7 IMAGE EVALUATION TEST TARGET (MT-S) 1.0 \^^^ IIIM I.I u 1^ iii Mm M 1.8 1.25 14 IIIIII.6 Photographic Sciences Corporation ^^ fV V \\ ^ #> %" ^ "f> '^X c^ 33 WEST MAIN STREET WEBSTER, NY. 14580 (716) 872-4S03 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. Coloured covers/ Couverture de couleur I I Covers damaged/ D Couverture endommagde Covers restored and/or laminated/ Couverture restaurde et/ou pelliculde r~~| Cover title missing/ Le titre de couverture manque I I Coloured maps/ D D Cartes giographiques en couleur Coloured ink (i.e. other than blue or black)/ Encre de couleur (i.e. autre que bleue ou noire) Coloured plates and/or illustrations/ Planches et/ou illustrations en couleur Bound with other material/ Reli6 avec d'autres documents Tight binding may cause shadows or distortion along interior margin/ Lareliure serr6e peut causer de I'ombre ou de la distortion le long de la marge int6rieure n n Blank leaves added during restoration may appear within the text. Whenever possible, these have been omitted from filming/ II se peut que certaines pagef blanches ajouties lors d'une restauration apparaissent dans le texte, mais, lorsque cela dtait possible, ces pages n'ont pas dt6 filmdes. Additional comments:/ Commentaires suppldmentaires; L'Institut a microfilm^ le meilleur exemplaire qu'il lui a 6t6 possible de se procurer. Les details de cet exempiaire qui sont peut-dtre uniques du point de vue bibliographique, qui peuvent modifier une image reproduite, ou qui peuvent exiger une modification dans la mithode normale de filmage sont indiquds ci-dessous. |~~| Coloured pages/ Pages de couleur □ Pages damaged/ Pages endommagdes I I Pages restored and/or laminated/ D Pages restaur^es et/ou pelliculdes Pages discoloured, stained or foxei Pages ddcolor^es, tachetdes ou piqu^es Pages detached/ Pages d6tach6es Showthrough/ Transparence Quality of prir Qualit^ indgale de I'impression Includes supplementary materii Comprend du materiel suppl^mentaire Only edition available/ Seule Edition disponible □ Pages discoloured, stained or foxed/ P* I I Pages detached/ ["^ Showthrough/ I I Quality of print varies/ j I Includes supplementary material/ I I Only edition available/ 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 6ti filmies 6 nouveau de facon i obtenir la meilleure image possible. The c to thf The ii possil of thi filmin Origii begin the la sion, other first sion, or illii Theli shall TINUI whici Maps differ entire begin right i raquir meth( This item is filmed at the reduction ratio checked below/ Ce document est filmi au taux de reduction indiquA ci-dessous. 10X 14X 18X 22X 26X 30X ■^ 12X 16X 20X 24X 28X m The copy filmed here has been reproduced thanks to the generosity of: Metropolitan Toronto Library Science & Technology Department L'exemplaire filmd fut reproduit grdce d la gin^rositi de: Metropolitan Toronto Library Science & Technology Department 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 soin, compte tenu de la condition et de la nettetd de l'exemplaire filmd, et en conformity 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 fiim6s en commenpant par le premier plat et en terminant soit par la derniire page qui comporte une empreinte d'impression ou d'illustration, soit par le second plat, selon le cas. Tous les autres exemplaires originaux sont filmis 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 h^^^rre on each microfiche shall contain the symbol -^^ (meaning "CON- TINUED "). or the symbol V (meaning "END"), whichever applies. Un des symboles suivants apparaitra sur la dernidre image de cheque microfiche, selon le cas: le symbole -^^ 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 dtre film^s d des taux de reduction diffirents. Lorsque le document est trop grand pour dtre reproduit en un seul cliche, il est film^ d partir de I'angle supdrieur gauche, de gauche & droite, et de haut en bas, en prenant le nombre d'images nicessaire. Les diagrammes suivants illustrent la mdthode. m i ^ ^ ^ t 3 t a 1 • . •■ V , 4 8 6 b^. DGTABTMEHT Or THE IHTERIOR-U. 0. OEOLOOIOAL BUBYBT CHARLES D. WALOOTI, DIHBCTOK RECENT EARTH MOVEMENT IN TSK GEE AT LAKES EEGION GROVE KARL GULBERT SZTBACT FBOU THK SIOHTBEXTH AMSITAL REPORT OF THE SURVET, IgtS^ PART U— PAPIRS OdIBFI.T OF A THEORETIC NATURE WASHINGTON OOVBRNHBHT PBINTINO OFFIOK 1898 %"•- ■=/ •u t \\ RECENT EARTH MOVEMENT IN THE GREAT LAKES REGION. 11 Y GS^ROVE KARL GILBKRT. 685 ^, J CO XT H NTS. Illtroiliu'tidli (!()l OliservntidiiM liy Mr, Stiiiitz (iOl Kartli niDvei'iuiits iliiiiiij; the closing opoclis rl' thr^ Ploistoeene period (i03 EeiiSdiiH for lej^iinliiif; ;i iirogressivo moileiii chiiiigo as probable (11)6 Goiicnil jiliiu of invrstiyiition (!09 Available ilatii 611 (!agc icii .lings fill HcndicH fil2 Selortion of stations anhie ihanges resulting from the movement 63!) Plans for precise measurement 641 Selection of stations 641 Conditions controlling ci|nipment 643 Kqnipment 644 Treatment of observations 645 Suiiplemeut. — Investigation by Mr. Jloselcy 645 597 i v.- *,.- i " l.llSTRATliiNS. Plati: C!V. Kstiiiiry :il tli itli "f Ouk Oicliaril CroiiU. iiiliMMs County. New Yolk Fill. '••: !. Miip of l.aurcntiiin liiki'i '.II. Ma|i of tlir Iiiiii" lis sliorc lino . "I."). M:i|) ol' till' slioro linr of lireiit l.iiko NipissniK f tlie sMrl'iir<' of tlic Laiircntian laU' Oil. Annual oscillation.^ o il7. OHi'illations of the .snrfaco ol I.aki) Michigan dni' to ulian^is in 111' iilunio of tlio lako 9S. lliajirani illnstratiiiK niellioil of ini'iisnri n;; oartli movements '.19. Airan^icmen t of selected .still ions Vnga (lilt I ill I t;o.-> llO'.t r.io lilO t>i:t 10(1. Kelalions of tlu^ sliores of tlie (in'a tliron;j;li tln'ir ontlels 1(11. I'ropi t Lakes to tlie isoliases ilrawii ■d svHtenis of stations for tlie iireeise meas nr'jment ot earth nioveiuent.s . u'J'J (>10 (112 %■- \; y RECKNT EARTH MOVKMKNT l\ HEIJIOX. THE (;|{EAT LAKES HV (1. K. GIJ.UKBT. INTltODUC'ilOX. Tlie geologic liistory oi" the oiiitli shows that in all i)arts of its sur- face tiiere have been great oseillation.s of level. Modeiii history also records upward and downward movements of the land at various points. Tlie modern movements are of small amount, hut it is lielieved that they are of the same Idnd as the ancient. an;;ioal cliiiiigea in iiortheasteru Wisconsin: I*roc. Am. Ass. Atlv. Sri.. Vol. XVIII, 1870, PI). 200-207. 601 602 EARTH MOVKMENT IN THE GREAT LAKES KEOION. farthir Wist. At tlio inmith of Itoia Uriili- the Baiiie tliiiifr 18 oxbiliited, only to a ureiiter extent. From tliis to tlio west cnil of the lake not only doen thr liiko set k into the valleys of thi' stieani.s. liiit the waters are making rapid encniiich- s on the hanks. So ra])iilly is the fillinj; hack, that the depoHits of the streams mn.i. kci'p jiaie with the lillinf? up. The eonseiiucnce is. that there is a large .ii:ii:.|i 11(1 )ioiilane of the Iroiiuois shore line in the basin of Lake Ontario descends toward the south-southwest at au average rate of .'iA feet a 604 EARTH MOVEMENT IN THE GREAT LAKES KE(iI()N. mile, the slope being steeper at the north than at the south.' The Oswego slioie line, in the .same basin, slopes iu the same direction at the rate of more than 3 feet a mile. The Warren shore line, traced from huiv?., New York, about the sides of the Ontario, Erie, and Huron basins to Pompeii, Michigan, is nearly level in the Maumee basin, but rises northeastward with a rate gradually increasing to 2 feet a mile. Its northward rise iu Michigan is 1^ feet a mile.-^ The present south- ward iiK'lination of the water plane of Lake Algonquin, which occu- pied the Superior, Michigan, and Uuron basins, ranges from a few inches to .{ feet a mile.' Great Lake Nipissing, which occupied the same basins after the disa])pearance of the ice and had its outlet at North Bay, conformed more nearly to the inesent slopes, the general iuclinacion of its water plane being about 7 inches to the mile.-' Fm. 04.— Map of tlio TnK[Uois ahuro line. Jfotlerii wntpr Innlics nrc Hhadeil. A lino allows tln-bound- (iry oC tho uncii'iit lake, 'i'lic jiurullol curves are i.-^ohiiHca. On the accompanying maps of Lake Iroquois and Cireat Lake Nipis- sing (figs. 04 and !).)) the character of the tilting is shown by means of isobases, or lines drawn at right angles to the direction of tilting. All points on one of these lines have been uplifted the same amount since the time of the corresponding lake. If we think of the i»laneof the water surface of one of the old lak?8 as having been deformed by uitlift or warping, then the isobases are cont J. W. Spencer, Trnna. Roy. Sw. Canada, Sertion IV, Igsg, pp. 12U134 1 a. K, Ullbert, Sixth Auu. Kepi. CniniiiinHiont'i-H of tile State Keaervation nt Niasara. Albany. IHDO. »!•'. I!. Taylor, Hull. (ieol. .Soe. Ani.r.ia. Vol. VIII, 1897, p. 6.1. " K. II. Taylor, A Sliort Ul»t, I incFi 1,500 feet. Figures give soundings in feet. The watci wjyb aiu bharply incised in the plain. PaMial rofillirg is shown by marshes. Slack w .t lake lev(>t roaches to Cailton, above which the creeks ate shal'jw, MOVEMENTS IN LATE PLEISTOCENE TIME. (i05 stream cliaiinels as they apitroach lake shores. The streams reacliing Lake Superiro)lorn water bodies are sliadrd. A line HhowH tlio huiintlary nl" ttio aii(-)<-iit laUe. U'ho iiaralli'l lines are IsobaHeH. the lake they become deep and sluggish, the change usually occurring several miles from the lake shore. Stated in another way, each stream, instead of deboucliing into the lake, enters the head of a long, narrow bay or estuary. The origin of such estuaries is well understood. They are found on all sinking coasts, and their meaning in this region is that the land has gone down or the lake level has risen, so that the waters of the lake occupy portions of the channels carved by tlio streams in the lowland plain. This description applies to the greater number of streams entering Lake Ontario between tlie Genesee and Don rivers and to those entering Lake Erie between Cuyahoga River and Maumce 15ay. ludividual mention may be made of Oak Orchard, 60fi KAKTH MOVEMENT IX THE GREAT LAKE8 KEGION. Eifiliteeninile, and Twelveinile cioeks in New York, of Twelveiiiile aud Twentyniilc creeks aiid the Ciedir and Jlnmbcr rivers in Ontario, and of Kocky J{iver, Black Hiver. Vermilion l.'iver, Old Wonians Creek, Pike Creek, Turtle Creek, and < )tta\va Ki ver in ( )liio. Kven the largest rivers of the district, inclndinj;- the (ienesee, Niagara, Cuyahoga, and Jlauniec. have features indicative of the same histny.' ]{y reference to the ma]) (lis:. !»