of the considerable height of the mean level of the country, the white silt formation is usually conlined to ' ' Report of Progress Geol. Siirv. Can.,' 1877-78, p. 153 B ; ' Quart. Joinn. Geol. Son.,' vol. xxxvii, p. 27.5. ■■' SiltyiloiKisils similar to tliose of tlie iniiin vulloy are fouml in limited quantity in tlio Koclcy Mountains to tlm east of it, in tlie narrow tributary valleys of tlio Kicking-Horso and Wigwam Kivons to a lieiglit of about \'.,'m feet. 1 am now inclinod to rofer tliese to an earlier iieriod tliau tbat of tlie main area of wliite aills, possibly even to tlio closing [K-riod of the lirst maximum of glaeiation. Their existence in tlinse mountain-valleys may, however, be accepted, in any rase, as evidence tbat the local glaciers devel(ji)ed on the Kocky Mountains projier during Iho second ixiriod of glaeiation were not of great size.—' Annnal Reimrt Geol. Surv. fan.,' 1885, p. 30 U. HH 46 (}. M. DAWSON ON TUB PIIYSIOGRAPIIirAL GKOLOGY the various troui^h-liki' valloys. It may be soeu from the line of the Canadian Pacific IJiiilway in charactoristic development, forming frayed terraces along the South Thompson valley ior a number of miles east of Kamloops. It is found also along Okanagan Lake, l)ut in the southern continuation of the Okanagan Valley, near the forty-ninth parallel, is ehielly represented by fine sands. Along the lower part of the Similkameen it has been observed in piitehes. On the Nicola it is often w^ell displayed. It extends down the main River Thompson to about the mouth of the Nicola, and appears again characteristically developed along the Fraser to the west of Clinton. It stretches far up the North Thompson, and runs back along the valleys of the main tributaries of this stream as well as of those previously mentioned. T' silt formation is often along these valleys a .striking feature, and is shown in terraces I'rom 100 to 200 feet or more in height. It is worthy of remark, that on the lower part of the Okanagan Valley, as well as on th(> lower parts of the main Thomp.son and Fraser, the silt is reduced in quantity and replaced by coarser arenaceous deposits, a fact tending to show the existence of rather strong current-action in i .-so main outlets of the plateau region. The silt formation is entirely absent from the upper portions of the valleys occupied by Adams Lake, Shuswap Lake with its various arms, the Arrow Lakes, and the northern part at least of Kootanie Lake.' Mabel and Sugar I^akes, lying l)etween Shuswap and Upper Arrow Lakes, have not yet been examined, but from their similar relation to the Gold Ranges they will probably be found to be equally free from the silt deposit. In previous publications I have stated the maximum observed height of the main deposits of these silts of the .southern part of the Interior Plateau at about 1,700 feet," but later observations show that they are developed to a notable extent at considerably greatt'r elevations. I will here only instance the valleys of Barriere River, a tributary of the North Thompson, Upper Nicola River below Douglas Lake, and Skuh-unh Creek, a tributary of the Lower Nicola, where thick deposits of white silt cut into terraces were observed at heights of about 2,250, 2,o00 and 2,450 feet respectively. It is thus probable that we may safely place the upper level of the silt formation in this region at about 2,500 feet, though it is still apparent that the more important developments of the deposit lie below 1,700 feet. This circumstance may be accepted as indicating that the region in question was subjected to elevation of a certain amount during the progress of deposition of the silts, the longest period of rest occurring at about 1,700 feet. The silt deposits are found in this part of the Interior Plateau down to heights less than 1,000 feet, but it is possible that some of the lower-level deposits have been secondarily formed from the denudation of the higher. In the Nechacco region, situated in the northern part of the Interior Plateau between latitiides 53° 30' and 54° 30', the white silt formation is very extensively displayed, covering an area of at least 1,000 square miles, with a thickness of 100 feet f.i some sections and probably exceeding 200 feet in certain places. The silts reach an < J.jvation of about 2,400 feet at the edges of the basin occupied by them, and where seen lowest (near Fort George) have an elevation of 1,900 feet. It is further possible that certain sandy deposits found near the sources of the Chilacco at a height of about 2,600 feet may ' The southern iwrtioii of tliis lake lias not yet l)een examined. '' ' Reiwrt of Progress Geol. Snrv. Can.,' 1877-78, p, 143 B ; ' Quart Joutn. Geol. Soc.,' vol. xxxvii, p. 275. 4 OF THE KOCKY MOUNTAIN IIKGION IN CANADA. 47 represent a littoral condition of the silts.' Some remarkable shingly beaches Ibund on the slope to the north oi'Tsa-whuz Mountain, at a height of about 2,100 feet, were taken to represent a shore-line of the Nechacco white silt lake or sea when at a lower stage.-' On the north side of the Nechacco basin, the silts go no further than the lower end of Stuart Lake, ceasing there at au elevation of aboiit 2,000 feet, and they are not seen along the upper part of this lake or on Babine Lake. The shores of Franyois Lake, to the west, (2,87o feet) are also free from them. If not interrupted in some unknown manner, the valley of the Fraser liiver must at this time have constituted a free connection between the basin of the Nechacco silts and the previously described development of the same silts in the southern part of the Interior Plateau ; while to the north-eastward, water standing at the level implii'd by the Nechacco silts must have extended by way of the Peace liiver gap through the Rocky Mountain Range. Though carrying us for a moment beyond the Cordilleran ri'gion proper with which we are at present occupied, we may here note that a great area of plateau country to the west of the llVth meridian, which is now cut through by the deep valleys of the Peace and its tributary the Smoky River, is covered with silty deposits precisely similar to those of the Nechacco basin. These generally rest upon the boulder-clay of the region and are embraced between elevations of 2,500 and 2,000 feet.' Though much less important in respect to area, the silty deposits of the Stikine and Tanzilla valleys, between latitudes 58° and 58° 30', may next be mentioned. These are sufficient to form a wide area of flat country, at a level of about 2,200 to 2,300 feet, and are scA'eral hundred feet in thickness.' The material is here somewhat darker in colour and more clayey than elsewhere noted, but corresponds in its regular bedding and in the abundance of calcareous nodules which it holds with many typical examples of the white silt formation. In descending the Dease River, typical white silts are first met with about latitude 59° 80' at an approximate lieight above sea-level of 2,400 feet.'^ In the lower part of the Dease Valley, and in that part of the valley of the Liard immediately above the mouth of the Dease, the silts pass into, and appear to be represented by, sandy and gravelly deposits, but on the last-named ri ver near the mouth of the Frances they become again well developed at a height of about 2,300 i'eet." No opportunity occurred, however, of ascertaining their highest level in this valley.^ ' ' Quart. Jonrn. Geol. Sc ,' vol. xxxiv, pp. 105-107. ^ ' Quart. .lourn. Geo!. Soc.,' vol. xxxiv, p. 110. •' ' Keport of TrogressXieol. Surv. Can.,' 1879-80, p. 142B; 'Quart. Journ. Geol. Soc,' vol. xxxvii, p. L'77. It may be added here, as at least siiftgestive, that over a great part of the area drained by the various trihutarifS of the Saskatoliewan, at similar elevations, the upj)er member of tlie glacial serie.s is constituted by similar silly beds. See ' Report of Progress Geol. Surv. Can.,' 1882-84, part C ; and ' Annual Reiwrt Geol. f-^r\. Can.,' 188G, part E ; ' Bull. Geol. Soc. Am.,' vol. i, p. 403. ' See illustration ' Annual Report Geol. Surv. Can.,' 1887-88. p. 67 B. The level here given is that of the main surface of the deposit, the height of its borders was not precisely ascertained. ^ ' Annual Report Geol. Surv. Can.,' 1S87-88, p. 90 B. « ' Annual Report Geol. Surv. Can.,' 1887-88, pp. 101 B, 103 B. ' Though not closely connected with the present discussion of the white silt formation, it may here be noted that the principal terrace fotmd along the shores of Dease and Frances Lakes, (forming the sources of the Dease and Frances branches of the Liard,) were found to be at the level of 3,180 and 3,200 feel respectively. These lakes are 48 f;. M. DAWSON ON TIIK IMIYSIOGKAPIIJOAL (H'.OWKiY Ciossiii'j liiially lo llii' l)iisin of thu UpiHT Yukon, \v«' iiTavel. They eontinuc !ilonuf?th, and are I believo sulliiioiit to indicate that Iht; whiti' Hills con.stituto a dolinite formation, and mark an important period in tho history of the glaciation. TAUI.H SlIdWINfl TlllO XdllMM. UlTKn TjKVl 1, (U' 1 UK WlllTi; SlI.T KdllM \TION IN V.MIIOCH I'AIiTH OK Till: ColiDlI.I.KRAN KtOlON, WITH SO.MI-; OUSHUVATKiNS on TKH LoWKU limit llF Till; WAMU I'OliMATION, AND InSTANCRS ok 1,0 couuectiou with the examination of dili'erent parts of this c a-nsive region during the past fourteen years, and their inter-relation has only by degree l.ecome apparent. Most of the heights assigned in the above tabic; iind on previous pages depend on barometric results, l)ut as these have been carefully checked and are in many cases means derived from a consider- able number of readings, they may safely bi' accepted as sulliciently rorn>ct to bo employed in a general view of the subject. The bodies of water in which the ..silts were laid down must in some cases have been bordered by sandy or gravelly beaches, which doiibtless passed in depth mo ^ or less gradually into truesilty deposits; and +hus it is that, while the main developments of the silts are striking and apparent, the precise position of their highest margin is not always delinitely recognizable, even where the cir(5umstances of observation are otherwise most favorable. Allowing, however, for these sources of uncertainty, there appears to be a greater amount of dilference between the highest levels of the main silt areas than can thus be accounted for. Though generally speaking contmnporancous, it cannot be supposed that the silts of the ditferent areas wta-e absolutely coeval, nor that the changes in level of so vast a tract of th(> Cordillera W(>re produced with perfect equalit .-. It seems probable, in fact, that the southern part of the Interior Tlateau was subjiicted to a differential elevation of sevtM-al hundred feet after the main time of deposit of the silts began, while the Columbia-Kootanie valley and the Upper Yukon basin, appear to have stood at an elevation of about 200 feet less than the mean level at the time. See. IV, 1800. 7. 