BULLETIN OF THE GEOLOGICAL SOCIETY OF AMERICA

V o l . 34, p p . 499-soe September 30, 1923

GLACIAL PROBLEMS 1

BY GEORGE H . CHADWICK

(Presented before the Society December SO, 1922)

CONTENTS Page and Lake Vanuxem...... 499 Lake Lundy (?) outlet...... 501 Restoration of Lake Amsterdam...... 501 Eastw ard reach of Lake A rkona...... 502 Westward reach of Lake Newberry...... 502 A new Genesee la k e...... 503 Hudson Valley ...... 503 Lake and Lake Emmons...... 503 The Algonquin lakes...... 504 Correlation ch art...... 505 Cattaraugus lakes...... 505 Some preglacial rivers...... 505

L a k e W a y n e a n d L a k e V a n u x e m

Year by year the story of the Laurentian glacial lakes grows more complicated. The steadily falling waters of our youthful innocence are being replaced by a pulsating rise and fall due to rhythmic oscillations of the ice-front. One of these pulsations was long ago recognized by Fairchild2 in his “free drainage” stage of lowered escape followed by the restoration of “Lake Vanuxem.” Fairchild has frequently predicted that the real history would prove to be more complex than existing knowledge revealed. Leverett and Taylor’s recent exposition3 of the caprices of the ice on the thumb of instantly involves western and central , for the moment the water levels fell below the outlet they must go out by the Mohawk. The control channels of “Lake Wayne” must be sought at either Batavia or Syracuse. Herein is introduced a new element in the recognized New York lake

1 Manuscript received by the Secretary of the Society May 1, 1923. 2 Bull. 118, N. Y. State Museum, p. 80; Bull. 127, N. Y. State Museum, pp. 50-59. 3 Mon. LIII, U. S. Geol. Survey, pp. 364-370. (499)

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/34/3/499/3414562/BUL34_3-0499.pdf by guest on 27 September 2021 5 0 0 G. H . CHADW ICK----GLACIAL LAKE PROBLEMS

succession. Hitherto it had been believed that the first of the Erie Basin waters to gain admission into central New York (the Genesee and Finger Lake valleys) was . In Fairchild’s conception “Lake Vanuxem” stood as the earliest water body with easterly (Mohawk) es­ cape, but he ascribed to Vanuxem no drainage from west of Batavia. Yet, if we correctly understand the relations, the episode of Lake Vanuxem and the “free drainage” interim is the'only known place in the New York succession into which “Lake Wayne” can be fitted. The admission of the Warren waters into central New York, past the Batavia salient, without conspicuous channeling on that salient presented a real problem to Fairchild.4 Practical coincidence of the merging water levels had to be subsumed. But if the Erie waters already flowed east­ ward, the restored Warren level would reach to Syracuse from its initia­ tion and this particular puzzle be eliminated. Bather, the problem is pushed back to Lake Wayne, leaving the apparent absence of channels north of Batavia still a matter of inquiry. Possibly the control point is actually somewhat farther east, near Le Eoy, where there are heavy channels just below the Warren beaches. None of the Le Boy channels appears, however, to fulfill the require­ ments for the long-lived Wayne spillway. They conform to the tem­ porary paths of waters rushing to a new confluence, not to a stabilized outlet. For that we must turn to Syracuse. Here we find at once a splendid channel at the right altitude (about 40 feet below Warren) hitherto unreferred to any fixed water plane. This is the “Gulf” west of Marcellus.5 It is confidently believed that the “Gulf” channel carried the Wayne waters prior to Lake Warren, and that it did not function again subsequent to the Warren flooding. A curious fact concerning the “Gulf” remains, nevertheless. Below its intake it is distinctly depicted by Fairchild as a “two-story” channel (the only one he so represents), with the lower story much narrower than the upper, thus indicating diminished water flow. If this narrower inner channel were cut during the ice readvance following the “free drainage” stage, as seems likely, then the restored level (“second Vanuxem”) would appear to have been robbed of the Erie drainage by readvance also at Batavia (or Le Eoy?). This would mean that while “first” Vanuxem included the Erie flow from Wayne downward to its minimum stand and extinction in “free drainage,” or even through the earlier rising levels of its restoration, there was no “restored Lake Wayne,” but only local waters of the restored (incorrectly “second”) Lake Vanuxem coursing

4 Bull. 127, N. Y. State Museum, pp. 51-52. 5 Bull. 127, N. Y. State Museum, pp. 26-30, pis. 4, 18, 19A.

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/34/3/499/3414562/BUL34_3-0499.pdf by guest on 27 September 2021 LAKE WAYNE AND LAKE VANUXEM 5 0 1

through the “Gulf,” while the Erie Basin waters were obstructed at Batavia and thrown back on the Grand River outlet as Lake Warren. This ice advance would obliterate the earlier channels at Batavia, but there is still the old difficulty of getting the Warren level eastward again past this point without a trace.