■!. ]). <">-) it will be seen that the outlets of these l;:i»es aie at their nortiieriimost ixiints, said this fact is related to the conditions of the stream channels. The water level of a lake is maintained by the balance between inflow and outtlow. It is just high enough to enable the outllow'ng stream to carry oft' the ex<'ess fr{;m inllow, and the height of water on all shores is thus detennined by the height of the outlet. 80 if thes basins are canted northward the out- lets are thus lowered with leference to other parts, and the waters recede on the southern shores. If they are canted southward, the out- lets are raised and the vntcrs are made to advance on the southeru shores. Reasoning from effect to cause, the fact that the lake water invades the new-made stream channels on the southern shores is evi- dence of the southward canting. It should not be ashumed that the "drowning'" of stream channels ia restricted to the traces mentioned above. Those tracts are specided because they fall within the range of the writer's personal observation aud are known to exhibit the phenomena in a striking way. It is believed that similar features may be found wherever the local con- ditions are favorable throughout the whole coast lines of Lake Ontario and Lake ICrie, about the head of Lake Michigan from Manistee, Mich- igan, to Kewaunee, Wisconsin, and about the whole of the American shore of Lake Superior. IJEASOXS FOK UEGAUDIXCJ A PKOfmESSIVE MOUEllJiT CHAXOK A« PUOJIAULE. Independent of the i)henomena described by Stuntz, there arc various considerations tending to direct attention to the question of the sta- bility or instability of the Laurentian area at the present time. The lirst to be mentioned is purely geologic. The epoch during which the overliow from tht upper lakes followed the valleys of the Mattawa aud Ottawa is detlnitely .associated with a certain stage of the Niagara River. The cataract of Niagara is at the present time increasing the length of the Niagara gorge at a somewhat ra])id rate. The fornnitiou from which the water lea])s is a liiin limestone (iO feet tliick, and beneath this are shales which are comparatively soft and weak. The cataract, by eroding the shale, underniines the limestone, which falls away in blocks, and these blocks are in turn utilized by the water as an instrument with which to grind the shale. WhirL'd about by the I .14" \' aiLiiKHT.] PROBAHILITY OF PROGRESSIVE MODERN. CHANGE. (i()7 water, tlie blocks not only woar away the face of the slialc clitf. liiit drill down deeply, .so tliat beneath the cataract there is a i>(>ol nearly or (jnite 200 feet deej). Working in this way, the cataract has extended the gorjie several hundred feet since the first accurate ineasurenu'nts were made, the average annual rate being between 1 and ."> feet. "With tlie jjpesent airangcnieiit of tlie drainage system the Niagara carries the surplus water from the basins of lakes Huron, l^rie, Miclii- gau, and Sui)erior; but when the upper lakes .sent their overtlow to the St. Lawrence by way of the Ottawa, the Niagara carried only the dis- charge from the I'lrie basin. Its volume was then only one-eighth of tlie present volume and its jiower was corres])ondingly less. It could not mo\o the great blocks of limestone which fell from. the dill', and, instead of scooping ont a dee]) jwoj, as now, it excavated a compara- tively shallow channel, whose bottom was cumbered with limestone debris. Owing to this ditference in method of erosion it is possible to discriminate the parts of the gorge excavated when the river was small and when it was large, and thns to determine the jdace of the cataracit when the outlet of Lake Huron was shifted from North Hay and the Ottawa IJiver to Port Huron and the St. Clair and Detroit rivers. That place is at the head of the Whirlpool IJapids, 11,(100 feet from the l)resent cataract. Assnming that the cataract worked at its present rate through this distance, we may compute the time consunu'd. At 4.J feet a year, it would be about two thousand six hundred years. ¥. li. Taylor, making allowance for various qualifying factors, esti- mates tiie time to have been not less than live thousand years.' When Lake Huron changed its outlet, the plane of its water surface exteiuled from the pass at Xorth I Jay to the i)ass at Port Huron, but the North Pay i)ass now stands 140 feet higher than the I'ort Huron. This dilference of altitude, anu>unting to G inches a mile, has, there- fore, been wrought within the period of about live thousand years. lu view of the gradual nature of such movements, this is not a long])eriod to assign to the measured chiinge, and it is natural to inquire whether the movement is not still in progress. Dr. J. W. Si)en ol'iil! Ilio ujiper Inkes into tlio Mississippi, by way of CliicaKO, just iis tlio iliiron w.itors wito lately tunioil from tins Ottawa into the Niagar.a drainai;i'; ami at tliu rrcent rate it would seem that aliont o. (100 oi- tl, OHO years at most will lie needed. 'I'lie ('lian;L;e of drainage slionid arrive liefore the cataract shall littve receded to BnlValo. ' A uliciit lilntorv (if tlie Omit I.nkca. 'I'lW. Am. Ass. Ailv. Sti., Vul. XLIII, 18l)t. ii.'JJIi. 608 EARTH MOVEMENT IN THE GREAT LAKES REGION. Ui I y Anotlibrcoiisidcratiou ot tlio saiiif tendency is found in the condition of tlie estuaries described in the lasr section. xMost ..1' tlie streams flowinj; into liiese rise in districts oi' unconsolidated drift and carrv forward in llond time a considerable load of detritus. TJiis is dei)osited in tlic estuaries, the coarser part making deltas at their liead.s, and the finer settling as mud in the deeiier water. The process tends to convert the estuaries, first, to marsiies, ami then to dry land, but in most instances little progress in that direction lias been made. There are a lew creeks rising in sandy districts which have succeeded in tilling their estuaries, changing them to marshes; but as a rule the delta at the head of the estuary invades it but a short distance, and the marshes which border it here and there at i)oints .sheltered from the Hood cur- rents are impassable except by boats, ami have the appearance of sub- merged flood pliiins. These characters, from their clo.se resemblance to the features observable along the subsiding i)arts of the Atlantic coast, give the impression that a slow flooding of the stream valleys is still in i)rogress. A third consideration is connected with the record of recent changes on the coasts of the continent. It lias long been known that the Atlantic coast south of Connecticut is subsiding, and I'rof. G. H.Cook was able to determine the rate in New Jersey as about 1' feet a century.' Dr. Robert Bell has recently collated a variety to 7 feet a century. If these two movements are parts of a general movement att'ecting the northeastern part of the continent, then the Great Lakes region, approximately intermediate in positicm between the rising and sinking areas, should be found to exhibit a southward tilting. These various facts, all tending in one direction, are sufVicient war- rant for the working hyiwthesis that the tilting of the lake region demonstrated by the slopes of the old .shore lines is still in progress; and the writer, who lias for many years been interested in the problems of the Great Lakes, has made repeated efl'orts to secure an investiga- tion by which the hypothesis might be tested. The mode of investigation first suggested was the establishment of elaborate observation stations at three points — Port Huron, Chicago, and Mackinac. Hy a suitable series of ob.servations at these points, the relative heights of benches might be established with high precision, the water surface being u.sed as a leveling instrument. Then, after an interval of one or two decades the observations might be repeated and any changes in the heights of benches due to dift'erential uplift detected. The matter was submitted in 18!(0 to the Superintendent of the United States Lake Survey and to the Superintendent of the United States Coast and Geodetic Survey, but, though it was received favorably by the latter i,flicer, the work was not undertaken. ' Am. Jour. Sri., 2(1 series. Vol. XXIV, 1857, »Am. Jonr. Sci., 4tli series, Vol, I,18!«!. PLAX OK INVKSTIOATION. COO Other plans were then ('(insirh^it'd, aiid it was lliially dpcidcd to makii a study of existiii}:' ri'iords of lake level, and, if necessary, sin)|)lenieiit tlieni by additional observations. The results of this investigation arc set fortii in tlie folh)\viny payes. <;KXKI5Ah I* I. AX or IXVKSTKiATIOX. Variations in tlie heiglit of the ocean level at any place depend ehietly on tides, winds, and atmospheric, pressure. I>y nu>aiis of lonjj scries of observations tiie ellect of tliese disturbing factors can be eliminated JAN. FEB. MAR. APL. MAY JUNE JULY AUG. SEPT. OCT. NOV. DEC. SUPERIOR MICHIGAN- HURON ERIE ONTARIO 1 ^ , ~"~^ y /" \ ^^.^ « y'' y /■"^ -^ y / ^■\ -- "~'^ \. ^ / / \ \. "^■■v. -..^ /■ -•- ^N s, 'V ~"— ■ . / \ \. / r \ — . — .^ ^ Fln.no. — .\imual n.-rilljitioiis nC tlii'siirCai'i's i-l" Itii' L.-iiirnitiiui liikcs. ('niii|iilril IVmn iiioiillilv riifanH puliUslx'il li.v IIm^ Cliifl' III' Kniiiiiii's. I'.S. A. Kuril viTlli-il »|iiiii' ivpri'siMil.i ii.\ inihis. Tlio oliserviitioiiK liii- I.aki' SiipiTliii vit the iiiTiml ISIi'J-lH'.l.l; I'm- .Mirliiniiiilliiniii. ISlili-lHU.'i ; Ini' Krii', l.s5.'i-|8'.t.'i lor (lutnrio, l«ill-lMi."i. and a mean level obtained which is ])ractically uniform from year to year and ilecade to decade. The height of the water surface must depend also on the (|Uantity of water in the ocean, but the actual varia- tions of volume are so small as compared to the extent ot the ocean surface that the resultinj^' variations of level may be nej^lected and the mean level used as a standard for the discussion of ditlerential movements of the earth's crust. With tlu! Great Lakes the case is materially ditlcient. There are variations due to wiiul, atmospheric pressure, aiul tides, but when these ha\ t* been eliminated by long series 18 (iKOL, PT 2 39 ! 610 KAHTII MOVKMKNT IN TIIK (iUKAT LAKES KK«iI(»N. i of oliscrviilions tlu' icsiiltiiiK irit'iiii li'vcl is far I'loin (•onrftiiiit. varying from season to si-asoii anil year to year with tlic voliiiiie of water. lu eacli lake tliore is an annual clianyc of more tliaii a foot, depending on tlie seasonal ineciualify bet wei'n gain by ])recii)itation and loss by eviii)oratic)ii (li«-. !M'.), and tliere is a still greater change resulting from tin' ciiniulative elfect of series of dry and series of moist years. The records siiow that the water surface in each lake has been .several feet higher in some years than in others. (See lig. !