80 (;. M l>A\VSON ON Till-; IMIVHlv^'JICAI'IIICAL <}i:oriO(iY It is ttiint Nsiiry in thiH papi-r to fiitiT into dt'ttiil respcctinf? tho cxti'iit of thti viiriouH mure or It'ss local f^IiuitTH ol" tilts second pi'.iod, at tho tinu^ at which their iinal ic'rcat commenced and tln' deposit of the white silt I'oKnalion ended. TIk^ distriljutiou id' these lilaciern coiil'orined cjoNcly to |1.«- tsecoiKlary orou'i'iiphic lealures cl' the ('.>;dillera, and was tiius too inliii ate to he kIioW!; on a .^niall map, and isliesides, as yet only partially known. It would appear that a reduced representative of tho Cordilleran ylaeier, prohahly more or loss hrokcii up and divided, remained aboui the position oi' tho lacicrs about the axial ref>'ions of thi> (Jold coast and other ranj^es, acting throughout tho long l)eriod whii'h the deposit ol'the while silt Ibrmalion implies, may he called in to a fount I'or the great depth of the lakes and iiords which occupy the more important valleys in thisse ranges. It may in I'act, on this hypothesis bo assumed, that in consotpionco of a depression a'le •'■!;' the axial parts ol' such ranges, tho i)re-existing valleys within thoir borders, became depi'ssed in their upper parts. Without some such exidauation, and unless we are prepared to admit that a greater part of tho t^xt^avation of these valleys was produced by the wearinn- action of Ww glaciers themselves, tho existence of some of them is very puzzling. Thus, to cite, only two instances, — Adams Lake, in the southern part of tho (Jolil lianges, is known to bo over !iOO feet deep in its upper portion, and the Upper Arrow Lake has a depth toward its head of more than TOO feet ; depths which in both cases much ei'ceed the probable depth of aay buried channel of oulllow. The phenomenon here all.ided to is, however, one of such frequent occurroucc in this part of tho Cordillera as to imply somo general caus(>, and is doul)tless the same with that which may bo found applicable to similar lakes mi other parts of tho world.' Tho abovt; solution is therefore merely noted as a possible ou<., and it will be observed, that in exact proportion tt) tho amount to whi(;h s\ich axial depression of ranges may be admitted it must bo considered as weakening that part of tho evidence already drawn from the depths of the fiords of the coast rt'gion, in respect to the amount of tho later Pliocouo and early glacial elevation of the Cordillera. The final decrease of the ice must apparently have been duo to some general cause connected with thi? close of tho Glacial epoch as a whole, for simultaneously with the ultimate retreat or disappearance of the glaciers, the elevation of the Cordilleran region was resumed, tiP it eventually reached approximately tho level which it now possesses. One feature to which I have called attention on several former ot;casions appears, however, to require explanation in connection with this final movement in tdovation, namely, tho almost complete absence of terrace-deposits on the seaward slopes oi the Coast llanges as well as in tho valleys of most of tho streams and rivers with short courses which drain these seaward slopes or pass through them. This circumstance is believed to dopeud on i • It is not the intention here to include all rock basins now containing lakes, hut mere'y those which seem to imply aroversalof sloiiein niountain-vnlleyB. The fact that diflferontial changes in elev itiou cannot explain all rock-basins, does not preclude the application of such a hypothesis to siiecial ca^jos. Cf. Prof. J. Geikie in "The Great Ice Ago," 2nd edition, p. 275. OF tin; ROCKY MOUNTAIN IlKdiON IN CANADA. 81 tho fact that thf gliiciorH of this raiifjc \v<'ri' still iil)li' to follow tlu' rt'troutiuf? hnnliT of tlio «t'a. liiliiij^' tho vallt-yM and covfrinjf thn HlopcH as tln'y ciin'ri^cd and svvot'piiiif Ix-foro thtiiii Hiich dctrital deposits as had hi'on fornifd, their action l)i'iiially the earibou, and 1 have been shown by him the skin and antUirs of one of these animals. The caribou is not now found any- where else in ihe region of the coast, either on the islands or in the Coast lianges, though it rooms over high plateaux to the east of these ranges. The shortest distance between any point of the Queen Charlotte islands and the nearest islands of the Coast Archipelago is thirty miles, and the intervening strait is subject to rapid tidal currents. The isolation of the (^ueen Charlotte Islands is in fact so complete that the dei'r, which inhabits all the other islands of the coast, is not found in this group. It is, therefore, in the absence of the caribou from the neighboring coast and its adjacent islands and in consideration of the width of the waterway whiiOi would have to be crossed, at least highly improbable that this annual reached the Queen Charlotte Islands under the present conditions. I arti thus led to bilieve tiiat the caribou colonized he islands at a time at which either the giaciers extending from the mainland attained to the Queen Charlotte Islands, or by a laud con- nection during a period of greater elevation.- Thi' latter is in every way the more prob- able supposition, and if it l)e entertained it may further be assumed thai, the animal came to the islands at the date of the immeJiately post-glacial elevation above indicated, and that it has since, &^ an 's(dated colony, succecHled in maintaining itself there. As ihe height of tho north"rn part of the Cordillera i)i: leased, after the period of rest marked by the white silt formation, the streams of the mainland were enabled to begin the cutting out of new channels in tho glacial debris with whicli their valleys had become more or less choked. In consequence of the strongly marked relief of the country, the principal vivcrs and larger streams were seldom forced by thi; glucial accumulations to leave theiv old valleys. Though the amount of this post-glacial river erosion is consider- able, it may be regarded as insignilicant in comparison with that of the great Pliocene period of denudation. To include any treatment of this subject here would, however, involve -appeal to an inadmissible amount of local detail. I will therefore state only, in general terms, that a large proportion of the rivers have up to the present time been unable to cut dov.'n to their pre-glacial beds; but that in other places this has been accomplished, and they are now engaged in more or less active erosion of the subjacent rock. Ou thai part of the Fraser Iviver between Lytton and Big-Bar Creek— a length of eighty \niles — the jjost-glacial excavation by the river varies in depth from 400 to 800 fet t, in accordance with the evidence of terraces still remaining in the valley. In several places the old channel has been locally departed from, and narrow canons with depths of 200 feet or more, have been cut across points of solid rock, and in some places (as betv/een Bridge River and Fountain C!reek), the Fraser has evidently regaimd its pre- fflacial level and is now at work in still further deepening the old channel. The condi- tious met m ith on the corresponding part of the length of the Main Thomi>son liiver are very similar. Having thus endeavoured to trace in their sequence the events of the Glacial epoch in the Cordillerau region, we may now consider briefly, and with that which has already • ' Kerx)rt of P- ^ktvbh Geol. Surv. Can.,' 1878-70, p. 113 15. '' Tl'.e miuimuiii amount of elovation reiiuirod woulil l)o about 200 feet above tbo iiresoul level. OF THE EOCKY MOUNTAfX EE(fION IN CAXABA. 53 been ascortainod as a basis, the coeval conditions in the adjoining- area of the Great Plains. The watershed of the Ko.ky Mouniain liaii^-e proper, has always constituted a physical and to some extent also a mental barrier and line of division l)et\veen our invesiigatious pursued to the east and west of it, and as already pointed out, this is in a measure justilied by the small amount of apparent connection or overlap of the evidences of the GHacial epoch of the two regions thus marked out. It is necessary, how'ever, to suppose that the occurrences of this period were to some extent homologous in both areas, or in other words, that if not similar and contemporaneous they were at least correlative.' In the earlier stages of the investigation of the phenomena of the glacial epoch, it was as a rule mentally postulated that in endeavouring to account for these phenomena we wen^ entitled to consider only such changes in elevation as are pr:\ctically of continental extent, or such as may have aflected the northern hemisphere as a whole. 01' Kite years, evidence has been accumulating on all hands, of diil'erential changes in elevation which, in time? very recent geolo^-ically, have affected comparatively limited areas. I do not propose to enter into any discussion of this evidence, or of the tlieories by which such clianges may be explained, but have found that in admitting their existence, it is possible to arrive at tenable hypotheses leading to reasonably satisfactory conclusions respecting the glaciatiou of the western part of the continent. In the history of the northern part of the Cordillera in Meso/oic and Tertiary times, given in the first part of this essay, we have already found indications of *hc- complemen- tary character, from a physical point of view, of the areas of the Cord:lle;a and the (iroat Plains. There is an antecedent probability in the belief that the elevation of a wide belt of country such as that comprised in the Cordillera, would lead to a correlative depression in some adjacent region, and though this compensating change may here be largely accounted for by supposing the occurrence of upward or downward movements in some not far remote part of the bed of i! i Pacific, respecting which we are without evidence, it appears to have been affected t- variable and occasionally to a very great degree in the area of the plains. I would refer particularly in this connection to a suggestive discussion and summary of correlative movem(mts in elevation and depression by Mr. Warren Upham, wliich appears in the form of an appendix to Prof Wright's " Ice Age in North America.'" This discussion appears to holdout the prospect of a solution for many of thedilhculties which have so far attended the explanation of the facts of the Glacial period, ])y putting in a con- crete form a series of propositions, which though doubtless j)resent in a more or less recognized manner to the minds of many geol.)gist.s, have been l)y them but little enqiloy ed as implements of research. Havin"' thus alluded in general terms to the fundamental conceptions involved, I may present the following tentative scheme of correlation of the principal events of the Glacial period in the area of the Cordillera and in that of the Great Plains, which I have ' Mr. T. C.Russell inliis "Notes on tlio Surface Geology of Alaska," refers lo the freshness of glacial striation on Lake Labarge, with other facts, as possible evidence of much greater recency in the glaciatiou of the Ooniillcra as compared with that of the oasleru purl of the