L aktc L undy (? ) Outlet

With the Wayne outlet at Marcellus (Syracuse), the fate of the Erie ' Basin waters during the “free drainage” interval remains to be consid­ ered. Eollowing westward (upstream) the capacious channels of this stage, and noting their extensive delta deposits where they cross -the Genesee and similar valleys, one is impressed with the belief that they carried more than local flow. At their upper end, north of Le Roy, the channel is a splendid rock-cut, over a mile long, a hundred feet deep, and & quarter mile wide. Its col appears to have been originally some 20 feet or more under the plane of the neighboring Dana beaches, but to have been somewhat silted up during the Dana stage.6 Back of this broad intake at Port Hill, with present altitude about 680 feet, must have lain a huge lake, stretching to Detroit, hitherto unrecognized as such. Its close correspondence in level suggests that its beaches in the Erie Basin have been confused with those of the long subsequent Lake Dana. Spencer in 18947 described the Lundy beach and once incidentally used the expression “the Lundy lake,” a name which thus seems to have no standing as against the properly proposed and worthy name of Lake Dana, subsequently given by Fairchild to his Geneva beach.8 But should it appear on further study that the Lundy beach correlates with the Fort Hill channel, as is very possible, then both names will stand. Indeed, the Belcoda and other channels on the same meridian may eventually explain other features in the complex of beaches at this general horizon in the Erie Basin.

R estoration of L ake A msterdam

During the “free drainage” stage the Mohawk Valley was necessarily unblocked at the east, giving passage to Fairchild’s “glacio-.” But Fairchild has shown9 that the east-leading channels of sub­ sequent date from lowering Warren and Dana terminate east of Syracuse

e See map, Bull. 127, N. Y. State Museum, pi. 2. 7 Amer. Jour. Sci., vol. 47, pp. 207-212. Said to be. 25 to 40 feet out of harmony with JLake Dana, in U. S. Geol. Survey Mon. 42, p. 772. 8 Amer. Jour. Sci., vol. 7, 1899, pp. 260-1; Bull. Geol. Soc. Amer., vol. 10, pp. 56-57. 9 Amer. Jour. Sci., vol. 7, 1899, p. 262; compare also 20th Ann. Rept. N. Y. State ■Geologist, pp. 112 et seq;, pi. .16; Bull. .160. N. Y. State Museum, p. 32.

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/34/3/499/3414562/BUL34_3-0499.pdf by guest on 27 September 2021 502 G. H. CHADWICK----GLACIAL LAKE PROBLEMS

(at Mycenae) considerably above the level of the Rome outlet, or higher than the free drainage channels. The inevitable conclusion is that they discharged into a restored Lake Amsterdam, due to reblocking of the Mohawk at its lower end (Schenectady). It is reasonable that any re- advance of the ice at Syracuse sufficient to restore the Grand River outlet must have been accompanied (or slightly preceded) by a powerful thrust in the Hudson Yalley, of which we here have the confirmation. Appar­ ently this forward shove was felt also (a bit later?) at Batavia, thus sundering Warren and Yanuxem as above intimated.

E astward R e a c h of L a k e A rkona

More puzzling questions revolve around earlier reexpansions of the ice. Taylor thinks that invaded central New York before the readvance extinguished it. The position of the overriding Alden in the wider portion of the Genesee Yalley (where the Warren beach is strong but single), and thence eastward to the Seneca Valley, is well above the Warren level and shows that the ice must have there destroyed any Arkona beaches, except far south up the valleys, where they would necessarily be weak. Careful search may. yet reveal these; probable beaches and notches occur, 20 feet above the Warren shore, at and east of Geneseo, but the best record is found in the large 850-foot delta terrace at the mouth of the Mount Morris canyon, which could not have been built in Warren waters, because the Canyon had been cut far back and deepened in the stage of lowered escape preceding Warren. The fact that the moraine of readvance barely sunders the Mount Morris and New­ berry levels at the critical points near East Bloomfield, Reeds Corners, and Gorham is evidence that even Newberry had been swallowed down into Arkona before that readvance occurred. (See the chart, page 506.)