>'.) For this reason the water surface of a lake does not afford a datum ]ilant' by reference to which the elevation lU- subsidence of coasts can ho directly determined. Jortu'iately, however, there is an indirect ISIO ^ r> mo A -^ iPy A A b\ /Y r k./-' ^ A A /\ i\ rA ^/rsA ^ A /N A r ' J >/ J ' Fill. UT.— (Ini'illaliiiiw 111' tlin siirfilr i»f I.jilic ^licliijinn due to rlmiipcs in tlio viilunit' «!' llic lake. A\'isr(i!isiii. t'nim Angtist. IH.'ll, to .Turn', IH'.t". Kiicli )ior-i/outal method by which ])ra(!tically the same result may be attained. If the mean level of a lake surface be determiiu'd for two parts of tlie coast at tlie same time, these tv. o planes maybe regarded as parts of the same level surface, and, through referent^e to this common datum, fixed objects on the land at the two localities can be compared with each other so as to determine their relative altitudes. If, then, after au interval of time, the measurements are rejieated, a change in the rela- tive height of the fixed objet^ts may be discovered land measured. The investigation described in the following pages made use of this method. l'"iti. y?*. — niaj;rain illustrating method of ineasnrin^ earth nioviMurntH. The fundamental i)riuciple of the method is illustrated by the dia- gram, flg. !)S, in which A B is the cross jjrofile of a lake basin. At a certain time the mean plane of the water surface occupies the posi- tion .W. r>y means of the engineer's levtd it is ascertaintd that a bench mark A has a certain height above the water plane at X, and that a bench mark 15 has a certain height above X'. The difference between tliest; two measurements is the difference in altitude between A and 15. After an interval of years the measurements are repeated. The water plane then stands iit some ditVerent level, say YV, The height of A above Y is measured, and the height of 1! above Y'; the AVAILAIILK DATA. (ill din'f'ience betweoii the two iiicasuienuMits ftivi's the relative liciKl't "f A sinil li. If earth lUDveiiiciits have owiiriTil duriiij;' the interval hcfweeu the two sets (if nieasm'eiiieiit.s, the seeoiid (leteriniiiatioii of tlic com- j)aiative hei;;lit of A ami H will ditfer from the lirst determination, and tiie amount of ditl'erence will nKsasiire the ditferoiitial earth move- ment. \V\l TjA lUu !•: I ) A'l' A . (iiii/r niiiliiKju. — In order toeliminate the temporary elfecrtsof disturb- \\\ii factors.it is necessary to have a series of observations ot tlif li('i;;ht of the water surface at i-ach of the localities compared. The ;;afi;es by T' " n — 1 — r — 1 — 1 — 1 — 1 — 1 — 1 — r^ N i .iuvA/VV^ V\a .^a:^^ - -^n:^A. . A A^ • j>re»i'titrt a caU'iiilai' year: t-arli vf-rtirnl spare, iinu I'liul. means of which such observations are made are of various kinds. One of the simplest is a graduated ])laidj, llxed vertically by attacliinj;- it to a dock or other structure, so that one end is above water and the other below. Sometimesthe plank isomitted and the n Lake Ontario; i'ort Dalhousie. Niajjara, Cinirhitte. Oswego, Sack- etts Harbor. Tiu'se records are for the most part i)ublisued in tlie form of montldy nu-ans, but the iuoses of comparison. I'revious to LS71 siu'h connection with benches was not nnide by the United States ]iake Survey, or, if nuide, the rec- ords are lost. There were, however, certain stations, notably Chicago, Jlilwaukee, Cleveland, i'ort Colborne, IJulfalo. Charlotte, and Oswego, at which this nnitter had received attention. The structures at Chi- cago on which the early bench marks were made are thought to have afterwards .settled.' At Milwaukee the early bench marks no longer exist, and although there is reason to believe that other benches were substituted with care, my researches have not discovered a satisfac- tory reiiord. Tlie same remark applies to l>uftalo; and the record of theorigiiuil bench at Charlotte has been lost. At I'ort Colborue and Oswego the zeros ol' gages are pernninent structures, which have ])rob- ablysiitVered no change; and at Cleveland, although the oldest benches no huiger exist, it is believed that the record of transfer is complete and satisfactory. In ISTO Gen. C. I'.. Coinstock was jdaced in charge of the United States Lake Survey, and the scientific methods introduced by liiiii 'ReiHiit 111! l.liii;aj;i) (.'Ity Daluui iiiiil Cit\ I!biicIi .Miuk.H. In W. II. Uwlj,'(s, Cliicn),'ii. 189j. i SELF.CTIOX OI' STATIONS AM) YKARS. ci;? included the t'stablisliriR'iit of a complt-to system of hciiclH's in <((nnni',- tion Willi tli(! K'lijics. Krom ISTli until the eoniplt-tion ol' the (ield worU of the Liiko Survey there was itn annual inspeetion (if the ;;aKes, and the relations of their zeros to the bench marks were redeternjined as otten as seemed necessary. I'rom ISTl to 1S78 tlio mipervision of {(ajii's and tlie reduction of records were in cbiirne of Mr. (). It, Wheeler, and from IST'.t to l.SSL', of .Mr. A. I.'. Klint. Tiie results of the present investi^'iition are largely indebteil to the care and thor- oughness with which these engineers jjerformed their work. .SKLlOmoX OF STATION'S AXn 'N'KAIJS. Tnder the general method outlined above the first stej) was the selec- tion of suitable pairs of stations on the shores of the various lakes. As '■-J^r-lp^^^ f'^X^A "-^ '\^ J J fj^j^Au \ i / ■"^ScuJceJt^'^^^ / -V \^ yi ^ CharlottB 9 1 A 7l ColSorne ^^yit'leycLoTi et Fin. 9n. — AiTiiiipement of ficlected Htations. the geologic data indicated a tilting of the land toward the south- southwest, or, more precisely, in the direction S. Ii7^ W,, it was desirable to have e.ich pair of stations separated by a long distance in that direction. As the hyi)othetic change was exceedingly slow, it was desirable to compare observations separated by the longest practicable time intervals. It was essential that the gage readings before and after the time intervals be accurately connected with the same benches. Consideration was also given to the fact that the results might be vitiated if use were made of observations taken during the ])revalence of storms, when the water is sometinies driven by the wind so as to staiul abnormally high on certain shores; i.nd in order that the use of such observations nught be avoided it was important to select years during which the force of the wind was daily recorded. With these cu I'.AHTir ^WOVEMKN'T IX THE (JREAT LAKES REGION. consiileratioiis in view the available data were examined, and tlie following si'lectioii was made of stations (see li<;'. !•'•>) and years: For Lake Ontario, Charlotte and Sacketts Harbor, 1874 and lS>J(i. For Lake Va-w, CIe\elan'o comparison was undertaken lor stations ou Lake Superior. SPKCIATi OJJSKIIVATIOXS IX 1 «()(!. Certain of the .s('lecteassagc of cyclones; and the impulses thus received are not (]uickly dissipated, but cause a long-continued swaying of the water. In large lakes these oscillations are so enduring as to cover the interval from one disturbing imjjulse to another, and keep the water i)erpetually in motion. Xear the ends of the lakes and in bays witii gradually con- verging sides the range of oscillation may be as great as I foot, and it ordinarily amounts at all lake stations to from 1 inch to I inches. For this reason a single observation may not a])proNinnite closely to the mean level of the water, and tlu^ actual nu'an level can be determined only by a series of observations at short intervals. In arranging the work of 18!Mi the observers were instructed to record the water level every five minutes for an hour each morning and evening of all days when the wind was light; and at Sacketts Harbor, where the seiche has an exceptionally long period, the length of the series was afterwards increased. ; I OILllKllT.l SACKETTS HARBOR AND CHARLOTTE STATIUNS. 615 I)ISCrS8I()X OK DATA I'KO^I PAlItS OF STATIOXS. SACKETTS HAKBOR AND CHARLOTTE. Ill I.S7-4 the zeros (if ga.ues at tlicse stations wen' points maiki'd on doc'ivs, and readinjis were made by means of j;iaduated vertical rods attached to lloats. They K'ive tlie distance of the water surface below the jiaiie zeros. At t'le time of each observation record was also made of th(^ direction and force of tlie wind. The work was under the direction of the Cnitcd States Lake Survey. Mr. A. Wilder was the oi)server at Charlotte, and ^Ir. Henry Metcalf at Sacketts Harbor. The gage at Charlotte was put in place in Novendiei', 1S71, and the measurements showed its zero to be 32.7 feet below a bench mark. In January, I.ST.'J, its zero was found to be .'iL'.O.'i'J tcei; below the same bench mark. On ^[ay 11, 1874, it was a.uain com[)arcd with tlu' bench luark, and the dill'crence was fouiul to have increased to o.i.OO.". feet. It is probable that this change of .Ot 1 foot was occasioned by the set- tling of the dock to wiiich the gage was attached. A manuscript report dated February .'i, 1875. says: "The baidc is here partly of tim- bers and partly of earth. The earth has been washed out and has fallen away from the timber in some jilaces."' Tia^ gage at Sacketts Harbor was also found unstable. The report of an inspection in .May, 1874, states that the zero of gage " has been lowered ((.."i."),') foot:"' and a rejiort dated February. •>, l.s7."), says: "This gage is fastened to tiie timbers of an old and unused dock. Tiie whole structure is ([uito dilapidated and unstable." The instability of gages determined the selection of time for the comparison of stations, I'.otli gages haxiiig been comiiared with benches in May, 1S74, that at (Jharlotte on the 11th and that at Sacketts Harbor pro'oably on the 1 1th, the computa- tions were based on a period including these dates. Within this period selection was made of those times of observation wlien tht; wind force at both stations was less than .'i on a sciUe of lo. Thus treated, tlio observations of ol days gave 51 eompaiisons. X GUI EAIiTir MOVKMENT IN THE (iREAT LAKES KElilOX. Comjiuldlioii of the hriiiht of Ihv giKjf :iio ill Sarlella Ilnrhnv, Seiv )'i)il:, above llir i/age :civ al Chuilolle. -Ycif York, in thv epiinij of 1x71. Iliiu ■ ri'iHlin;;. D.i.v. Hour. Sllrlivltd Harbor (■Ii:irl..lli i DillcnnciN | 1 1874. i Fert. Fftt Fftt. A])!-. 17 1' 11. 111. T). i:! :i. i.-i 2.2S ' Ill !i ]p. 111. :>. liii a. (11 2. ;f2 22 1) li.iii. 5. 