continent. ' Bull, CJool. Sue. Am.,' vol. i, p 142. My oh.servations in the northern parts of both regions do not lead me to concur in this iuggcstiou, which in view of all other circumstances seems so improbable. ■^ Apiwndix A. " Probable Causes of Glaciation," hy Warren Upham. ssaam 84 (;. M. ])A\VSO\ ox Tin-: PHVSIOGI?ArillCAL GEOLOGY ])('fii led to iidopt as at least a tonablo working hypothesis, after mature consideratiou of the u\ aihihlf cvidcncr relating- to both those great areas, and in the light of personal f'ainiliarity with the eir.umstances in both. In giving to this scheme, the defiuiteness implied by a tabular arrangement, I would, however, premise that though for the pur- pose of making it clearly understood it is necessary to adopt this form, there is no inten- tion of presenting it as a complete or final solution of the problem under discussion. In view of the multiplicity and more or less indeterminate character of many of the observa- tions which it has been end(>avoured to embrace under this generalization, as well as the vast extent and diversity in leatures of the region to which it relates, it is scarcely pos- sible to hope that it may not eventually prove to be in error in some points. It appears however at present to agree with the known conditions and to indicate at least the mode of their explanation. f-'oilHMB 01- COUUKI.ATIOX M' TIIH riIi:X0M13XA OF TUB Gl.AdAI. rKUlOD IS THE CoKDlI.LEUAN' UEdlON AND TUB lihxiioN 01' THE Gkkat Plains. dmlilh'rdn Rti^ii»i. Cor<1iII«riin /one at a liij;h iilovati'>ii. rcriod of most sovoro jrlaciiiUoii iiml inaxiiimni devolopmeut of the great Cordillerivn jilacier. (irailiml siibsi(len((? of tli(( f'ordilleran region and decay of the. great glacier, with depusition of the bouldor-day of the Interior Plateau and the Yukon P>asin, of the lower boulder-clay of the littoral and also at a later stage (and with greater subniorgonce) of the iiiterglacial silty beds of the same region. Ke-elevation of the Cordilleran region to a level prob- ably a.s high as or somewhat higher than the present. Maxinnim of .second period ofglaciation. Partial subsidence of the Cordilleran Region to a level about '.',60(1 feel lower than the present. Long stajje of stability, during which the H7iiV< Silh were laid down, (.ilaciers of the i^ocond period ecjiisiderably reduced. ITpfier hoidder-cla\' of the coast probably fornK^d at this time, though perhaps in part durinj; the second uiaxi- mmn ofglaciation. Kenewod elevation of the Cordilleran region w ith one well marked pause, during which the littoral stoixl aliout 200 feet lower than at present. Glaciers much reduced and diminishing, in consociuence of general amelioration of climate toward tlie close of the Glacial jieriod. Region of Ike Great Plnins. Correlative sulisidenco and submergence of the Great Plains, with possible contemjjoraneous increased eleva- tion of the Laurentian axis and maximum develop- ment of ice upon it. Deposition of the lower boulder- clay of the j)lain8. Correlative elevation of the western part of the ( ireat Plains, which was probably more or less irregu- lar and led to the production of extensive lakes in which inter-glacial deposits, including j»at, were formed. Correlative subsidence of the plains, which (at least in tlie western part of the region) exceeded the first subsidence and extended submergence to the base of the Rocky Mountains near the forty-ninth parallel. Formation of second boulder-clay, and (at a later t-tagei dispersion of large erratics. Correlative elevation of the plains, or at least of their western portion, resulting in a condition of equilibrium as between the Plains and the Cordillera, tlieir rdalive levels becoming nearly as at present, l^robable formation of the Missouri Coteau along a shore-line during this period of rest. Simultaneous elevation of the Great Plains to about their present level, with final exclusion of waters in connection with the sea. Lake Agassiz formed and eventually draioed toward the close of tliis period. This simultaneous movement in elevation of both great areas may probably be connected with the more general northern elevation of land at the close of tlie Glacial period. 1 ! OP THE EOCKY MOUNTAIN REGION IN CANADA. 55 A Referring to the several correlative movements of elevation and of depression of the Cordillera and the Great Plains implied by th(! hypothesis set forth in the above com- jjarative scheme, it may I think be admitted as not improbable on theoretical grounds, that such conditions of oscillation once set up, in consequence of the interact! n of what- ever causes, might have a tendency to continue for a considerable time before a stable condition was regained ; a state of equilibrium being in the end aitained either by the general decrease in intensity of the forces involved or by the final preponderance of one class of these. It may further be pointed out, that the supposed sequence of events is generally in correspondence with the view that the periods of maximum glaciation of the Cordillera were those of its greatest elevation, while the decay of its great glaciers was in both instances accompanied, if not caused, Ijy subsidences leading to the encroachment of oceanic waters. The flooding of the Crreat Plains by arctic water, while the Cordillera stood as a much elevated land between these and the warmer ',vaters of the Pacific, would in itself go far to explain the conditions under which the excessive precipitation required for the production of the Cordillcran glacier occurred. The supposed sequence of events is further.aorecompatible with the belief, which has been advanced by several g< logists, that the weight of a ponderous ice-cap is in itself sufli- cient to produce subsidence of the laud, and with the idea that such an ice-cap may thus eventually become self-destructive, by obliterating the elevation to which its existence is in the first instance, largely due. It is also of interest, though as a minor point, to n^ie that the extensive orcurrence, in the interglacial beds in the area of what is now one of the most arid parts of the north-western plains, of peaty deposits aboundinu- in the lemnins of ferns and semi- woody plants, ' maybe accepted as betokening a ■ nusiderable local rainfall; a circum- stance which would be in accord with the postulated depression of the Cordillera and relatively elevated position of the Groat Plains at that time. Though independently based upon and primarily iiil nded to in.'lude the observed phenomena of the G-lacial period in the north-western j- .vtion of the continent alone, it is further worthy of remark, that the elevation believed to have alfected the Canadian Great Plains during the interglacial episode, is to some extent confirmed by the fact that Messrs. Chamberlin and Salisbury find evidence of a similar elevation of the Upper Mississippi Valley, at a corresponding time, to an extent of about 1,000 feet.- It is also noteworthy that the two great correlative movements of elevation of the Cordillera and depression of the Grent Plains here admitted, correspond, at least in a general way, to the principal similar, though smaller, changes in level accepted by Mr. W. J. McGee as explaining occurrences on the " Middle Atlantic slope," as indicated by the Columbia forma- tion.'' As in the case of the Great Plains, the two marked changes in the region of the Middle Atlantic slope were in the sense of depression, but as there is reason to believe that depression in one region may have been made up for by elevation in another, the main point to which allusion is here made, is merely that both maxima of glaciation appear to have been marked on the Pacific as well as on the Atlantic side of the continent, by con- siderable disturbance in level. In previous publications on the glaciation at the Canadian portion of the Great Plains, • 'Bull. Geol. Soc. Am.,' vol. i, p. 332. " " Seventh Annual Report. U. S. Geol. Surv.," p. 039. 'Sixth Annual Report U. S. Geol. Surv.," p. 214. 36 fi. M. J)AWSON ON Till': PIIYSIOGRAPIIICAL GEOLOGY dealing only with the ovidenco available at the time, which did not include any con- Kid(!rablt' body of facts relating to the Cordillera, I had arrived at the opinion that, as a wholi', this glaciation w^as attrihutal)le to the action of water-borne ice rather than to that of the icts of a great confluent glacier. At the same time in deference to the views maintained by many geologists, I pointed out in Avhat manner the circumstances might ])e explained under either hypothesis and noted the difliculties incident to both.' At this time I fully recognized the fact that the Laurentian axis had been buried beneath a con- fluent glacier, and the evidence upon which the existence of a similar confluent Cordil- leran glacier is based, as noted in previous "pages, has since been obtained. ■ Later and more detaih'd and systematic^ examinations of the area of the Great Plains, carried out with a knowledge of the views of other geologists on the glaciatiou of different parts of the northern hemisphere, have, however, resulted only in further confirming my belief in the glacio-natant origin of the glacial deposits of these plains. I may be excused for stating here that I have had every opportunity of becoming familiar with the effects of confluent glacier action in regions to the east and west of the plains, my apology for making this statement being found in the circumstance that the attribution of all the widespread and important phenomena of the Glacial period to the action of a confluent glacier (a lineal descendant of the polar ice-cap) has become almost official ; with the natural if scarcely conscious tendency, which for a time appears to follow the acceptancHi of any wnde generalization, to ignore or doubt any outstanding or unconform- able observations rather than to appreciate these at their full value and to follow them up as useful clues toward a position of fuller knowledge of the phenomena as a whole. Some reasons for the opinion held by me that the effects met wath on the plains gener- ally must be accounted for by floating ice, are alluded to on a later page. It may here be noted, however, that the theoretical mode of accounting for the gla- ciatiou of the (ireat I'lains which appeared to me most probable at an earlier date, and which was discussed in the publications just alluded to, has since become subject to con- siderable modifications, principally by reason of the discovery of boulder-clays of two distinct periods and the definition of certain intervening interglacial de^wsits. Here, as in the Cordillerau region, the existence of evidence of the action of water at high levels might be simply referred to the occurrence of a si)igle and contemporaneous time of great depression of the land, but though the existence of such evidence at heights of between 4,000 and 5, .SCO feet in the Cordillera and on the eastern slopes of the Rocky Mountains, may be referred to as favoring the theory of a general depression of this kind, it is ' " (teolngy and Keseiiiros of Uio HoKion in the vicinity of tho Forty-nintli riirallel " (1875). ' Quart. .Tourn. Gool. !S(ji'.,' v(il. xxxi (ISTo). -' Jt beinti now duiinitely known, as a result of work done under the auspices of the Gooloj^'ical Survey of Canada and more jmrticularly of observations by myself and my colleagues Messrs. McConnell and Tyrrell, that tfie extreme borders of the eastern and western pilaciated areas of the continent barely overlap, uiul then only in a limited rej.'ion ; it becomes manifestly necessary to recognize two distinct main centres of glaciatien, both of which were, at the maximum (jf the glacial conditions, more or less completely covered by great confluent glaciers. Cf. 