W estward R ea ch op L ake N ewberry

In his later writings and maps10 Fairchild carries Lake Newberry into the Genesee Valley from the east at this same moment in the history preceding Lake Hall. While the overriding ice has destroyed the records, making it impossible to say that Newberry did not have such an exten­ sion prior to Arkona, yet it seems to the writer unlikely that the New­ berry level entered the Canandaigua Valley before its restoration, and ■ not immediately even then, being momentarily excluded by a restoration of the Potter Lake with new outlet east of Gorham. This presently fell to Newberry, and when the latter almost immediately coalesced with the

“ Bull. 106, N. Y. State Museum, p. 32; Bull. 127, N. Y. State Museum, pi. 35; etc.

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/34/3/499/3414562/BUL34_3-0499.pdf by guest on 27 September 2021 WESTWARD REACH OF LAKE NEWBERRY 5 0 3

Genesee waters these were at the Pearl Creek outlet, so exactly coincident in elevation that a divided escape ensued until the unblocking of the Linwood channel established Lake Hall.11 It should be noted that Hall thus becomes a fairly fixed water body, not a series of lowering stages, though it may have been let down once, some 20 feet, to the Stafford channels. A n e w G e n e s e e L a k e

Studies now proceeding on the complicated history of the Genesee Valley lakes indicate that Lake Dansville was admitted into the Canan­ daigua Valley by way of the Hunt Hollow strait following Lake Naples, and was then lowered down on this pass, instituting a new member in the Genesee succession, here named the “Livonia Lake.” By opening of the Bethany channels Livonia became the Mount Morris falling waters, whose intricate history involved one more brief escape to the Canan­ daigua Valley by the Cheshire channels, as well as temporary separation from the Wyoming Valley by lowering on the Pearl Creek outlet channel.

.H u d s o n V a l l e y L a k e s

Greater interest attaches to the behavior of the Hudson-Champlain Valley during ice-waning., Woodworth’s interpretation12 of this behavior involved a peripheral bulge and a wave of uplift. Fairchild, whose early conception of the uplift as a rigid tilting was opposed to Woodworth’s, has found independent evidence of the peripheral bulge in his work on the Susquehanna River.13 Antevs, Daly, and others assert the existence of the bulge and of lakes restrained behind it. Of these lakes one must have occupied New York Harbor and its environs, perhaps continuous with that in Long Island Sound; for this the name “Lake Manhattan” is here suggested. As the ice waned and the pursuing bulge raised Man­ hattan, the Tappan Zee would have held a water body which we may call “Lake Haverstraw.” Perhaps even a third stage, north of the High­ lands, preceded the inception of , and this might suitably be called “Lake Newburgh.” The writer finds it difficult to conceive that any of these waters were salt, or with marine organisms.

L a k e V e r m o n t a n d L a k e E m m o n s

It becomes necessary at this point to reinstate Woodworth’s Lake Ver­ mont, for, like the precedipg, this was sundered from the tides by the

31 Bull. 106, N. Y. State Museum, p. 33, pi. 6. 12 Bull. 84. N. Y. State Museum, pp. 229, 232. 13 Science, n. s., vol. 57, no. 1465, p. 113.

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/34/3/499/3414562/BUL34_3-0499.pdf by guest on 27 September 2021 5 0 4 G. H . CHADW ICK----GLACIAL LAKE PROBLEMS

migrating bulge. That it had no free communication with salt water via the Hudson Valley is an inevitable corollary of Fairchild’s discussion of the Round Lake region.14 Analysis of that paper shows that long before the unblocking of the north end of the Adirondacks, while still Lake Iroquois was discharging through the Mohawk and cutting the Round Lake channel, the Hudson Valley at that point had lifted so far as to confine the to the narrow trench it was reexcavating in the clays of the Lake Albany filling. Lake Albany was gone. There was no open strait of salt water (or any other water) from New York to Plattsburg, but a river of fresh water flowing (with a gradient) south­ ward and holding to superoceanic level. The true marine waters (Hochelagan Sea) entered the Champlain Valley only when ad­ mitted from the northeast, past . By the original definition “Gilbert Gulf” is the portion of these marine (tidelevel) waters confined to the Ontario Basin, and the name should not be extended over the region in which Woodworth’s “Hochelagan Sea” has clear priority.15 The Vermont beaches are finely developed at 740 feet above tide and down, both around Covey H ill and far into the Ontario-Saint Lawrence Valley. The true marine beaches are those originally recognized as such from 523 feet above tide down, on Covey Hill. Between these two sets is a gap of nearly 100 feet, marking the elevation of Vermont waters above sealevel just prior to their extinction. ■ Above the Vermont beaches, between them and Iroquois, in the Ontario-Saint Lawrence Valley, lie the “Emmons” beaches of Fairchild, rediscovered and renamed “Frontenac” by Taylor.16