2;f 2. 118 1 2. 25 2:) 7 :i.iiL. .-.. It; 2.80 2. :!(! 2 |i. III. .-.. l.s 2. 7X 2. 10 21 7 :i. 111. ."i. 2(1 2.8(1 2. Id 2."'i 7 11. 111. 5. l.s 2. .SO 2. :i8 2 p. 111. .■i. 1 1 2. .SI 2.3(1 27 7 ii. III. ."i. IIH 2.78 2.3(1 !• p. III. 1.K8 2.78 2.10 2X 2 p. 111. .-.. IM 2.87 1 2.20 ! !l ]i. III. .">. no 2.88 ! 2.18 May ;! 7 ii. in. .■>. ()(i 2. il") 2.11 1 7 .-1. 111. .-l.lli 2.!!.". 2. 48 L' p. III. 5. (HI 2. !U 2. 15 i» ji. m. -). 10 2. !t4 2. 22 5 i) ji. 111. ."). IH 2.91 1 2. 27 6 it p. 111. 5. 12 2. 85 - 2.27 7 7 a. 111. .-). 0(! 2.S7 2. li) X !l p. 111. ."■. 17 2. 8.-. 2.32 11 '.1 |>. 111. Ti. 17 2.82 2.35 11 'J p. III. .-.. 12 2. 83 2. 29 l."i 7 11. 111. "). 17 2.84 2.33 2 p. III. ."i. 2r> 2.82 2.43 IK 2 p. 111. n. la 2.82 2. 31 !l ]). III. 5.18 2. 8.'i 2.35 20 7 11. 111. "i. 0.") 2.83 2.22 2 p. 111. r.. 10 2.82 2.20 21 7 11. III. 5.02 2.81 2.21 22 11 p. III. .^02 2. H2 2. 20 21 7 a. III. .'). 12 2. !)5 2. 17. I 2 I). III. 5.08 2. !)2 2.10 !' p. III. 5. 08 2. ill 2.17 20 !» p. 111. ,"i. (18 2. 80 2. 22 27 1 7 11. 111. r.. (M) 2. 80 2.14 !* p. 111. 5.10 2 84 2. 26 2H 7 a. 11!. 1.!I8 2.83 2.15 U p. III. 5.02 2.83 3.19 QiLBEiti.] SACKKTTS HARI30K AND CHARLOTTK STATIONS. fil7 Comjiulation of the hiiyhl of flieijagc :ero at Siifketla Ihirhor, Xiiv York, almri- tlie ij(i(je :ero illen'iuo. Snckettn Uarbor. ; Chiulolte. 1 1871 Ftet J-..V. 1 Fret. \ May 30 7 ii.iii. .->.oo 2.81 2,11. j !t ]). m. 5. OS 2.82 2. 21 ; ,)mii^ 1 !t ]). 111. 4.117 2. 82 2.1,-) >> 7 11. 111. ,"), (10 2. 8;{ 2.17 !) ]). 111. .-.. 02 2.83 2.17 I '.) )i. m. 5.10 2.82 2.28 Ti !t ]i. 111. .-.. 00 2.78 2. 22 <; 7 :i. 111. ,-).00 2.78 2. 22 !l ]>. til. .-). 0!) 2. 70 2. -M 7 7 ii. 111. ,-.. 1(1 2. 711 2. ;n '2 p. 111. .■>. 00 2.81 2.11) 8 ',1 |i. Ill 1.117 2.81 2. 10 it Mciiii 7 a. Ill 1.117 2. 82 1 2. 15 2. 217 J- .008 1 111 189(1 the si'S't' i'f CliiU'lotte. was a graduated plank s])ikpd to a pile just uovtli of tlie western abutuicnt of the Koine, Watertowu and Ogdensburj; Hailroad biidge. Tlie readings give the distance of the water surface above tlie zero of the gage. At Sacketts Harbor the arraiigeiiieiit was similar, the gage being .spiked to an unused dock. The observer at Charlotte was .Mr. .1. ^Y. Preston, harbor master; at Saekett.s Harbor. :Mr. Wilbur S. McKee. Observations were made morning, noon, and night, tlie morning and evening ob,servatioiis being extended into series whenever the water was so little agitati>miliiin ..f ri'iKlin^s. DilltTi'l.i'*'. Bate. Sncki'tts Harbor. Ch.ir- 1 l< .^.ick. Ills Clmr- Ii..r. Siickitlli ll;ilh..I-. Clinrlottc niiarliittf iii1nii.'4 Sack- ill« Harlior, 1896. 1 1 rret. Frft. Alls;, s 7. 1.5 :i. 111. 7 a. 111. iH i:i f.!l8l (1. iW-' —0. OL'2 S (t. .'!() ]). in. li p. 111. i:! ii; (i.»12 ().!«! (1. OL'L' Sept. !l .~i. ;tll ». in. 7 .'i. 111. 13 i:i 0. :i.".i 0. i-M (1.077 n 5.(1(1 p.m. (1 ]). 111. i;i i:i (t. 'J7(l 0, ;!(iH 0. Oils Oct. 27 X. l."> a. 111. 11 .... 7 11. ill. \r, u —0.1 IS —(1.0 IS 1 0. 1(10 \Iej -f 0. 05.'. 1 ±o.ou The beiicli at Olmrlottc is a iiiarlv (Hi tlic iiiiper surface of tlu' water table of till' old lijilitliimse. The walls of the hiiildiiif;' show no ciacks, and there is every reason to believe the beiieii stable. On May 11, 187 1, the zero of gajj^e was found by Mr, !•>. S. Wlieeler, assistant engi- neer I'liited States Lake Survey, to be .'>.'?.003 feet below this beiieh mark. On .liiiie.'iO, ISOO, f leveled from the zero of the present nage to the beiu'h mark, obtaining 38. !>.')(» as the mean of two measurements. On July 11. bsi)7, .Air. Warner W. (lilbert obtained 38.)).14 feet as a mean of two measurements. The only bench mark existing at Saeketts Harbor in 1S71 and l.S!H» is a point on the upjier ourcr edge of the water table at the northeast eorner of the stone building known as the Masonic Temple. lu May, 1871. this was determined by Mr. Wheeler to be IL'.^L'.") feet above the zero of gage. On June l.'S, iS'.Ui, by duplicate measurements, I found it to be i;().42."» feet above the zero of the i)resent gage. The building bearing this mark rests on a foundation of bed rot.'k, but -.levertheless has yielded to such extent that its walls are cracked. 1 was informed that the eraekiiig and repairing of the walls took place some years previous to 1871, and regard it as probable that there has been no change since that date in the height of the bei.ch mark. ciiiHKKT.] SACKKTT8 HARliOR AND CHARI.OTTK STATIONS. 019 These sovcral data arc combined in tlio Ibllowinf,' table: CoiHjiHlalion of till- heiijht uf Ihe henili marh- .S is not describoil in the records I have seen. Neither is the name of the observer given, bnt various circumstances indicate that the readings were made either by Ool. Charles Whittlesey or under his immediate direction. The readings give the distance of the water surface below the high- water level of 1838, ami that level was adopted by the United States Lake Survey as the ]ilano of reference for all observations on Lake Erie. At Port Colborne the upper sill of Lock No. 27 of the Welland Canal was the zero of nn-asurement. and the measurement was made by the lock master, .Mr. John Mc(iillivray, by thrusting a graduated jjole into the water until the end rested on the lock sill. As the reference point at Cleveland was above the water surface and that at Poi't Colborne below, their difference in height is obtained by adding the two readings. Most of the observations at Cleveland were nmde at 8 a. m. and the observations at Port Colborne at noon. At Port Colborne the direc- tion of the wind was recorded; at Cleveland, the direction and iovv.e., I do not know the scale of force emjiloyed, but the record numbers range from (» to 5. All observations at both stations were rejected wlieu the wind force at Cleveland was recorded as greater than 1. Th(^ gage zero used at Cleveland in ISOo was the upper edge of a cleat nailed to a jdaidc forming one wall of a well in a wharf. From I Ann. Itciit.Chli'IcilKnginciTii, U.S.A., lur 187-, i>\i. 10;i;), 1U3.\ 1040; 1873, up. 117:i, Illi'J, lliU, (IIUIKIIT.) PORT COLHORNE ANIl CLEVKf.AND STATIONS. r,2\ this till' ohsorvia- measured to tlie water surface in tlie well with a graduated rod. Tiic ga^e zero was set at the level of hinh water in 1.S3.S, wliieli is mentioned in the records as -tlie i)huie of reference."' Three observations were made daily, at 7 a. ni., 1 p. ni.,and 7 p. in., the work heinj;' under the direction of the liiited States I'lnjiineers. At I'ort C'oll)orne observation was made by means of a lloat connected through a chain with a counterpoise, and was therefore indirect; but the readings were ehei^ked by occasional observations with ajiole, after the method of l.sr»S. An index on the connterpoist^ was so adjusted as to indicate on a scale the depth ol' water on the lock sill. I inspected the gage in l.S!»(i, linding it in close adjustment, except that an error in cither direction of a fraction of an inch might arise from friction. The observer in 1S!»."» was ,Mr. .lohn llenshaw. Jn the following table the readings at \\>vt '^'olborne, which were ict^orded in feet anil incl",'s. have been convertetl to feet and hundredths. The record of wiiul at the two stations was the same as in l.S,"»S, and there were also available the wind ami ])ressure observations of the Tnited States Weather linieau. From ail inspection of these data three jjerious were selected tor coni- liarison: June L'S to -Inly .'!, July IS to liS, and August 3 to IS. These lieriods are so related to the tidal cycle as nearly to eliminate tidal error. ■^1 CiniijintitliDn ftf' llif! Iititjiif nf thr ' plinir of rrl'irt'iu'r ^' at ^'Irrrltanl. (thiti. nh'H'r fhi nill (if I. ml,' \o. ,.'.' '(/' llic U'eUantl I'tiunl iil I'art (iilhiirnv, Onlario, in IS'oS and 76'.'',7. Diilc'. 18.'W. Alljf. 20 Si'pt. 1 :) fi (( u ii> :iu ; Oct. 4 i 10 I i?ii: at iiiii ill Clfv... I'.iiiCcil IiiidI. Iinnit'. I;i;niiii;is ill Cii'vi'Iiirnl. Dill. It. 8.11 S. ,1 .'i. 8.(1 9..'. 11. 1 I..'. II, !l l.li 3.U i n. In. u : u i; U (1 11 II u :; II I II (I l:l 11 II 1 U U n .■( i:i 11 i:i '.I U :i U II i:i 11 i:i .'j U II 11 -J II .•. 14 II 13 lU 4.0 ■-'. II. l."i !,■. l.'i U. .'■ II ,-..11 II 11.11 11 11.2 1 1 ^■. II 14 .^.L' 14 111.4 II .■(. 7 11 0,7 1 1 ■<. l;i 11. :i II '.1.7 14 II. :i 14 11. H 14 II. .•> l.T 11.11 ir. :i.8 1,". 7..') 1.) 1.0 15 1.0 I 7 ii. 111. l|..iii. 7]i. 111. 1 Menu, 'ml C.il- liiniH-. Mini. r,,i. IWI. IWI. IWI. /■.,/. F.'l. L'S :i. :io :i. .-.o 3.411 111. nil 14.3-.' •Jll :i. :<:< :i. .-11 :i. ;i3 :i. .■.3 10,7,-i 14. 'JS :iii :i. .'m ;i. .'11 3. y.i !>. j.i 1 1 , lis 14. Ut 1 ;i. (Ill ;i.4ii :i. .'ill 3. Ill 111. 7.'i 1 1. l.-i *_' :i. 47 :i. 4.". 3. :i:> 3.411 111, S3 14.32 :i .;. .'i7 ;i. Iiu 3. r.s lii.lli; 11.. -lO IH ;i. .'.7 ;i. cii 3, .■i:! 3.73 11.1111 11.73 HI ;i.!i5 :i. .-.0 3. 70 3. OS 111, 7,-. 14.43 •Jll :i. .^0 ;i. OS 3. .'.:i 3. 07 ill:,-. 11.112 '_'! ;i.:i:i :!.87 3. (VI 3.01 11. ml U.iil 'J:; ;i. iiu ;i. .'.II 3.011 3. 07 111.75 14. 12 •j:i :i. I'll ;i, 70 3.70 3,07 111. S3 H..MI ■-'1 :i. ii.-i ;i. on 3.711 3. i;ii 111. S3 11, 111 nr. :!. ~- :i. 7:i :i. S'J 3.70 10.07 II, 13 X :i. "11 3. 7:i 3. 03 3.00 11.1111 11.011 ■J7 ;i. 7.-. '.'. "11 3. 3'.' 3. .V.I 10. '12 14. .-il •-'« 3. .•lO :i.4:i 3. Si) 3. .'■iH 111. 7.-1 14.33 :{ 3. -It 3.,«11 3. Sil 11,1111 ll.sil 4 :i. 811 ;i.7:i 3. 73 3. 7.-. 111. 7.'i 14..>l ,-, ;i.:3 ;i. iw 3. OS 3.711 111. .".-i 14.28 n ;i.!iii 3. 73 3. 70 3.7S 11.33 13.11 7 3.03 3. IP.-i 3. 00 3. S3 11.7.-> 15. .'iS G22 KAHTii mo\'Kmj;nt in the gkeat l.\ki:s kegion. Cmni>iit,iti r 14. 501 ( ■ . 022) ) I tl IllilllCi 1" ri'l'f ri'iHi aliovt) Tort Colbnnui loek- .ill- 1H58 .. 14.800 .t.057 1 1 MI'i .. 14.50 ±.022 . . 0. 23 ) ±.00 Tlie zero of gage at Port Colborne, being submerged iiiasoury, i.s of iiiiqne.'^tioned .stability. The canal was constructed in 1S33, and if any fit'ttliiig followed coii.strnctioii it was doulitless coiiiplote bel'Ore LS.jS; but the appearance of the inasoni'y above the water gives no sugges- tion of yielding. Tlie earlier v.-ork at Cleveland was connected with several beiuih inailis, all of which have been ilestroyed, but before the disaiipeaiance of the last one the datum was transferred by leveling to otliei' points, Tlie chain on which the record deiiends is as follows: 1. "Top of coi)ing of the northeast wall of the Ohio Canal lock where it Joins tlie river." The high water of 1, 1S75 (as sliown by niiinns(n'ii)t records in tlu- oHieeot'the ('liiet'ot' l"-ngincers, U. S. A.), Assistant ICngineer T. "\V. Wri<;lit, I'nitetl States Lake Sur- vey, leveled from fliis bendi nuirk to the canal lock copinn' (1). (indinjj tlie diliercuice (1 above L') to be .'