'Geological Jlagazine,' Dec. iii, vol. v, p- 1'50. It is therefore no longer apiiropriato to refer to the great conlluont ice-nia.'-s which spread out from the region of the Laurentian axis or plateau as the " Continental Glacier," even if it be supposed that it extended as far as the western e.lge of glacial deposits on the plains. In view of this fact and in tho interest of precision of expression, I venture to suggest that this may be named the LaurcrUidc Glacier. I'or the great ice-mass of tho western regiou I have already proposed tho name Cordillerau Glacier. OF TUI'] ROCKY MOUNTAIN IJMGION IN tJAXADA. 57 + believed that the proof of the hig-h level of tlit^ Cordillera and the relatively low eleva- tion of the plains at tlie initial stage of the (ilacial period, constitutes an insuperable dilli- eulty in the way of a simple explanation based on sxnh a supposition. As the scope of this paper does not ineludc any extended description of the glacial phenomena of the Great Plains, I will here only outline, in the most general terms pos- sible, the character of thi' observed facts ' The surface of that part of the G reat Plains which is characterized by deposits due to the Glacial period, is at the present time hii^'hest at the base of the Rocky Mountain Range near the forly-niuth parallel, having there an eh'vation of about 4,000 feet above the sea, and sloping thence in a north-eastward direction to the base of the Laurentian axis or plateau on the east (distant, about TOO miles), with an average fall of about 5'4 feet to the mile. Further north, in the Peace and Liard River countries, the width of the region representing the plains ))etweeu Ihe Rocky Mountains and the Laurentian plateau is mucli less, but it still preserves an easterly slope, the highest plains to the east of the foot-hills of the Rocky Mountains being at an elevation of about 3,000 feet in the first mentioned district and still lower in the second. The maximum elevation attained by the Laiirentian plateau in the corresponding part o.f its length is vxnder 2,000 feet, and probably averages about 1,(')00 feet only, and from the rocks of this plateau a considerable portion of the whole material of the drift, even of tliehighest levels, has been derived. The glacial drift reaches quite to the foot of the Rocky Mountains, near the forty- ninth parallel, but soon leaves them to the south, while further to the north its margin is constantly found at a somewhat lower level and at some distance from the base of the mountains. Disregarding exceptional areas of small dimensions, from some of which the glacial drift has been more or less completely removed by denudation, and, in a few eases, upon which it has never heen deposited, the whole of this vast tract is covered with remarkable uniibrmity and with little regard to the elevation by di^posits referable to the boulder-clay. Overlying these, stralified sands or gravels are almost every where found in continuous and connected spread, with, very frequently, bedded silty deposits ; some of which are evidently due to post-glacial lakes, others apparently to more general causes. Large erratics are in all parts of the region observed to be most abundant on the siarface, or in connection with some of the upper superficial deposits. In the western part of the plains, ])ut so far as ascertained to the east of the extre ne western margin of the drift deposits, the boulder-clay is found to be separable into two parts, between whicli interglacial silty and sandy deposits, which in some places hold peat, are extensively developed. It is as yet uncertain whether the division thus evidenced as between the lower and upper boulder-clays extends to the eastern half of the plains, the area over which it is known to be Avell marked, stretching from the vicinity of the forty- ninth parallel northward to th(> North Saskatchewan River, with a length of about 250 miles and a maximum actually known width of about fifty miles. ' Tor dotrtils of observation in tliLs urea reference may be niaile particnlarly to "Cleology and Resouri^gs of tbo Forty-nintb I'arallol," pa^'e 201!, '(Jnart. Journ. Geol. Soc.,' vol. xxxi, p. ti03, 'Report of I'rogre.ss Oeol. Surv. fan.,' 1879-80, p. 13() H., ' Report of Proiircss fieol. Surv. Can.,' 1882-84, p. 139 C, ' Annual Report Geol. Surv. Can.,' 1885, p 71 C. (R. G. IStcConnell), • Annual Report,' 1880, p. 139 K. (.1. 1!. Tyrrell); also to 'Bull. Geol. Soc. Am.,' vol. L, p. 395 (J. B. Tyrrell), and p. 540 (R. G. McConnell.) Sec. IV, 1890. 8. 38 G. M. DAWSON ON TUE PlIVSIOfJllAPIIlCAL GEOLOGY The relation of tho low»>r boiilder-(;lays to certain underlying and immediately pre- glai'ial gravels has been referred to on a ibrmer page (see p. 22.) In accordancf with thf gi'neral hypothesis here advanced, it may be supposed that the submcrgenc(! of the- plains during the iirst period of glaciation (that of the maximum of the Cordilleran glacier) extended as far westward as the limit of the lower boulder-clay. Between the forty-ninth and liftieth parallels it must in this case have reached to a posi- tion forty miles or less from the base of the Ivocky Mountains, .and further north to an undetermined line, which, however, in all cases appears to have fallen short of the maxi- mum Avestward spread of the glacial drift.' The interglacial episode seems to have been marked by the elevation of the western part of the Great Plains, accompanied liy the production of shallow lakes, in or around the liorders of which peaty dt'posits were locally formed. The width of the region raised above the sea at tliis time also remains indeterminate, and no evidence of this episode has yet been obtained to the north of the North Saskatchewan. It appears probable, if the r(>gion of considerable uplift extended much further to the northward, that it may have been there narrower and less important. In connection with the description of the Missouri Coteau, I have noticed the existence of systems of now more or less completely abandoned old valleys cut in the Laramie plateau of the vicinity of Wood Mountain." These have been explained as pre-glacial valleys, but in view of the additional light on the sixb-division of the Glacial period since obtained, it is now believed that the greater part of the erosion of these wide old valleys in the higher portions of the Great Plains, both in the locality heie particularly cited and in the Cypress Hills Plateau and elsew'here, may probably be assigned to tlie interglacial time of elevation. This would accord better also with their fresh and unbroken outlines, which, in attributing them entirely to pre-glacial action, are difficult to understand. Before leaving the subject of the interglacial deposits of the western part of the plains, it should be pointed out that, while they indicate that this was in part a land- surface at the time of their deposit, they also afford a certain amount of definite evidence showing that at the date of their formation, this part of the plains had not had impressed upon it the slope from west to east which it now possesses. This is shown particularly along the Belly River, where interglacial beds evidently formed in a shallow lake, may be traced continuously for forty-five miles from west to east. These beds may safely be assumed to approximately represent a Avater-level of the time, but their position is now no longer horizontal, but conforms instead to the slope whiih the plains now possess.' To the renewed subsidence of the Great Plains wdiich is supposed to have been contemporaneous with the second elevation of the Cordillera, and correlative to this, is attributed a second access of floating ice, derived chiefly from the front of the confluent glacier of the Laureutian axic (great Laurentide glacier), to Avhich the deposition of the upper boulder-clay is due. We are as yet unable to form any dt finite opinion as to the extent to which the action of such floating ice and the continued ieposition of boulder- clay was limited in the eastern part of the plains during the interglacial episode. It would further appear, that while the elevation of the Cordilleras belt was probably ' i.e., The aviiiliU)lo evidonco at least does not carry ;'. so far. '^ " Geology and Kesources of the Forty-ninth Parallel," p. 230 Qnar^. ,^ouia. Geol. Soc.,' vol. xxxi, p. 615. ^ ' Report of Progress Geol. Surv. Can.,' 1882-84, p. 151 C OF THE EOCKV .MOUNTAIN KEGION IN CANADA, 39 uot greator aud possibly not as groat as during the iirst period, the extent to which the western part, at least, of the plains was depressed was in excess of that by which they had been alFected before. Various conjectural explanations of this might bu hazarded, but in the abs(>nce of any complete knowledge of the circumstances, these would pos- sess little value. It would appear, however, that the ice-bearing water at this time reached quite to the base of the Rocky Mountains near the forty-ninth degree of latitude, bringing with it debris derived from the Laurentian axis, of which the nearest partis over 500 miles distant. ' It does not of course necessarily follow that such debris was actually derived from the nearest possible source, though in previous piiblieations I have spoken of it as thus originating in order to simplify the conceptions involved. Recent investigations by Mr. J. B.Tyrrell, have in fact shewn, that the direction of move- ment of the ice over the eastern part of the plains may gemn-ally have been from north to south or from north-east to south-west," which might imply an even greater distance of carriage for the erratics here particularly alluded to. Many of these erratics along or near to the base of the mountains between the forty- ninth and fiftieth parallels, lie at heights exceeding 4,000 feet, while the highest observed instances of tlicir occurrence, are at an elevation of 5,280 feet above the present sea-level,' the erratics being hen^ stranded upon moraine ridges due to local glaciers which have llowed out from the valleys of the Rocky Mountains, probably during the first maximum of glac'iation. From the sam(> south-western portion of that part of the plains which is characterized by glacial deposits, further evidence, shewing that the elevation succeeding