T h e Algonquin L akes

However intricately interwoven may be the Algonquin beaches, in point of historical sequence there are several distinct water bodies at present passing under the name “.” Leverett and Taylor discriminate (1) an “early Lake Algonquin” confined to the south end of the Huron basin, with Detroit (that is, Port Huron) outlet; (2) doubtfully a tripartite lake with outflow divided past Detroit and Chi­ cago, which seems to be an expanded ; (3) the true Algon­ quin, tripartite, with Trent River escape, and (4) the second Algonquin, restored to Port Huron outlet. To avoid confusion of speech, it is sug­ gested that the first stage be called rather the primitive Huron, and that

14 Bull. 195. N. Y. State Museum, pp. 12-15. and map p i.; compare also Bull. 215-6, N. Y. State Museum, pp. 28, 46, figs. 10. 11; Bull. 154, N. Y. State Museum, pp. 30-3, fig. 5; Bull. Geol. Soc. Amer.. vol. 33, p. 525. “ Bull. 84, N. Y. State Museum, p. 220. 16 Bull. 158, N. Y. State Museum, p. 34 ; Bull. 164, N. Y. State Museum, p. 22 ; Mon. LIII, U. S. Geol. Survey, pp. 325, 445 ; Proc. Roch. Acad. Sci., vol. 5, p. 138.

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/34/3/499/3414562/BUL34_3-0499.pdf by guest on 27 September 2021 THE ALGONQUIN LAKES 5 0 5

the almost resundered Huron waters during the later Nipissing-Ottawa outlet stage be named “Huronipissing.” The temporary lifting of Port Huron to level with or above the Chicago outlet and subsequent abate­ ment witnesses to the migrating bulge hereinbefore advocated.

C o r r e l a t io n C h a r t

The relations of these water bodies as above understood are exhibited in the accompanying correlation chart, in which contemporaneity is hori­ zontally represented and time reads from top downward. Proportional time values are not closely attempted with present inadequate knowledge, but attention is called to the rhythmically spaced ice maxima, which may have a coefficient of approximately 26,000 years. The purpose of the chart is to focus attention on outstanding problems and provoke fresh attack. A similar chart projected for the minor lakes of western and central New York is deferred to a future writing, but the following pertinent note is retained here:

C a t t a r a u g u s L a k e s

Exploration in the Cattaraugus country during the past summer re­ vealed a fine channel leading from the South Branch Cattaraugus Creek westward toward New Albion, with summit elevation about 1,430 feet above tide. As this is intermediate in altitude between the Lime Lake (Machias) and Persia outlets already recognized by Fairchild,17 the suc­ cession of lakes in the Cattaraugus Valley gains a new member. Another probable outlet examined is the pass north of Ellicottville at about 1,640 feet aneroid (sheet not issued). It would seem that the Dayton pass used by the Buffalo and Jamestown division of the Erie Railway must also have carried some overflow for a time, but there is no good channel across the col.

S o m e p r e g l a c ia l R iv e r s

In the nature of a supplementary proposal, the following names are tentatively put forward for certain fairly established rivers of preglacial drainage: “Allegowanda” for the ancient upper Allegany, with discharge by Gowanda to the lower Cattaraugus Valley; “Geneseraga” for the ancient Genesee past Sonyea into the Canaseraga trunk valley and, thence lower, through the Irondequoit trough, and “Senecahanna” for the ancient upper Susquehanna flowing out by the Seneca Lake Valley to the Ontarian River.

« Bull. 106, N. Y. State Museum, pp. 16-17, 35.

Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/34/3/499/3414562/BUL34_3-0499.pdf by guest on 27 September 2021 on 27 September 2021 by guest Downloaded from http://pubs.geoscienceworld.org/gsa/gsabulletin/article-pdf/34/3/499/3414562/BUL34_3-0499.pdf

p

CHADWI ------GAIL AE R E S LEM B PRO LAKE ------K IC GLACIAL W D A H C .w

xpS-a m a /-/on I ' “ Division -fines crossed 6y drainage I ■ Divisions not crossed j——^ Boundaries o f Arfiss/js/pp, dra,na_ge * “ /Direction of f/otv **| [-> fr7oif n o t cross/ny c-o tu'nn LofceS m /fcifics, Rivera in Roman Moha*vk cofutnn drops out a fter /ror~incipa f readi/ances o/ ice-fror,f 7~ime ratios no* cfosefy •sAox'f

F i g u r e 1.—Evolution of the Laurentian during Ice Waning