[.(IT feet. This bench mark is still in existence. The walla of the building are cracl; 'd in such manner as to indicate some settling of the northeast corner, and the broad llagstoiio on which the bench is marked stands (in 1SK7) (t.Ot foot lower tlian the next stone of the water table toward the west. As the lower stone supports part of the building and the higher stone cari'ies no load, the latter maybe assumed to show the original level of the former. It is impossible to say whether this settling atVeets the record of water levels. The building was erected iu 1812, and istherelbre ,'),"> years old; it was :V.i years old in 1875 when the datum of levels was transferred to it. The datum remained with it eighteen years, until ISIKS. If settling has jirogressed at a uniform rate, the datum was affected (»,(i ■ ". foot, but it is e(|ually possible that the settling belonged to the early history of the building and that a condition of practical stability was reached prior to 1.S7."). ;{. " ISoltom of west angle iron, on bottom of ncu'th longitudinal jdate girder, middle of first full depth bent, close to stone pier, new L. S. v\: .M. S. It. K. drawbridge, now [1S9.'5] being finished. " As the briflge is symmetric and reversible, this description applies to two dilferent points, but measurement shows that they have the same height. It was copied from manuscrii)t records iu the I'niteil States ICngineers oBice at Cleveland, courteously placed at my servicer by Col. .Tared A. Smith. The records show that in .Tune, IS'J.'J, the bridge bench (.'!; was connected by leveling with the .Tohnson IFouse bench (2) aiul also with the gage zero, and that the gage zero was checked by the bridge in ISilG and found correct. The gage readings in IS!*") (used iu our comiiuta- tions) are thus referreil to the bridge bench. Thelieight of the bridge bench is given as l.'.H feet above the "plane of reference." and by implication as 1.71 feet above the .lohnsou House bench (2). The draw- bridge rests on a stone pier many years older than tlie present bridge, and there can b(^ little (|uestiou of ils stability. In these re(!ords of bench marks and levelings in Cleveland there is certainly much to be desired, bur. the ]iiesumption is nevertheless in fii\or of good work. It api)ears from the eomjiutation that the ground at Port Colliorne has risen, as compared to the ground at Cleveland, (),2.>!» foot, or about 2; inches in thirty-seven years. The probable error of this nieasure- nu'ut, as indicated by the discordance of gage data, is three-fourths of an inch. Asa check upon this result, a third computation was made from gage readings iu the summer of 1S72, a year in which the gage zero at Cieve- V: 624 EARTH MOVEMKNT I\ THE (iUKAT LAKES UEfilON. liiiiil WHS roiiiiccicd with tlie ciiimllork bi'iicli iiiaik by instruiiu'iital Icvcliii;;-. That c.piniiiitatioii j;ivi's tor the hi'i-;ht of the ])laiie of refer- (•net' at ('Icvchiiid ahovf tl,.- lock sill at Port Collionu' 11.711 fi-et. If wo assunif a j;ra(liial ciiaiif'*' IVoiii is.js to IS!ir>, and Interpolate between 1 l.sttO feet, tiie (leterniinatioii lor IS.'iS, and 1 l..')(il, the determinatioTi for iS!»."), wo obtain for the snninier of 1.S72 the value 1-1.710 feet, which dillers from the result of that year's observations by only (MHt-l foot. The observations on LakifrtTillri 7h. m. 1 Feel. 1 1>. m. Feet. ]i. m. Ferl. 7 a. ni. Fret. 'i p. III. Fret. U |i. in. F.rl. Fe.l. IHTt! Feet. ! Feel. July 11 2.17 1 2. 23 2. 12 7. 22 7.31 7.47 5. 05 5. OH 5. 35 12 2. 12 2. 03 2. II! 7.51 7. 25 7. 22 5. 42 5. 22 5. (Ml 13 2. 20 2. 05 2. 35 7. 35 7.37 7, 35 5. 15 5. 32 5. III! 14 1. i»5 2. 12 2. 10 7. 50 7.41 7.30 5. .55 5. 29 5.20 15 2. Iti 2. 15 2. lit! 7. 33 7. 35 7. ,30 5. 17 5. 20 .5.21 111 2. 13 2. 12 2. 11 7.31 7. 25 7.40 .5. 21 .5. 13 5. 20 17 2.15 2.20 2.21 7.37 7.37 7.30 5. 22 .5.17 5. (I!l 18 2.20 2.07 2. 20 7.21 ■ 7.25 7.28 5. 01 5. 18 5. US in 2.18 . 7.31 5. It! All}.-. Iti 2.29 . 2. 20 7. .50 [■■;:35- ,-). 21 5. 15 !7 -'. Ill . 2.21 7. 33 ' 7. 50 5. 14 5. 20 18 2.23 2.02 2.27 7. 50 7.35 7.40 5.27 5. 33 5. 13 19 2. ''2 7. 45 5. 23 L'O 2.18 2. 23 7. (.2 1 . 3i 5. 4 1 5. 1 1 ''1 2. 25 2.20 2. 21 7.41 7. :!4 7.37 5, 10 5. 1 1 5.13 22 2. 32 2. 1« 2. 33 7.3X 7.51 7. 10 5, 11(1 5. 33 5. ( 17 23 2. 2(1 l.'Jl 2. 10 7.38 7. fiO 7. .50 5. 12 5. 1)0 5. 10 I'l 2.25 . 2. 42 7.4!l 7.40' 5.21 5. 07 .M. is;i Lira ti.Hlli ■M A. M. .-.. 70.'t 1. 3X5 7.088 29 A. M. 1'. M. .-|. 170 1.3.51 1.375 I!. 824 6.935 j Aug. 1 ; ^1 A. M. 1'. M. 5. 1120 1.2!'7 1.125 (i.733 j 7.015 i 7 1'. M. "). 117 1. 120 6.^07 'J A. M. .-...-, 19 l.ilS 0. 907 11 A. M. .">. ."i7."> 1. 133 7.008 j 14 A. M. 5. :w8 1. 455 0.793 1 20 ' A. •■. 5. .•>87 1.310 0. 927 20 I'. y\. 5. 571 1. 40() fi.977 21 A. M. 5. .-|.-)S 1.330 6. 888 1'. M. 5. .".X« 1.391 ti. 979 22 ; A. M. 5. .")()5 1.2,V.l 6.701 1'. M. 5. 5 10 1.22!l 6. 769 21 1'. M. 5. .")9."> 1.375 6. 970 2,"> A. M. 5.721 1.221 6.912 25 1'. M. 5. 792 1. 21 W 7.000 28 P.M. 5.721 1 . 271) 7.000 30 V. M. 5. 7!I7 1.239 7.036 Sept. 1 2 P. M. P. M. 5. 725 5. 74X 1.259 1.2113 6.984 6. 951 4 P. M. 5. 720 1.2 IS 6. 968 i 5 P.M. 5.515 1.134 6. 649 7 P. M. 5. 7;i!l 1.275 7.014 8 A.M. 5. t)49 1.203 6. 8.52 P.M. 5. 595 1. 139 6. 734 9 P. M. 5. 5K5 1.077 6. 662 14 A.M. 5.5S-1 1.208 6. 792 15 A.M. 5. .500 1.181 6. 74 1 20 P. M. 5. 892 1 1.2H1 7.173 23 A.M. 5.791 1 1.307 7.098 25 A. M. 5.932 ! 0. 803 6. 735 28 P. M. 5.755 1.167 6.922 29 A. M. 5.615 I.OIJ 6.628 I POKT AUSTIN AND MILWAL'KEK STATrONH. 627 Computation of liiiijlil of ijui/e :eio at Milirmdrr, ll'idiiiiiHiii, abuie ijiigu :eri, nl I'ort AitHlin. Miiliii/aii, in the kiiiiiiiipi- ni' /,. 5t)« 1. 2.50 (!. Hlli 3 t A M. 5. 594 i.4a5 7. 0.5!l 1 P. M. 5.574 1.1«6 t'l. 7liO 1 P. M. r,. ()3L> 1. 101 (i. 73H 5 1'. M. 5. 705 1.0H5 (!. 7!tO 10 P. M. 5. 50(i 0. 8«» ti. 3! 15 ir. A. M. 5. 781 0. 7t>9 • 6. 553 17 A M. 5. VA2 1, 144 7.0«fl IX A M. 5. 720 1, 3!W 7.11H 1!( A. >[. 5. SKi 1.215 7.0t!l ''li A. M. fi. 1m:! 1.212 7.;«ti :'"> A M. li. Hi!) 0. 800 (i. !t3!l 26 P. M. 5.! Kill 0. «58 11. SIS L'7 A. M. 5. ni« 0. 7.50 (i. i'A'iH P. M. 5. MOO 0. 722 (1.531 l''J A. II .M. 5. XBl 0. 724 1 (1. 5f>!S Jlci G.875 I ±. 019 iiaiks, and it i.s Milwaukee is well provided with enjjineer beiicli probable that thorough research would establish tiie coiiiii'i^tioii of tiie gage zeros at eaeh oiiocli with several of the bench marks; but after iuspeetiou of the data readily accessible, I thought it best to make u.se of only one bench, that called the '■ check jjoint." This consi.-^ts of the toi» of a copi)er bolt leaded into the north side of the center ))ier of the swing bridge over the river between Chestnut and Division streets. The gage observer is refjuired at stated intervals to check the stability of the zero of liis gage by means of this check point. Using two rods, with the aid of an assistant he makes a series of simultaneous meas- urements from the check jjoint and from the gage zero down to the water level, and from these measurements the relation of the gage zero to the check point is determined. Their relation has also been determined by means of the engineer's level at various times, and was so determined on August 8, 1876, by Assistant Engineer L. L. Wheeler, who found the check iioint 0.843 foot above the gage zero, lu 1890 the check observations by the observer were very thorough, series of twenty simultaneous readings being made every fortnight, m 628 EARTH MOVEMENT IN THE GREAT LAKES REGION. and from five of tbeso series tlie relation of the two points is computed as follows: Coinpiitdtion of hdiiht of Milwdtdre check poiiit aboi-r Milwaiihef -no of gage in the mtmmer of lSHi:. I-'i-i't. Julv 1-' iiiu'iiu (if twenty coiiipiirisiins bv siiriultaiicoiis readings) l.L'0;{ July-i; I I-^l^ Anj;iist 11 l--'^f A 11 K 1 1 » t L' X 1 ■ 2t« ye)iti'nilpiT 16 l--Ot) Mran 1-^1'^ i.0(l2 In response to a letter of in, and October, 187itling, and the assumption made was that it had been at nnirorm rate through the whole period. The corre, I obtained as the height of the gage zero above the clieck point in July ami August, 1870, 7.4()() feet. T'>e various data thus described are combined in the foUowiug table. • 1 ( PORT AUSTIN AND MILWAUKEE STATIONS. G29 Compiiiation of height of Milwaukee check point ahoie Port .tiialin chcfl, point in the mimmern »/' ISTH iind !S9C. 1870. Feet. 1.205 fi.S75 — .-|. 125 ! Feel. Milwaukee check point above Milwaukeo gage zero : 0. Hi;i Milwaukee gage zero above Port Austin gage zero ■ — 5. 210 Port Austin gage zero above Port Austin oheik point 7. 4tiO Sum of above = Milwaukee check jioint iibove Port Austin check point 3.(l'J:i Difference — i). 138 This result iudicate.s that the grouud at Milwaukee, as cDinpared to the ground at Port Austin, has subsided 0.138 foot in the twenty years from 1876 to 1890. It is the algebraic sum of six measurements, of which three are levelings by water surface and three by the engineer's level. The probable errors of the water-level measurements are 1; 0.010, ± 0.013, and ±0.002. The probable errors of the Port Austin levelings in 1890, as indicated by the discordance of two independent results, is 1:0.008. If the probable error of each of the other measurements was ±0.(I10, the probfible error of the result is less than J .03 foot. There is also an uncertainty arising from the possibility that the stone pier to which the Milwaukee check mark is attached has settled, another uncer- tainty dne to the possibility of river floods, and a third involvt' B. Oliver. Tlie following tables give tbe coinput.ations for the two year.s : Computation of height of (jaije zero nl Miliidiikee, ll'iscoiiain, dboie gaije ziro at J.'aviinabu, Mivhiijan, in the isiimmir of IS^H. Date. liead ngs nt EKCnnaba. . Iteadiiigs at Milwaukee. 7 A. m. 2 p. m. 9 Jl. III. 7 a. 111. t 1 p. ni. p. III. 1876. Feel. Fett. Feet. Feel. Fett. Feet. Feet. Feet. j Feet. July 11 1.86 1.90 1.78 2.17 2.23 2.12 0.31 0.33 ' 0.34 12 l.ilO 1.75 1.78 ■ 2.12 2.03 2.16 .22 .28 ..S8 13 2.00 2.07 1.95 2.20 2.05 2.35 .20 —.02 .40 14 2.10 2. 15 2. 05 1.95 2.12 2.10 —.15 —.03 .05 15 1.!I6 1.95 2.00 2.16 2.15 2.06 .20 .20 .06 16 1.90 1.95 1.85 2.13 2.12 2.11 .23 .17 .26 17 1.89 1.80 1.85 2. 15 2. 20 2.21 .26 .40 .36 18 2. 0,5 1.85 1.95 2.20 2.07 2.20 .15 .22 .25 1!1 Aug. 16 1. 7.5 2.07 1.94 1.90 2.18 2.29 2.30 2.20 .43 .22 .36 .30 17 2.00 1.95 1. 