the second subsidence was unequal in character, is afforded by the difference in heights of the upward limit of Laurentian material in a few conspicuous instances. Thus, taking a belt of country running eastward from the base of the mountains and approximately included between the forty-ninth and fiftieth parallels, we find the furthest western erratics at the great elevation just quoted. Seventy-five miles to the east, the summit of Rocky-Spring Plateau is driftless, the upward limit of the eastern drift being found at 4,100 feet. Twen!y-five miles further east, the limit reaches a height of 4,660 feet or possibly slightly more on the Three Buttes or Sweet-Grass Hills. ' Fifty miles still farther to the east, M'-. McCouuell has ascertained the limit to be at 4,400 feet on the Cypress Hills, the higher pari of the plateau so called being again driftless. One hundred and sixty-five miles due north of the Rocky-Spring Plateau, Mr. Tyrrell has found the summit of the Hand Hills to \^^ similarly wdthout glacial debris, the limit- ing height at this place being 3,400 feet. ' The localities thus enumerated include all the known driftless areas within the western '^jorder of the drift-covered area of the plains, aud it Is probable that there are uot any other instances remaining to be discovered. ' As indicated on the ma]) wliicli accompanii's luy first paper on tlie silaciation of tlie plains, tlie actually nearest part of tlio edjre of the Laurentian plateau, on a noith-oast bearing, is about 550 miles distant. The dis- tance of 700 miles given in that paper, refers to that part of the Laurentian niar;.:in between E. and E. N. E., from which alone it was at the time (in conformity with indications of direction obtained from striffi in the Lau- rentian region), snpposed that the travelled material of the part of the plains in que.stion was derived. - ' Bull. Geo). Soc. Am.,' vol. i, p. 40L ■' 'Keport of Progress Geol. Surv. Can.,' 1882-84, pp. 140 C, 148 C. * ' Eoport of Progress Geol. Surv. Can.,' 1882-84, p. 148 C. 6 ' Annual Report Geol. Surv. Can.,' 1885, p. 75 C. 60 (i. M. DAWSON UN Till': I'JIYSIOCiJlAPllIUAL GKuLO(;Y Jt is worthy of reauirk tlmt the Three Buttes or Sweet-Grass Hills form lui isolated case of the occurrciict' of volcanic roi.'ks hi this part ol' the plains, representing the stumps of volcanoes wiiich are referred to the Miocene period (see p. lo), and occupying thts centre of a wide low anticlinal, the inlluence of which has now hi'en traced northward to the North Saskatchewan. The greater amount of elevation by which the upper limit of the glacial deposits has been all'ected in the central region of this anticlinal swell, as compared with tlie instances to the east and west uf it, together with the still greater height to which tlie same limit has been lifted on the Hanks of the liocky Mountains in the same latitude, is very suggestive of the inlluence of tangential pressure in the earth's crust, which has had most elfect in increasing the elevation of old lines of uplift. Iloeky- Spring I'lateauand the Hand Hills may be considered as about on the nodal line between this anti.linal and the wide synclinal which separates it from the base of the Rocky Mountains. The depth of submergence of the general surface of the plains surrounding the Cypress Hills and Three IJuttes may thus lie sinM>osed to have averaged about 1,500 feet, and to have exceeded this by oOU leet or more ai a distance of al)out liity miles to the north where the general heiglit of the country is less.' Following the western margin of the liuulder-clay to the northward, wi' lind it to occur at lower elevations and at a more or less considerable distance from the base of the Kocky Mountains. About the lifty-lirst parallel its lieight is from 8,400 to 3,800 feet, at a distance of thirty-live miles from the mountains, while still further lo the north it crosses the Ked Deer and North Saskatchewan liivers at an elevation of about 3,000 fei't, approxi- mately following a contour-line at that height. Scattered Laurentian boulders are, however, found at somewhat higher levels and to the west of the recognized margin of the l)()ulder-clay up to 3,200 and 3,400 feet.' Beyond the North Saskatchewan, the country toward the base of the mountains has not been so closely examined as to ensure that the highest western levels of the Laurentian drift have been accurately determined, On the watershed between the Athabasca and Peace liivers, however, (lat. 54' 12', long, lit ), Laurentian boulders are found at a height of 3,300 feet. In the Peace River valley, such eastern erratics have beim noted to h^'ights of 2,300 to 2,500 feet only, (in the vicinity of the; D'Echafaud River in latitude 55 45', longitude 120",) being there at a dis- tance of about seventy miles from the eastern base of the mountains proper. ' According to recent observations l)y Mr. McConnell, Laurentian boulders are again found in the Liard River region to occur some distance to the west of the margin of the boulder-clay, their greatest observed height being 2,300 in lat. 60" 15' long. 123". ' In the tabular comparison of the events of the Grlacial period in the regions of the Cordillera and the Great Plains, it is suggested that the remarkable monument of this period known as the Missouri Coteau,"' may with probability be referred to that time of ' Mr. Upliam accepts the oliservations liere referred to as evidences of the lliicknoss of ii confluent glacier which he supposes to have swept across the pl:iin.s to the haseofthe Kooky Mountains. 'Am. Geol. ^Magazine,' vol. iv, p. L'l."). For rcasrins olsewliere stated, I am unable to concur with him in tliis hyiiotliosis. '' 'Annual Report (ieol. Surv. Can.,' lS8(i, p. 14;! E, 'Bull. Ceol. ^oc. Am.' vol. i, p ;!'.t7. •' ' Heport of I'logress Geol. Surv. Can.,' 1879-80, pp. i;!9 13, 140 B, ' Quart. Journ. Geol, Sue.,' vol. xxxvii, p. 277. ' ' Bull. Geol. See Am.,' vol. i, p. 542. The boulder-clay in this region would appear to be confined to levels below a height of about 1,000 feet. '' This natural feature was noted and named by the early French voyageurs Grand Cukau du Missouri. OF TIIH KOl'KY MOUNTAI.V RllGfON IN CANADA. 61 ^ rest and stability which is marked in tlic CordiHoru by tlic whili> silt Ibrmation. This period si'eins to have been initiated by a partial siibisidcnrc ol' the Cordillera, aecoinpanied by a correlative elevation ol'the western part at least of the Ureal riaiiis, durin'^- which or as a consequence ol" it, a condition ot equilibrium was' established ;is between the Cordillera aiid the plains, which practically closed the series ot correlative oscillaaous which it is supposed were contemporaneous with the conditions of glaciation in the west. It is of course possible that the Coteau may have l)een produced aloni-- the western margin of the interior continental sea during the elevation by which the plains were alfected in the interglacial episode, but so far no evidence with this tendency has been obtained, while the slight degree to which the irregular ridges aiul mounds of the Coteau have since been affected by denudation, constitutes in itself an argument in favour of a later date. North of the forty-ninth parallel, the thick and irregular drift deposits of the Coteau, forming a marked feature on the plains, and running in a north-west by south-east bearing, have been examined in some detail for a length of about 200 miles, or to the elbow of the South Saskatchewan.' Throughout this part of its length the deposits forming Mie Coteau are piled on or against a more than usually abrupt slope of the surface of the plrins. To the north of the South Saskatchewi:n, the Coteau is probably move dilfuse, in conformity . with the less marked edge of the thirci steppe of the prairie, but it appears to rea.li the North Saskatchewan near its elbow. Id' material consists largely of boulder-clay, but this is generally covered with stratilied san Js a.ul gravels and strewn with many erratics.' Its height is nearly uniformly maintaiiicnl at from 2,000 to 2,200 feet above the present sea- level, and from its rougher and higher parts it subsides tow^ard the north-east into wide swelling undulations, which slope gradually down to merge into the general hwl of the plains of the second prairie st(>ppe. The whole appearance of this remarkable glacial deposit is to my mind that of a zone against which heavy debris-bearing ice stranded for considerable period during which no great ihange in level occurred. When, in 1875, I first recognized the glacial origin of the Coteau, it was pointed out that it might be explained either as the moraine of a continental glacier or in the manner just outlined, but subsequent investigation appears to me to add further probability to the hypothesis that it was ibrnred along a sea-margin. The line of the Coteau is practically concentric, for a considerable part of its length, with that of the outer and higher border of the glacial deposits of the Great Plains; and in this connection it is further worthy of remark that practically the entire area of the plains which is (diaracterized by glacial deposits lies on the arctic slope of the continent, thus rendering it probable that the water by which it appears to have been flooded, was in direct communication with the Arctic Ocean and may in fact be regarded as having been an expansion from this northern sea. It may be added, that in assuming the Coteau to represent a water-line, of which the relative eleva- tion of different parts has since suffered little change, we are led to believe that the line of maximum slope by which the surface of this part ot the Great Plains was affected (which slope was evidently increased subsequently to the formation of the Coteau) must lie at right-angles to its lenu'th, or from north-east to south-west, a belief which accords precisely with the actual great height of the south-western portion of the drift-covered area. ' "Geology and Resources of the Forty-iiinlh Piirallel," pp. 228, 250 ; ' Quart. Jouru. (Jool. Soc.,' vol. xxxi, p. C14; 'Annual Report Gool. Surv. Can.,' 1885, part C. ■' ' Annual Report Geol. Surv. Can.,' i885, p. 74 C. 62 (i. M. DAWSON OX Till", PIlYSiOdllAPIlIOAL GROLOflY I i It lias aliciidy licrii sliitfil thai there is no piddl' that, any part of tho groat Cordilloriin illariir at any time Jlowi'd eastward to the phiiiis liy the passes ol' tin; Korky MountaiiiH. 'I"iie I'lilire ahseiiee, wliellier ill the lioeky Mountains or on the plains, of material derivtid lidiii the (lold K'an^es, is alone almost snilieieiil to show that nothiiif^ ol' th(! kind hap- |)ene(l. Lo. a! yhu ieis ol' considerable siz(; wero, however, produced in thest; mountains, and sueli •.'