95 2.19 2.13 2.21 1 .19 .18 .26 18 1.70 1.9.S 2.00 2.23 2.02 , 2.27 .53 .09 .27 19 1.78 2.00 2.13 2.22 .35 .22 20 1.95 1.90 1.85 2.23 2. 18 1 2.23 .28 .28 .38 1 21 1.85 2.05 2.10 2.25 2.20 2.24 .40 .15 .14 22 1.83 2.05 1.90 1 2.32 2.18 2.33 .49 .13 .43 23 1.91 1 1.83 1.75 2.20 1.91 ' 2.10 .35 .08 .35 24 Men 1.95 1 n 1.85 1 1.90 2.25 2. 10 2. 42 ! .30 .25 1 .52 0. 2,55 ■■■"1 i.012 «l ( OltBKKT.] ESCANABA AND MILWAUKEE STATIONS. 631 Computation of height of gage zero at EKcanaha, Michigan, above gage zero at Mihiaulcee, Wisconsin, in the summer of 1896. Date. Time. ' 1 KcniliD^^s (means of dericH). DilTcreiiie. Milwankeo. Gscaiinba. 1896. 1 Feel. Feet. Feel. ■July 2 P.M. 5.465 5.917 0. 452 7 A.M. 5.434 .-).907 .473 8 A.M. 5.505 5.920 .415 9 A.M. 5.348 5.837 .489 P.M. 5.356 5.7ft5 .409 10 A.M. 5.442 5.771 . 329 P.M. 5.567 5. 694 .127 11 P.M. 5.576 5.771 .195 13 A.M. 5.411 5.869 .458 P.M. .5.493 5.776 .283 14 A.M. 5.574 5.750 .176 17 A.M. 5.431 5.865 .434 19 A.M. 5. 524 6.007 .483 P.M. 5.496 5.887 . 391 20 A.M. 5.528 5.803 .275 21 A. M. 5.573 5.973 .400 j 23 A.M. 5.645 5.908 .263 25 A.M. 5.601 5.856 .2,55 28 A.M. 5.703 5. 857 .154 31 A.M. 5.446 5.938 .492 Ang. 1 P.M. 5.360 5.859 .499 4 A.M. 5.654 5.912 .258 8 P.M. 5.317 5.954 .607 9 A.M. 5.519 5.658 . 139 1 10 P.M. 6. 328 5.546 .218 13 A.M. 5. 273 '>. 616 . 34!) P.M. 5. 378 5. 752 .371 14 A.M. 5. 338 5.670 . 332 15 A.M. 5. 360 5. 730 .370 P.M. 5. 402 5. 710 .308 19 A.M. 5.414 5.878 .464 21 A.M. 5.558 5.935 .377 P.M. 5. 588 5. 872 .281 22 A.M. 5.505 5.698 .183 Sept. 4 A.M. 5.734 6.028 .294 13 P.M. 5.452 5.848 .396 1 14 A.M. 5.584 6.762 .178 16 A.M. ,->.500 5.937 .437 il « 632 EAKTH MOVKMENT IN THE GREAT LAk'ES REGION. Conumtation of height of gage zero at Escanaba, Michigan, above gage sero at Milwaukee, Wisconsin, in the summer of ISnC— Continued. ])ate. Tinio. Headings (moans of «orit'B). Difference. Milwfiiikue. Ksi-anabn. 1890. Feet. Fett. Feet. Sept. 18 A.M. 5. 701 6.047 .346 L'6 A. M. ,-).itU 6.187 .273 28 1'. M. 5. I'm 6. 224 .469 20 P. M. 5. 510 0. 133 .623 Oct. 7 A. M. -...-.M 6.164 .6150 M A.M. .-.. 731 6. 270 .539 14 P.M. 5.813 fi.157 .344 17 P. M. r<. fi22 C. 160 . .->38 lit A.M. r,. 84r) 6. 287 .441 1 20 A. M. .->. 8.^7 ;i!i ..S.-.7 ' I'. M. (i. MX 6. .-)40 .392 i 26 A. .M. (!. 080 (!. 471 . 391 P. M. 5. i)80 6.11 14 ..")6l Menn. 0. 374 ;i-.012 Tlie bench employed at Milwaukee lias already been described. At Escanaba there were three bench marks in good standing', as follows: No. 1, the top of the water sill on the southeast corner of the Adler building, northwest corner of Ludingtou street and Drusemau avenue; No. '2, the top of the water sill of the Escanaba lighthouse at the north side of front dooi-, against the brick wall; No. .'i is described in 1876 as the "center of a copper bolt set horizontally in the foundation of the light-house, west side, north corner, A feet north from steps." In a description by Oapt. (ieorge A. Zinn, dated -lune liO, ISOO, tlie top of the bolt is specified. I am informed by Mr. ( 'liiiton H. Oliver, the gage observer, that the diameter of the bolt is thiee eighths inch. The rela- tive heights of two or more of these bench marks have been determined in at least six difterent years, the inoasurements being made independ- ently with the engineer's level. It is advantageous to compare these nieisurements, not only to learn what confidence is to be reposed in the individual benches, but for the sake of whatever light ipay be cast on the general precision of such data. ■ ( ESCANAHA AND MILWAUKEE S^TATIONS. Comparison of Encanaba lieiich marks irith one anothir. 633 Tear. Above zero of gage. Xo. 3. Ditlon'iice bi'twvi: Ih'iii-1i inarkH. Feet. ! Feet. Feet. 2.100 a(i.23.-» Feel. Deviation r'-<.477 1 +0.0(« .">. 405 fi.482 t 1.017 — .0011 +0.001' +0.(105 5.476 a(!.K13 ial. 367 + .002 ,").46« 0.473 ! 1.005 — .006 — .011 — .0117 5.483 0.407 1.014 + .OOil + 013 + .(1(12 .Meiin 1 5.474 6.484 j 1.012 aNiit uHed ii) ('(iiiiputii)}; iiieanH- In this table the le.idiiiff of the height of bench Jiiavk No. .'5 in 187(» is corrected for tlie distance between centei' and top f Itolt. In the first division of tlie table the benches are reterred to zero of gage, but as the gage was not ;o zito Escauaba (taKf zero j'' )Vc Milwaukee gai;e zero Milwaukee gaj,'e iser l-ore Milwaukee cheek poiut . . .•^uiu of aliove = E9canal)a bench Xo. 1 above Milwaa- kee check point "• T"" Difference + •1''3 I). 949 The result iudicptes that the ground at Escaiiaba, as compared with the ground at Milwaukee, has risen 0.173 ' .)t in twenty years. This quantity is tlie algebraic sum of six measurements, of which three were made through water leveling and three by instrumental leveling. The probable errors of the water levelings are ±0.012, iO.Oll', and ±0.002 foot; the estimated probable error of the instrumental levelings at Milwaukee is ±0.010 foot, and of the two levelings at Escauabaeach ± O.OOS foot. This gives as the probable error of the result ± 0.022 foot. A similar computation, using bench mark No. 2 instead of bench mark No. 1, gives 0.155 foot instead of 0,173, and a compution based on bench mark No. 3 gives 0.150. The mean of tlie three results is 0.101 foot, with a probable error of ±0.022 foot. The only important uncer- tainties to which this result is subject, besides those indicated by the discordance of meiisurements, arise from the possibility of the settling of the bridge pier to which the Jlilwaukee check point is attached and the possibility of ri\ er floods. mSCBEl'ANCV MOTED BY CAPTAIN MARSHALL. In the later work of the United States Lake Survey all determinations of lake level were referred to the high-water level of 1838, which is called the " plane of reference.'' That plane was directly observed by Dr. I. A, Lapliam, the geologist, and with the aid of a bench mark on his house at Milwaukee was permanently recorded. For other stations on Lake Michigiin-Huron its position was determined by assuming that the level of 1838 had everywhere the same height above the mean lake level as determined by long series of observations. For the deter- mination of this plane at Esc<;naba use was made of observations for the period from January 1, 18')0, to December 31, 1875. In 1887 Capt. W. L. Marshall, U. S. E,, under whose direction the gage readings at Milwaukee and Escanaba were then nmde, detected a discrepiincy, which he reported to the Chief of Engineers in a letter dated October 1.' I Ami. liipt. Chief of KnglnoerH, U. 8. A., for 1H87. part 3, p. 2417. _ ■ I i 1 v^ SUMMARY OP KESULTK. 635 In former reporta the zero of Kscanalia i^iijiB has been assuiiiod as O.Tti font above the piano of reference, Imt a comparison of correctcil readings at MihvaukceandKHCMualia shows that th<^ deterniinatious of tlie plane of reference at Milwaukee and Kscanaba vary 0.187 foot, the Kseauaba plane bcintj too high or the Milwaukee determination too low. In the light of present knowledge it seems i>robiible that the dis- crepancy thus noted by Captain Marshall as au error was occasioned either wholly or in chief part by the progressive tilting of the land. This conclusion isdirticult of verification, because little record survives of such checks as may have been made upon the heights of gage zeros during the i)i'riod 1800-1875; but the indicated change agrees in direc- tion, and approximately in rate, with the change deduced from the present investigation. From the middle of the period 1800-1875 to the summer of 1887 was an interval of twenty years, equal to the interval 1870-1890 here used, and the discrepancy of 0.187 foot discovered by Captain Marshall differs irom the change of 0.101 foot here deduced by a quantity little greater than the probable error ascribed to the latter determination. SUJOIAUY OF RESULTS. In the following table are assembled the numerical results as to changes in relative height of the four pairs of stations. Besides the measured changes, the table includes the periods intervening between dates of measurement and distances between the stations of each pair. The lines connecting pairs of stations have a southwesterly direction (fig. !>!•, p. 013), and it is the northeastern station of each pair that appears to have risen as compared to the other. The results thus show a general agreement with the working hypoth- esis, that the latest change recorded by geologic data is still in progress. To make the comparison quantitative there should be substituted for the direct distances between stations the corresponding distances in the assumed direction of tilting, S. 27° \V., and the measured results for various distances and various time intervals should be reduced to a common basis. Ju the third column of the table are the reduced distances, and in the sixth the reduced r ites of change. Assuming the change to have a uniform rate and to be the same for all parts of the region, the measurements at the dflerent pairs of stations give for a distance of 100 miles and a period of a century the iwantities of the sixth column. The seventh column coni^ins the probable errors of quantities in the sixth, and is based on the jirobable errors of the measured changes iu pairs of stations. .A\ I , ■fl-c G30 KAIiTII MOVEMENT IN THE GREAT LAKES REGION. Sumtnari/ <)/ ilinl,iiiaH, time iiilcrritU, ami mvamirdninlH of tUffircnlial enrlli mortmciilt. ,, , '^''-'-o'lr;:™: Change „,,..,Hure. ,„„^„, iliHtani'f. tjon I Sacki'tts Harbor iiiiil I t:harlotte Port Coll)ornc and ! Cleveland I'ort Austin and >Iil- wankee I liscanalia and Mil- llilfi. I mU: I Years. | Feet. •J2 88 76 ir.s j 111 259 17(i Feel. Feet. 0.061 0.37 j 0.18 0.2.30 0.46 0.11 .'0 0.137 I 0.39 0.09 .'0 0.161 0. r! 0.06 wankce 192 180 Mean 0. II Weiglited mean 0.42 , ±0.044 I.S TlIK LAXI) tiltixg:' With the iiuiuericiil results of the investigation before »s we may now recur to the main subject and ask whether the evidence warrants the condusion that a general, gradual tilting of the basin is in prog- ress. In the discussion of the data used in coini)aring the several pairs of stations it has been found that, taken at their face value, tlioy indicate a tilting in the hypothetic direction, but it has also been fouiul impossible to resolve all doubts as to the stability of the gages and benches and the accuracy of the measurements. By reason of these doubts the result from no single pair of stations is conclusive, but when assembled they exhibit a harmony which argues strongly for their validity. As tabulated, there are four results, but these are not all independent, since ol)servations and measurements at Milwaukee are used twice. There are, however, three results wholly indeiiendent and a fourth partly independent. To these may be added a fifth nartly indei)eiHlent, namely, the determination of change between I^ort Col- borne and Cleveland for the shorter period, 1872-1895. Not only do all these results indicate a change of the same sort, but they agree fairly well as to quautity. The computed change for 100 miles in a century ranges only from 0.37 to 0.46 foot, and the greatest deviation of an individual result from the mean of four is 12 i)er cent. This measure of harmony appeals strongly to the judgment, and is also susceptible of appro.xiinate numerical expression. If the four determinations tabu- lated in the sixth column are, in fact, measures of the same quan- tity — that is, if the tilting has been uniform throughout, as we have assumed — then the probable error of tlie determined value of that quantity (0.42 foot) is less than ±0.05 foot. VI i OILIIEIIT.I RATE OF MOVEMENT. 637 Tlio most important factors tending to throw doubt on the (•oiulnsion are the possibilities of accidental change in the various beuclics to which the measurements are referred. The bcncli at Port Austin, being a marlc on bed roclc, is trustworthy, and the agreement between the three benches used at Escanaba is good evidence of their stabil- ity; but the bench at Milwaukee, with which bntli are conijjared, is a pier of a bridge in daily use and may, perhaps, be slowly settling. If it is settling, the comparisons with benches at Kscanaba and Port Aus- tin may merely reveal that fact and not measure tlie subsidence of the land. The fact that the swing bridgoon the jiitr has not required re lev- eling is certainly favorabh* to the stability of the pier, especially when it is considered that a change of fully lA inches is to bo accounted for; and there is further conliriuation in the discovery of a discrepancy between ^lilwaukee and Escanaba by Captain ^Marshall, whose data are probably independent of the check mark. Of the benches on Lake Erie, the one at l*ort Colborne is satisfactory, but those at Cleve- land may have settled at critical times, and if so their change would intiuence the result in the direction found. Of the benches on I.ako Ontario, the one at Charlotte is eminently stable; the oidy practical question atTccts the bench at Sacketts Harbor, which is on a building that has not been wiiolly stable since its construction, although pre- sumably s(t since the making of the bench. If the buikPr, at Sacketts Harbor scttleil between 1874 and ISitG, the etfect cf the lowered bench was to i)ro(luce, not sucli a change as appears from the measurements, but one with the opposite sign. It seems to me that the harmony of the measurements and tlieir agreement with inedictioii from geologic data make so strong a case for the hypothesis of tilting that it should be accepted as a fact, despite the doubts concerning the stability of the gages. 1 il KATK itV MOVKMKXT. The deduced mean rate of change — 0.-12 foot to the KM* miles in a century — depends on assumptions which are convenient rather than probable. These are: (1) 'hat the whole region moves together as a unit, being tilted without internal warping, and (2) that the direction of its present tilting is identical with the direction of the total change since the epoch of the Nipissing outlet of the upjier lakes. What wo know of the general character of earth movement?^ gives no warrant for such assunqitions of uiiilbrnfity, but no better assumptions as to this region are now available. I'liler the law of probabilities, the close agreement of four measurements, three of which are wholly inde- pemlent, gives a good status to their mean, but there are other con- siderations tending to weaken this status. The probable errors of the individual measurements are rather high, ranging from It to .")» per cent, and this suggests the possibility that the closeness of their cor- respondence nniybe accidental. It should bo remembered also that at Ml i ^^*' 638 EARTH MOVKMENT IN THE GREAT LAKES REGION, two or throe stations there was reason to beliftve that the gage zeros were settling during tiie period in wliich tlie observations were made, and till- residts involve the doubtl'ul assumption that the rate of set- tling was uniform. There is room for doubt as to the precision of the instrumental leveling; in only a few instances is the fact of dupiifute measure nents recorded, and single measurements are notoriously insecure. Va-vov was doubtless admitted by ignoring the eft'ects of bar etrie gradient. Hiver floods may have introduced errors. In the absence of Hood records the records of rainfall at Rochester (near Charlotte), Clevelaml, and Milwaukee were compared with the gage readings, the results showing only that if fiood errors are involved they must be small. There may also be per.soinil eiinations of observ- ers, especially as the gages at pairs of stations were not in every case of the same type, i'^or all tb' se reasons I am disposed to ascribe only a low onler of precision to tin .educed rate of change, atid regard it as indicating the order of magr.tude rather than the actual magnitude of the diti'erential movement. The rate of change indicattil by Stuntz's observations is more rapid. As already quoted, he stares that at a time when Lake Supe- rior was exceptionally low at its outlet, it was nevertheless so high at its western extreu)ity as to obliterate from the St. Louis River a rapid which had been visible oidy a few years before. This statement involves no definite mejisures, but it implies that the change within the memory of individuals involves feet rather than the inches deduced from the studies in the other lakes. Similar inferences may be drawn from his statement as to submerged stumps. The recorded range of water level in Lake Superior is about n feet, and trees would grow little if any below high-watermark. If, then, with low stage at the east end, stumps are submerged at the west, a change of ."» feet or more would seem to have occurred during the period covered by the growth of a tree and the survival of its stump. Tlie differences between the inferences drawn from this evidence and the result based on gage readings on the other lakes is so wide as to suggest the possibility of error in the Lake Superior observations. It is certainly important that they be verified. Unfortunately I have not been able to visit the region, and the gage records accessible to me are not so connected with bench marks as to give a satisfactory basis for computation. The United States Lake Survey nmde observations of lake level at Superior City from 185!) to*187], and then trsMisferred the station to Duluth, where it was continued for two or three years. No bench mark at Duluth is described, and the only recorded bench mark at Superior City is ui)on a wooden structure, Johnson & Alexander's sawmill. If this bench survives, a good test could be made by renewing tiie gage station at Superior City. At the other end of the lake, at Sault Ste. Marie, there are authentic benches dating from 1855. If we assume that the rate of 0.42 foot per 100 miles per century is aiLniBT.] RESULTING OEOOKAPHIC CHANGES. 63!) uiiiforin and secular, and project it backward to tlie time when tlie drainage of Lake ilnron was sbifted from North l>ay to Port Huron, we obtain for tlie period since that cliange about 1(\<)()() years. Fronj studies at Niagara, Taylor has estinuited the same p. "'od as between 5,000 and 10,000 years;' and the comparison indi(;ates that the rate of nio. (ItO). Atiramilton and Port l)a"iousie it amounts to inches in a century. The water also advances on all shores of Lake ICrie, most rapidly at Toledo and Sandusky, where the change is 8 or inches a century. All about Lake Huron the water is falling, nn)st rapidly at the north and nofth- east, where the distance from tiie I'ort Huron isobase (CO, fig. 100) is greatest; at IMackinac the rate is (» inches, and at the mouth of I'rench Kiver 10 inches, a century. On Lake Superior tlio isobase of the outlet (1)1), fig. 100) cuts the shore at the international boundary; the water is advancing on the Amev !an shore and sinking on the Cana- dian. At Duluth the advance is inches, and at Heron I>ay the reces- sion is 5 inches, a century. The shores of Lake Michigan are divided by the Port Huron isobase. North of Oconto and -Manistee the water is falling; south of those places it is rising, the rate at ililwaukee being 5 or (i inches a century, and at Chicago !) or 10 inches. Event- ually, unless a dam is erected to prevent. Lake Jlichigan will again overflow to the Illinois Kiver, its discharge occupying the channel 1 ll I Dull. GboI. Soc. America, Vol. IX, IsgS, p. 83. G40 KABTir MOVKMKNT IN THK (iRKAT LAKKS RKOION. ciuvcd l.ytlie outlet ofii ridstoconc, Klaciiil laki'. Tlio smnmit in that cliaiiiiil is now « I'eet above the mean level of the lake, and the time before it will bo overtopped (under the stated assumption as to rate of tiltiiifr) may be computed. Evineutly the flr.st water to overflow will bo tliat of some high stage of the lake, and the disehargo may at (Irst be intermittent. Sueh high-water discharge will occur in 500 or «00 years. For the mean lake stage such discharge will begin in about 1,000 years, and after 1,500 years theie will be no iuterruptiou. In about 2,000 years the Illinois Eiver and the Niagara will carry e(iual jiortions of the siu'i)lus water of the Great Lakes. In 2,500 years the (lisihaige of f lie Niagara will bo intermittent, failing at low stages of the lake, and in .i.-^OO years tiiere will be no Niagara. The basin of lj;| OiiUiM^ ^tmAtLsky Fi(f. 100.— KelntioiiH of tlio Bliori's of tlio Great LiikOB to the i^obascB drawn through their outlets. Lake Erie will then be tributary to Lake Huron, the current being reversed in the Detroit and St. Clair channels. The most numerous economic bearings of tliis geographic change pertain toengineering works, especially for the i)reservation of harbors and regulatiim of water levels. But the modilications thus produced are so slow as compared to the growing denumds of connnerce for depth of water that they may liave small imi)ortsesot'scien<'(' tlieordcrofniagnilndo of tlni ciianye is more iniiioitant than its precise nn'asnrenuMit. hut there are involved nr^at economic interests, and tlie.