•laeiers, dehouchins" to the eastward, formed moraines amoni;' the foot-liills and even Itcymid these on the margin of the plains. Moraiut's duo to thoso glaciers have been traeed, in a more or less degraded eondition, for al)out ton miles l)oyond the base of the mountains near the forty-ninth i)arallel. A glacier of larger dimension, than most of those derived from the mountains, evidently followed the liow Vallisy for a distance of at least twelve or iourtcen miles beyond the mountains.' These largo local glaciers are biievcdto hiive been contemporaneous with tho maximum development of the Cordil- Icraii glacier. The material ol' their moraim'S lias been derived entirely from tho llocky Mountains, but Laurontiaii erratics are occasionally found resting upon them, and they are frc(|ui'iitly surrounded by si ratilied deposits of later date. Some evidence has already l)eeii noted which tends to j)rove that tho llocky Mountain glaciers were insignificant duriiiu' th(! second nia.\iniuiii of glaciatioii of ilio Cordillera. (See i). 45 ) The correlative and comi)lem"ntary elevation and depression of the Cordilleran bolt and tho (Iri'at Plains which has been investigated in foregoing pages, necessarily implies either the existence of extensive contemporaneous faulting along the north-eastern margin of th(! Cordillera, or a hinge-like inovomont about tho same line, the dill'oron<'e in eleva- tion, in tho latter case, being accounted for by slopes of varying degrees of inclination. In tho absence of any t'videuce of such recent faulting on a large scale, tho latter remains tho more probable hypothesis. If the sequence of events which has been advanced be in the niaiii correct, the greatest change in height between neighboring tracts must have oi'curred during the second period of glaciation, in tho southern part of the region undtir discussion, and may have amounted to between 5,000 and G,OOU feet. There appears to be, hovvevt'r, oven here, ample room in which to distribute the resulting slope. Under tho assumption that tho dill'erenco occurred between tho lino of tho Columbia-Kootanie valley and the oxtrome western margin of tho eastern drift, a slope of one in sixty would ro(|uire to l)e impressed on tho intervening tract, while, even if it be assumed that the whole of this dill'eronce occurred between tho axis of tho Kocky Mountains and the same oastorn-drift margin, this slope would only be about doubled, and would thus amount to about 200 feet to the milc.- The last stage of thi^ (ilacial period appears to have been marked by a general move- ment in idevation of both the Great Plains and tho Cordillera, which it is presumed may have been connected and contemporaneous with a still more general movement in the same sense, by which tlu^ entire northern part of the continent was affected. Mr. Upham's researches in connection with the later-glacial or post-glacial " Lake Agassiz," which occupied tho valley of the lied liivor and tho basins of Lake Winnipeg and asso- ciated lakes, proves that in the eastern part, at least, of the plains, this (devatory move- ' ' Report of Progress Geol. Siirv. Can.,' 1882-84, p. 145 C. -■ Consiilerublo relative clmnges in clovatiou between tlio western and eastern sides of the Rocky Mountain Kango may be found to explain some pu/./.ling features of tlio actual drainage system of this range. See ' Annual Koi)ort Geol. Surv. ('an.,' 1885, p. 27 B. T 1 OF TIIK IKK'KY MOHXI'MN ftWilOX IN CANADA. 63 T mi'iit was yroatcsl to tlic novtli. Tlic north-eastwurd slope ol' the Wfstwanl i>arl ol' tin- plains may very prol)al)ly havis boeu I'urthttr iiicn'ascd at this timo. Ill the "(Jooloify Mild Kesoiin't's of llic Forty-iiiiilli I'avallfl " fai'ts arc luldiiord tciid- iiitf to slunv that a relative cl'valioii ol' lliu southern part ol' tin- western portion ol' tlie. plains is the latest change ol" this kind of whieh any record is I'oiind.' Tlie evideiire of this is derived ehielly from alterations in the courses ol' rivers to more northerly directions, and KU(^h evideneo has since heeii much added to and extended f-o as to include nearly tho entire area of the Districts of Alberta and Assinihoia." This elevation of the soutliern, or relative depression of the northern part of llie plains may v\'ith great prol)al)ility he assumed to have immediately succeeded the cliiiiige in an opposite sense last mentioned. To the wi'st of tlie Rocky Mountain I'mige, similar evidence of southward elevation or northward depression in immediately \)ost-glacial time, is ail'ord(Kl by tlu! Columbia-Kootanit> Valley,' and it is even possible that llie out- flow of thedreat Shuswap Lake 1)ecame clianged at the same time, from th' Spalliim- sheeii and Okanagan Valley to that of the South Thomi)soii.' The t)bseivalions here alluded to are at least sullieient to show that a depression to the north or elexatimi to tho south was very widespread in tln^ western part of the continent. In view of the fact thai an undoubted majority of the ucologists who have interested themselves in the phenomena of the period of glaciaiion in America are at |iresent dis- posed to attribute these phenomena almost in their entirely to the nctiuii of one or more great eonlluent glaciers, while my study of the glaciation of the northern part of the Great Plains leads me to hold the belief that this region has as a, whole been submerged, and thai floating ice has been the main agent in its glaciation, it appears to be necessary to add here, by way of justification for this belief, a synopsis of the jjrincipal fatits upon which it is based. This must, however, necessarily be brief and in scfmc measure imper- fect, as in the case of many circumstances observed on the ground, suliicieiitly precise d;ita have not yet been obtained to enable deliuiii! statements to be made respecting them in a summary form. Prof. J. S. Newberry has stated that " the track of a glacier is as unmistaka])le as thai of a man or a bear, and is as siguilicani and trustworthy as any other legil)l(! inscription." ' I would propose to extend this forcible statement of tin; cas(! by adding that the evidences of the action of glacier-ice and floatinn-ice arc as diU'erent as are the tracks of a man from those of a bear, though cases may nevertheless occur in which it is impo,ssible to decide to which agent certain traces should be assigned. Proof of the existence of a great con- fluent glacier on the Laurentiau axis, and of a second similar glacier on the Cordillera, appears to me lobe complete, while over the greater portion of the extent of the northern part of the plains the eviden(H' seems to have a diUereni meaning. One of the most striking circumstances in favour of a belief in the sub-aqueous origin of the glacial deposits of tht; plains, is their extraordinary pcrsistenct^ and similarity in character over the vast area which they cover in an almost uniform sheet. The area of ' 0}>. Cit. p. -'(i4. -' Report of Prof^rca.s Geol. Surv. Can.,' 188'J-84, pp. 14 (', 150 C; 'Annual Kepert Ciuol. Suiv. Can.,' 188."i, p. 75 C ; 'Annual UeportGool. Surv. Can.,' 1886, p. 140 E. ■"Annual ReiK)rt Geol. Surv- Can.,' 1885, p. :il B. * ' Report of Proaross Geol. Surv. Can.,' 1877-78, p. GO 15. ■' ' Geol. Surv., Ohio,' " Geology," vol. ii, p. 2. (54 (1. M KAWSON MN rilK I'llYSIOC KAI'IIK 'Ah (IKOLOdY ilhil pniiinii mIomi' nf lliis i'c'4'i(in whii li is iiwliidcd Itdwci'ii lh(> forty-niiitli and liriy-fourlh l);ii;illi'ls (if iioilli liililiidi', t'xci'i'ds 'J.'i(),0(io s(| niir.'-iiiili's, hciiii^ ;d)(nit Iniir tiiiif.s iiis irroat IIS llh' tuhd iin-a ol'lli'' Ni'W I'liin-hmd Sditivs, Over this cntirti trad, tho l><»ul(lor-rlay and oIliiT ii\ I rl\ iir^' d"|)(isils i'uriii ii iii'inly uiiWrokfii siii>rrliri;d liiy<'i' oi' idiiiONi, inliniti'Hitnal lliirkiii'ss ill tompaiisoM to its boulder-clay of the plains, besides its far-travtdled constituents, always incdudes a consideralile and oiteii a larye proportion of relatively or quiti^ local material. Owing to tin' imperfectly consolidated character of the Cretaceous and Ijuramie deposits of the plains, this important local element is I'ound most al)undantly as a portion of the paste, which varies in color to a considerable extent in correspondence with that of neighboring strata.' Notwithstanding the observal)le diversity thus caused, th(^ composition of the l)ouliler-A. 65 I AW J f tircuiiiHliiiicos. Tht'Ntnitilii'ddtiposils olNiiiKl, gifiv<'l aiulsilty miitrrial wliiili slill iiciuly I'VcrywIicrc ovcilin Ihf houldcr-fliiy, thoiijjfh in hoiiu- fuses no douhl in purl rcrfralilc td p(isl-jrliit;iiil iakcH ol" liuiitt'd diinrnsioiis, an- too widc-siircad and an- ionnd ill too many dill'crciit luvi'ls to allow this lo he acrcptcd as a '^cnciiil explanation of their oi lurn-nee. Irregularly stratified deposits ol' this kind even cover most ol' the hills and ridj^es ol' tho Missouri Coteiiu, respecting' ii part ol' \vhi: somewhat in the depressions,' In tlie southern part ol'the Dislrictol' Allx-rla, Ihesesuperlicial deposits are very i^'enerally I'ound to consist ol' yruvels and sands below, wilh sandy or clayey loam abovo. That theoverlyin<;- well-stratified deposits have not been formed l»y the > OF THE ROCKY MOUNTAIN RE(JION ]N CANADA. 67 complicated this question, and has to some extent weakened the force of the evidence referred to, by shewing that ranch of this Rocky Mountain material (which I had d(>signated in a general way as Qvarlzile Drift) has probably been secondarily derived i'rom the Miocene conglomerates. The earliest opportunity was embraced of alluding to this new factor in tin; problem, ' which, however, does not appear to me to be such as to invalidate the general argument derived Irom the mixture and overlapping of Laurentian and Rocky Mountain drift on the plains. The places above named are the only ones in which the Miocene conglomerates have been discovered, and their coarse character is such as to lead to the belit>f that they were formed in Miocene river-valleys or in the estuaries of rivers, and that the areas covered by them were never large. The existence of such conglomerate-covered areas appears in fact to have been suilicient to determine the existence of the plateaux mentioned and the limited size of these is probably connected with the actiial small dimensions of the original deposit. The Cypress and Iland-Hills Plateaux are situated at distances of 230 and 120 miles respectively from the base of the Rocky Mountains, while the material found in the glacial deposits and derived from the same mountains, is recognizable in these deposits, near the forty-ninth parallel, eastward to the 101st. meridian, 580 miles from its source and only 200 miles from the nearest part of the base of the Laurentian plateau.- It may be added, that quartzite IVom the Rocky Mountains may occasionally bo noted in the glacial deposits in the form of boulders of considerable size, siich as are not found in the Miocene conglomerates, while, moreover, a similar mixture of Laurentian a)id Rocky Mountain drift occurs miii-h further north in the Athabasca and Peace River basins, where no trace whatever of Miocene beds of any kind has yet been observed ; the Rocky Mountain material being there likewise found at times as boulders of some size. ' AVhile therefore I am not prepared to insist on the absolute character of this evid- ence, it would appear that no discoveries which have yet been made, are such as to negative its general ac(;uracy, the limited and lo(;al character of the Miocene conglo- merates bearing but a minute quantivalent proportion to the wide spread of the Quartzite iJrift. Because of the absence in the area of the Great Plains of rocks capable of aifording direct proof by their striation of the direction and character of the movement of the gla- ciating agent, I may be pardoned, before leaving this subject, for alluding to yet a further instance derived i'rom the transport of material. The rocks forming the central part of the Three Buttes or Sweet-Grass Hills (which have already been mentioned), are of vol- canic origin and dilfer from all others met with in the glaciated an-a of the Great Plains. A detailed study of the distribution of fragments of these rocks might therefore possess considerable importance in connection with the direction and manner of transport of material generally. We are as yet able to refer only to incomplete observations on this point, but even these appear to have some importance. Li the immediate vicinity of the Buttes, trappean rocks derived from their summits are I'reely mingled with erratics of eastern and western origin, but at some distance to the east, north and south such frag- ments are rarely found. A few specimens of these rocks were, however, discovered at a ' ' Report of Progress Geol. Siirv. Can.' 1S8l'-84, pp. 142 C, 143 C. ' " Geology and Resources of tiio Forty-ninth Parallel," p. 225. ' Quart, .roiirn. jleol. Sou.,' vol. xxxi, p. 013. ' 'Report of Progress Oeol. Surv. fan.,' 187!)-80, p. 140 B. '(Juart. .fourn. Ge( i. Soc.,' vol. xxxvii, p. 277. 68 G. M. BAWSOX OX TIIK I'lIVSIOOHAPlirCAL GEOLOGY point sixty miles due east. Fnigmeuts are also moderately abundant in the drift seven- te<'n miles north oi' the East Butte and .seven miles or more to the north of the West Buttt'. ' The country to the south and south-west of the Euttes, beinj^ in the northern part of Montana, has not been examined by me, Imt it would appear to be in this direc- tion, that the main stream of drbris from the Bultes must be looked for. Thi' uuoxidised and uuweathered character of the material of the boulder-clays of the plains, may be quoted as a further evidence in favor of their sub-aqueous origin ; and whereA-er they have not been manifestly subjected to recent subaerial action, this character is always apparent. They frequently contain great numbers of more or less completely rounded shaly fragments or pebbles, derived in n-e particularly from the clay-shales of the Pierre group of the Cretaceous, which so soon as they are subjected to the weather, split up and crumble to pieces. Boulder-shaped masses con.sisting of sand have also been found in railway cuttings near Medicine Hat, on the South Saskatchewan, which it is dif- ficult to account for except on the supposition that such sand, in a wet state, had been frozen into solid masses, which sank to the bottom of water and were immediately enclosed by earthy or clayey material there. The direct evidence of submergence of the plains afforded by terraces and analogous phenomena may next be brielly alluded to. Terraces found along the river-valleys of the plains, with those met with in the valleys of th(^ IJocky Mountains, may here be omitted from consideration, as being for the most part due to local circumstances, such as post- glacial river-erosion and the stoppage of mountain-valleys l>y local-glaciers. Terraces evidently frei^ from suspicion in these n^spects are, howt>ver, found in considerable number and often well develo])ed, in the higher part of the plains h^ss than one hundred miles from the base of the mountains, and among the foot-hills. On this point I may quote as follows from my " l^eport on the Region in the vicinity of the Bow and Belly Rivers." "On approaching the mountains, however, true terraces of a more significant character presi'ut themselves in many places. Terraces in the entrance to the South Kootanie Pass at a height of about 4,400 feet have already been described in my Boundary Commission Report. In the valleys of Mill and riu(^her Creeks, p.nd those of th(! Forks of the Old ]\lan, east of the actual base of the mountains, wide terraces and terrace-Jlats are found, stretching out from the ridges ol the foot-hills, and running up the valleys of the various streams. Actual gravelly beaches occasionally mark the junction cf the ter- races with the bounding slopes, and they have no connection with the present streams, which cut through them. Their level varies in dilFerent localities, but the highest observed as well characterized, attains an elevation of about 4,200 I'eet, In the Bow valley near Morley (4,000) and then.x' to the foot of the mountains, similar terraces are found wliich are quite independent of the modern river ; and in the wide mouth of the Kananaskis Pass a series of terraces was seen from a distance, which must rise to an elevation of at least 4,500 feet." - Many years ago Sir James Hector, wrote us Ibllows of the higher western border region of the G^reat Plains:— "On approaching the Roc'y Mountains, the extreme regu- "Geolotiy and ReKOiiiviw oftlie Foily-niiitli Parallel," pp. 2-10, LMl. Also MS. notes of 188:!. ' Report of Progress Geol. Surv. fan.,' 1882-84, p. 146 V ■ see also '• Ueology and Re.sources of the Kortv-nintl. Parallel," p. 244 ; ' Quart. Jonrn. Geol. Soc.,' vol. xxxi, p. t>lS. „ — liinBrri iifci^a^ OF TIIH ROCKY MOUNTAIN RKGION IN CANADA. 69 larity with which these deposits have been terraced by the retiring waters, at once attracts attention. * * * The terraces may b;' considered us ranging on the east side of the Rocky Mountains from 3,500 to 4,500 feet above the sea." ' To the east of this now more elevated part of the plains, evideuci's of water action of a similar character are to be sought rather in the existence of extensive tracts of almost perfectly level prairie, which occur at various elevations, and wliich nuiy be regarded as more or less perfect examples of " plains of marine denudation " impressed on the soft and originally not very unequal surface of dcpositioi;. From this category, level plains formed by the iulilling of post-glacial lakes, are of course excluded. Some evidences of water action are, ho-" -ever, also to be found here along the flanks of projecting plateaux. One distinct terrace of this kind is noted by Mr. McCounell on the Cypr.'ss-Hills Plateau, at a height of 3,700 feet,- while the south-eastern front of the Uocky-Spring Tlateau, a few miles south of the international boundary, is strewn with Laur.'utian and Iluronian erratics in a manner highly suggestive of an ancient shore line,' and the upper limit of the drift on the northern slope of the plateau, at the height noted on a former page, shows shingly materials resembling old beach deposits. In order to explain the covering of the Great Plains with boulder-clay and other glacial deposits in conformity with the theory of the deposit of these from a great ice- sheet, we must adnitthat a great confluent glacier spread continuously across these plains from a gathering-ground situated on or beyond the corresponding portion of the Lauren- tian plateau. It has been sugg(\sti'd by some authors that such a glacier extended across the plains only as far as the line of the Missouri Coteau, and that a great lake, held in on one side by the front of the glacier, stretched thence to the wi'stern margin of the drift, a further distance of 300 miles. Such an explanation, however, appears to me to be inad- missible on account of the practical identity of the phenomena and deposits met with to the east and west of the Coteau, and more particularly because of the homogeneous character of the boulder-clay deposit of the two areas.' Others ' have advanced the con- sistent, though bolder view, that the supposed great confluent ner reached quite to the western limit of the drift, or at least as far as the margin of ^.hc ;ognized boulder-clay, which does not fall far short of this limit. Holding this view, Mr. Tyrrell, in the light of facts still requiring the action of water in that part of the plains near the base of thi; Rocky Mountains, supposes that here at least lakes may have occurred in front of a great ice- sheet." On either of these forms of the hypothesis, we must admit that the plains werii extensively covered by water at the inception of the Glacial period, (when the iSaskat- ' " ICxploration of Britisli North America," p. I'l'i'. - ' Annual Report Geol. .'^urv. Can.,' 1885, p. 74 C. ■' ' Uoiwrt of l'rojrres.s Geol. Suiv. Can.,' 18,S:.'-84, p. 148 C. ' There is mo such (ieneral dillerence as between the drift dupo.'iita to this east and west of tlio Missouri Coteau, in this region, as to admit of the separation uf the latter from the former and their inclusion as a whole under tlie designation of att«naated border deposits (Chamberlin) or fringing deposits (howis and Wrigiit.) My earlier pxaminations of the plains near the forty-ninth parallol,l)oing,to the west of tlie Missouri Coteau, principally confined the higher par's of the region, apiieared to allord some indication that a general dillerence might be established, but subsequent more extended observations have not l)orne this out. ■' I'articularly Messrs. Upham, Wright and Tyrrell in publications already referred to. " ' Bull., Geol. .Soc. Am,' vol. i, p. 401. 