-e demand irioro delinite inl'ornnition. The account of the present in vestitcation is there- lore snpphMnciited hy an ontiine ])lan of the more elaborate iiivestij^a- tion which appears necessary to give nu'asiu'ements of tiie inecis'on that is desirable. Mxistin;;' tlata are ntithei' fall enongli nor exact enon{,'h to ^ive satis- factory measures ot' tiie small ([nantities son};ht. Doubtless a more elabcnate disenssion would yield better results than I have obtained, but the improvenu'nt could not be y;reat. Observations by the I.ako Survey were eoinhu'ted for purposes notdemandin^ahiK'h order of pre- cision, and hij,'h relincment was not attempted. The supplementary v.ork done in IH'.HI attemi)ted only to bo good eiioujili for use in comlii- natiini with the work of 1S74 and ISTti, and can not serve as the lirst term of a lU'W comparison. The problem rerjuires a new set of hi;;h- fiiad'. observations at each station of a carefully jilaiined system, to bo f( Mowed, after an interval of at least adecade, by asecond set of observa- tions at the same stations. Foreseeing no oi»i)()rtunity to undertake such a work niysclt. J havo formulated in the following jiaragraphs a [)lan embodying the results of my experience — iiplan iuteinled to allbrd useful suggestions to some investigator by whom the work nmy bo actn.ally undertaken. Silccfidii of HfatioHx. — To measure the rate of change in any given direction, observations at two stations siitlice; but todetermino also the direction of change, it is necessary to use three stations grouped in the form of a triangle. The longer the sides of the triangle the better the ineasurenn'ut of rate, ami the larger its smallest angle the better the determination of direi;tion. A brief inspet'tion shows that tho shores of Lake Michigan and Lake Huron give the best opportunity for the planning of ii well-conditioned triangle. Though the narrow- ness of their connecting strait has led to the giving of separate names, they are really a single lake, and the stretch of their water surface is in every direction greater than that of Lake Superior. For the ])urpose in view the point of first importance is the outlet of the lake at I'ort lluron. This is peculiar in that the plane of mean water level has here a C(mstant relation to the adjacent land, a relation altogether independent of the progressive deformation of the basin. This station should not be on the St. Clair Eiver, but on the shore of the lake near by. 18 GEOL, I'T L' 41 Mlii: f i! G42 EAUTH MOVE^rKNT IX THE fiREAT 1,AKES REGION. Tlie second point of vantage is < 'hicaso. As economic interests are more sciionslv allectod l).v tlio ocfrniphic cliaiifre at that point than elsewlii-if, it "is desirable to determine diroetly, l)y comparison with Port Huron, the rate at wliidi ilie lake is encroacliinjj- on the land. A tliird point of prime importance is the Strait of Mackinac. Altiiongb the e(piilibrinm levels of the surfaces of the two lakes are the sanre. there are considerable periods when their equilibrium ia dis- turbed, and during such jieriods a current Hows in one dirc<'tion or the other through the strait. Only when this current is nil is the whole water body in perfect oiiuilibrinm, and it is essential to p cise levelinjf through t iie water surfa.e eitiicr that times of e(iuilibriniu be chosen or Vlii. 101.— I'mpuned s_\ rtleins of staliniis lor (lio jiriTiso nii'ai^lirtMiii'iit <<(' i-jirlii riM»\iini-ntH. that due allowance be made for the ;fradients associated with tlow. Observations must therefore be nnide on the current in the strait, and it is best to connect them witli the work of a complete station. As a|i])earsby the annexed diaj-ram (fi};. 101). the trian};le formed by these tliree stations is well conditioned as to size and form ; the len<;ths of its sides are appioximately 225, 275, and JHO miles, and its smallest anf;le is about lii degrees. While the j)roper use of these three stations will give answer to tlio questions of greatest ec(niomic and scientilii; importance, there will be nmterial scientillc advantage in adding a fourth station to the system. It should be))laced somewhere on the north shore of (ieorgian Hay, and, 01I.nF.IlT. 1 PLANS FOK PRECISE MKASURK>I?:NT. 643 ^, J •*' . giving consideration to accessiliility as well as };i'<'};'''>iplii"' ])osition, it is |)robable. that I'arry Sound slioiild be selected, liy addinj;' this sta- tion another well-conditioned triangle woidd be (uimpleted, and there would result an additional determination of tlie rate and direction of tilting. If rate and diiection vary from iilace lo jilace the fact will ]irobal)ly be l)rought out. There would be additional advantage in the fact tiiat I'arry Sound and Chicago are separated by the greatest \n'iw- ticalde distance in the direction of ma.xinnim change, so that a coiii- ]taratively short ])erioil of time might afl'ord a valuable measiireiiient. The approximale results of tlie present investigation indicate tliat the change in tlie relative leiglit of Parry Sound and Chicago in ten years would be about 2 inchc! . Conilitioiis ciiiitrolliiiii ciinipmiiit. — In order to jilan intelligently the system of observations, full consideration should be given to the con- ditions arte<'ting the iiroblem. and pro\'ision should lie made for all jtossible sources of error. Prominent among tliese are the various fac- tors which modil'y the water level at points on the lake shore. Sin'h factm's have been considered in the jireceding discussion of gage data, but they are assembled lieii^ in a more systematic way. Tlie lake continually receives water from streams and from rain, and continually parts with water by discharge at its outlet and by eva!>ora tion. In the long run gain and loss are eipial, but for short periods tliey are usually iineciual: so that Irom ibiy to day. from season to season, and from year to year the volume of the lake and the cnnse- (pient mean level of its snrtace are <'ontinuall\' changing. lu bays and estuaries there are local temporary \ ariations occasioned by the Hoods of tributary sti'eams. There are solar and lunar tides, snnill as compared to those of the ocean, but not so small that they may be neglected. The wind pushes the lake water before it, jiiling it u]) on lee shores and lowering the level on weather sliores. During great storms these changes have a magnitude of several feet, and the effect of light wind la distinctly appreciable. I'^ven the land and sea lireezes, set up near the shore by contrasts of surface temperature, have been found to pro- duce measurable ell'ects on the water level. There is also an iiilluence from atmosiilieric pressure. W'lien the air is in eipiilibrium, if that ever occurs, the iiressnre is the same on all parts of the lake surface, and the eiiuilibrium of the lake is not continuously recorresent paper are in hand. As will be readily understood from the lidlowing abstract, the data he has gathered constitute an important contribution to the subject. 'MiHlltli'utiiin cil' I hi' ilroiil Lakes liyiMirtli mciviniini i Nal. lioog. Mog., SBptonilii'r, IHiiT. \'ul. VTtl, lip. 2:::t-'J47. ■' Liikt' Krin iMiIrtruinjj: the i-*Irtiuls si'imnited rniiii tho nminluiiit In rcreiit tiini'^; \>y Iv I*. Mtiscleyj Tku Lukonide Mui;u/.iiii', Liikosiilo, <>liii>. April, IK9K, Vul. I, |i|), 14-lT. ■ I ( I i I'f 646 EARTH MOVKMENT IX THE (;',; -T LAKES RE(iION. North of Sandusky Buy. near the west end of Lake Erie, is a cluster of islands, of which the live largest are each several miles in extent.' About them tlie water is shallow, and if the lake were lowered 30 to 35 feet they would all be connected with the mainland. On these islands grow many species of wild plants, and the origin of this Hora is related to the geologic history of the islands. There was a time during the ice retreat v.lien the whole l)asin was covered by a glacial lake. If the water was gradually lowered fiom the plane of the glacial lake to the present plane of Lake I'hie, the islands were at first barren and were eventually occui)ied only by such jdants as were in .some way conveyed across the intervening straits, from I'.i to 3 miles wide. As ^loseley points out, theic arc many modes of such adventitious intro- duction, but they could not be expected to give to tlie islands a liora so varied as that of the adjacent mainland. If, on the oilier hand, as inferred IVom the slopes of the old shore lines and other data, the altitude of the land was dillerent when the glacial lake was drained away, the origimil Lake ICrie occinpied only the eastern part of the I'hie basin, and tiie western part, including the dis- tiictof the ishmds, was dry land. Subsequently, from the tilting of the land, the lake walers advanced westward so as to Hood the straits and convert the lowland hills into the present islands. !n connection with such a geologic history the islands would have accpiired their flora a'' the same time with the mainland, and siionld now present the same variety of s])ecies, so far as local conditions iiermit. iMoseley has careful ly compand the insular Hora with that of the nniinland, and finds that Ihe only nniinland si)ecies which do not occur on the islands are such as do not tind there a congenial soil, Tiie botanic, evidence thus supports the geologic, and verilies the conclusion that the land has been tilted toward the southwest since the birth of Lake Erie. The islands are conii)osed largely of limestone and are surrounded by limestone cliffs. In f-'outh Bass or Put-in Bay Island there are caves opening at the water's edge and \ tly (tccupied by lake water. Exploring these, Moseley finds stalactites exteinling from the roof down into the water, and stalagmites lying 3 or i feet below the present sur- face of the lake. Compiring the present water level with the lowest levels known in recent times, it appears that these stalagmites have not been above water during the present century, and as stalagmites are formed only in the air. it is clear that the lake has encroached on the land since they were nuide. These data show only that a change has occurred, and ascribe no date, but other phenomena observed in tin- neighborhood of Sandusky indicate clearly that change is now in progress. A tract of land on which hay was nuule in 1828 is now permanently under water. A tract of land onclmlf mile scpuire, surveyed in 18(){>, has since become marsh, with water and niiid 12 to 18 inches deep. Various parts of Sandusky ' Korlhc' reliitiiin ot'thuaii jhUiiiIh to tliv laku anil the iMubiimi iiI'Iih