70 (I. M. DAWSON ON TIIR PlIVSIOUIfAPIIICAL GEOLOGY clu'wan gravels, wen' laid down) ; again in the interglacial episode the deposits of which separate the two boulder-clays ; yet again during the principal period of gjaciation, (when the supposed glacicr-damnicd lakes were i'orined) ; and finally, we are called on to suppose that large parts of the surface were covered by post-glacial lakes, of which the latest and lowest in level was that which Mr. IJphani has named Lake Agassiz. It is thus to account lor the boulder-days alone that we are on any hypothesis asked to admit the passage; of a great glacier over ihe plains, and in view of the circumstance that no positively definite distinction in rega'd to physical characteristics has as yet been established as between boulder-clays of supposed sub-glacier origin i nd similar deposits produv;ed by ice-laden water,' 1 cannot but think that the assumption of such a glacier invasion as that supposed is scarcely warranted or required by the obser ved facts. Admitting for the moment the existence of a glacier co-extensive with the boulder- <^lay deposits of the Great Plains, we may note the required dimensions and movement of such a glacier. The main direction implied by the abundant Laurentian and Iluronian drift of all parts of the glaciated area may be stated as from north-east to south-west. While it is possible that the principal direction of transport w^as more dire<;tly from the east or north, it must have lain between these bearings, and its precise direction is not important in the present connection. "Whether the supposed great ghKuer originated on the Laurentian jilateau or crossed it from some further point, it must at least be assumed to have moved continuously from the south-western edge of the Laurentian area to the furthest point at which material attributed to its action is found ; or in other words, each identical transverse section of the glacier, charged or covered with Laurentian debris, must have been bodily pushed forward at least to the western edge of the boulder- clay deposits. This involves the conception that the glacier was thrust forward for a distance of over 500 miles from the margin of the Laurentian as a minimum, for no glacier-mass produced upon the surface of the plains and merely reinforced by accessions of ice from the Laurentian plateau will account for the carriage of the material from the latter. In earlier dist'ussions of this subject, I have assumed that a great glacier-mass thus moving across the plains would have to ascend their present gradual upward slope to the south-westward,- but if I am correct in the belief now entertained that the a'^tual inclination of the plains was produced at a tin^e subsequent to the main period of glacia- tion, we have to suppose only that the glacier was thus pushed forward over a nearly level surface, and we may admit that it was possibly assisted in its motion by a greater contemporaneous elevation of the Laurentian plateau. Even under such conditions, however, the glacier must have met with serious impediments, not the least of which is the escarpment of the Porcupine, Duck, Kiding and Pembina "Mountains,"' composed of the (ut-oif edges of the Cretaceous strata, which faces the Laurentian plateau at an average distance of 130 miles from its margin and rises to a height in places of more than 1,000 feet above the intervening low tract. It is true that Mr. Tyrrell, while accepting the existence of an almost universal glacier such as that above indicated, quotes observations in favour of the belief that the ilow of the ice descending from that part of I Cf. Wright, " Ice Age in North America," p. 116. - "Geology iind Resources of tlic Forty-ninth Parallel," p. 261 ; 'Quart. Journ. Geol. Soc.,' vol. xxxi, p. 620. •' 'Geological Magazine,' Dec. ii, vol. v. y>. 211. t ■ ! ■ 1 -:• OF THE EOGKY MOUNTAIN EEGION IN CANADA. 71 J ,1 the Laurontian plateau which is more or less routiimously screened from the plains by this Lsiarpmout, on reaching the intervening hollow turued to a southward or south- t'astward direction.' These observations of direction of stria;, howevi>r, render it only the more difficult to account for the flow of the supposed great mer de glace across the plains, for if the ice from this nearest and most obvious source was thus largely or altogether turned aside, for a width of about 300 miles measured along the base of the Laureutiau feeding-ground, the deficiency thus produced must obviously have been made up by a greater accession of glacier-ice to the plains from the north, which ice, cro.ssuig the fifty- third parallel with a width of about oOO miles, succeeded in attaining the forty-ninth parallel with a width of GOO miles and still with a very great thickness. The local contributions to such a glacier by precipitation over the area of the southern part of the Canadian plains would be insignificant, while the loss by melting must be supposed to have l)een great, as the region in question is in the central part of thi- continent, and must then as now have been one of small rain or snow-fall.- The hypothesis of such a vast mid-continental glacier and the effects attributed to it appear to l)e particu- larly opposed by this circumstance. In addition to these more general aspects of the conditions implied by the theory of a mer-de-glace extending over the area of the Great Plains, reference may now be made to still another circumstance which to me appears in itself to bo destruiJtive of the hypothesis, this being dependent on the efFe<;t which such an ice-nuiss mi;st have produced on materials over which it passed. Doubt has been felt and expressed by competent authorities as to the possibility of the formation to any considerable extent of till or boulder-clay beneath the mass of an ice-sheet in motion.' It is obvious on a priori grounds as well as on reference to known instances of glaciation on a large scale, that the principal elfect of a moving ice-sheet or glacier is in the direction of the removal of all incoherent deposits and the baring of the underlying and more resistant rock-surface, the face of which is striated and smoothed, or even abraded and channelled, wht'ii the ice-covering has been heavy or its action long continued. It is thus only under somewhat exceptional circumstances and loi-ally that till can be proved to have originated beneath a glacier-mass, or that antecedent deposits o!' a non-indurated character can be shown to have maintained their position below^ such a mass. Such instances have, how- ever, l)een made the basis for an almost general theory of the origin of till or l)oulder-clay, as a bottom-moraine, and in some instances this has been carried no far as to (iractically rele- gate the whole of such deposits to this origin. The extension of su(;h a theory to the vast area of the plains appears to subject it to a strain greater than it is capable of bearing, and such an extension can in fiict be mentally approached only by a hypothetical series of increasing approximations, beginning with the minor observed instances which have been referred to and leaving other circumstances out of consideration. Prof James Geikie, in oifering an explanation of the preservation of certain deposits under till on the hypothesis that the latter was itself formed ])eneath a great glacier, writes:— "In the open lowlands and in the broad valleys, where the ice-sheet would ' ' Bull., Geol. Boc. Am.,' vol. i, i). 401. ' Cf. W. ,1. McGee, "On Maximmn Synchronous Glaciation." 'Proc. Ain. Assoc. Adv. Sci.,' Boston, 1880, p. 447. Also Von Wcuickoll', ' rrocecdings Geol. Soc, Berlin,' 1881. ^ Cf. ' American Journal Science,' iii, vol. xxxiv.p. 52, where Mr. 0. P. Hay attempts to explain such formation. 72 (J. U. DAWSON ON TJIK PHYSIOCKArillCAl, (JI'lOLOGY lulvaiici! willi diininishod l)iit mort' oqiiahlc uiotiou, we come upon wulospioad and often derp er. ii, vol. v, )>. 75. OF THE IIOCKY MOUNTAIN REGION IN CANADA. 78 glacial nor the interglaoial deposits referred to have beeu observed to be disturbed or contorted.' From what has already b'eii said with respect to the Cordillerau region, and more particularly in couuectiou with the meaning which the White Silt Ibrniatiou appears to have iu that region, it seems probable that the water by which the northern part of the Great Plains is supposed to have been Hooded, was in connection with that of the sea.- In discussing the results of my earlier investigations of the superficial deposits of this part ot the plains, in reference to a theory of their submergence, I have stated, that alter a certain stage the waters entering from the north and south must have formed an open strait between the Arctic Ocean and the ocean to the south.' This was written, however, under an assumed limitation implying an ecjual subsidence of the continent, and at the time no satisfactory information was availal)le respecting the position of the margin of the glacial deposits in the corresponding western part of the United States, such as has since beeu supplied by the work of Chamberlin, Salisbury, Todd, Wright, McGee, Upham and others. The result of these new facts appears to show that instead of opening broadly southward as well as to the north, any body of water covering the northern part of the Great Plains could have had only a tortuous and comparatively narrow communication with the sea to the eastward, round the front of the great confluent Laur -ntide glacier, and that even this communication was probably formed only at the time during which the plains stood at the lowest level indicated by the greatest spread of the drift deposits. If sut-h conditions may be assumed as probably representing the facts at the time, they go far toward explaining one of the greatest ditficulties against the acceptance of the hypothesis that the waters by which the plains were flooded were in communication with those of the sea. The difficulty alluded to is the complete absence, so far as yet ascertained, of the remains of marine organisms from the glacial deposits. While prolonged weathering and the action of sub-aerial waters might result in the removal of calcareous organic remains from certain parts of these deposits, the condition of much of the boulder- clay, together with the occasional actual occurrence in it of fragments of Cretaceous or Laramie shells, is such as to show that any contemporaneous molluscs might have be(>u preserved. If, however, the body of vi'ater in question, though communicating with the sea to the northward, was almost throughout closinl to the south and in receipt of large quantities of lluvial water, it may well have been in great part brackish, if not almost entirely fresh. Adding to this the conception of its frigid temperature due to the great abundance of ice with which it must have been laden, and the vast amount of Hne sediment which must hu 'e been carried into it by sub-glacier streams, it will be apparent that the conditions were singularly inimical to the existence of life of any kind, whether that characteristic of salt or fresh water. Somewhat similar conditions, though on a much ' Piof. J. W. Spencer lias lately adduced much additional evidence to show that modern glaciers, near their terminations, produce little efl'ect even on loose material lieneath them, ('Trans. Roy. Soc. Can.,' vol. v, sec. iv, p. 89.) but the mode of action of a glacier at its decaying extremity and when in contact with materials in a position to acquire more or less heat from the atmosphere, and by radiation and conduction from neighl)oring warmer bodies, must be very dilferent from that which would be found far back beneath tlie mass of even a small glacier. ' It may still, however, be a« d <^ 4> ^ 3 3 o H ■1 Tiitiis. U.S. C, 1800. Sor. IV. "late IIT. m '^^ «<«r fe:--. i^^^il ^J fiasiii .♦.'*; .^ - ,K ,*%'»1 /lll'i" iw# fis :^4: y*«<5 ijf' M mm %^ 4/i «■'■*«> *\^*Vi,^».. ■■•■•J ■■-■>■■ # -...s^ *^ ■--'^'^ *u#,. ■^i,. fd -^ 11^ 1^ V^: i:\^^^i| r-:^^^1V, .^pi^^ I c <5 Cu C o (/) o J 3 o 2i. tf o -7 :" w UJ ° > OC •^ r., i; 1 o o rt (A) Ul I u o T) d o §1 13 O U j: rt Q J3 ii •f- o u -^,f. . li U- O rt .c t- 2 1 >- UJ j: o _) 1 _j J= c . a 1 > cc .1^ ■A UJ z Ui (J c o eft E o ^•3 1 O -C .« Q- _1 1 > -J tr 1 t o UJ o .^ c)^ "- UJ >, 4-1 t*. I Tl C — < 3 J O .-J z 1- 1 i ul tft 8 ill I rt ?^^ 1' 0) •rt , 1 3 Q- *-• »« > Ul c 0-" ;^ D. ^ ilJ u.. O c 1- ;3 o 4-1 To illustrate Dr. G. M. Dawson's Paper on llie Rocky Mountain Region.