Final Rise Above the Sea.-From the account given in a former chapter of the making of the bed rocks in this area it is seen that the area was sea bottom most of the time until sometime in the Silurian period. If it remained longer beneath the sea, or was again submerged later, the beds thus formed have since been eroded away and the record lost.
The Making of the Peneplain.-The history of the making of the present surface begins with the final uplift above the sea. It is not probable that the land in this region ever rose very high. Erosion reduced it again to a peneplain, that is to a nearly flat surface very near to sea level. Whether this occurred once or many times cannot now be told, for in the last complete cycle of erosion the records of all former cycles were necessarily obliterated. The last peneplain produced was at least as flat as the flattest uplands now surviving and probably not more than 200 or 300 feet above the sea. The streams of that time meandered with very little fall over very wide flood plains. The surface then, as now, was independent of rock structure, though since the rocks are very nearly horizontal the surface was very nearly parallel to the beds. Nevertheless it cuts across the beds at a very small angle. As seen on the geologic map this surface rests now on one formation, now on another. This slight discordance of surface and bedding is one of the indications that the present surface is a peneplain and not an original sea bottom which has been lifted up and not eroded.
Age of the Peneplain.-This peneplain was complete at some time not very long bef ore the glacial epoch, that is, in the Tertiary period. If it was complete earlier than that it continued unchanged into the Tertiary for want of any uplift, which would give the streams renewed cutting power, and thus bring about the destruction of the peneplain.
Uplift and Erosion.-Later, this peneplain was uplifted. This must have been late in Tertiary time, because the amount of erosion following the uplift was not great as compared with the entire task of base leveling. In the new cycle of erosion, maturity was reached only near the larger streams. Back a few miles from them the uplifted peneplain was preserved in considerable patches. Despite the later events of physiograpbic history these remnants still survive at altitudes about 900 feet above the sea being 250 to 300 feet above the Miami near Hamilton, and 350 to 450 feet above the Ohio in the southern part of the area. Remnants of the old peneplain rising to a common level are sufficiently numerous so that in any extended view from one of these uplands the horizon is flat.
Miami River.-Three streams of considerable importance united their waters in this area (Fig. 48). Probably the largest was the preglacial Miami. The name must be used with reserve, since the outlines of the drainage basin and even the course of its northern portion were different then from now. But for most of the distance from Dayton to Venice, Ohio, the pre-glacial stream followed nearly the course of the present Miami.
East Fork of Little Miami River. -Another stream entered the area from the east along the line of the East Fork of the Little Miami and flowed west past Madisonville and Norwood. Judged by the breadth of its valley it was smaller than the Miami, but quite as large if not larger than the stream which entered the area along the line of the present Ohio.
Licking River.-From the south came the ancestral Licking, essentially in its present valley, with this exception that it continued northward through the Cincinnati basin and the narrow part of Mill Creek Valley to near Elmwood Place. There it joined the stream from the east, the ancestral Ohio described below. It thus appears that the broad and now very important Cincinnati basin was not made by the great river which now traverses it, but by the relatively small Licking, which, both south and north of this basin, flowed in a much narrower trench, having an almost constant width of one-half mile. It is quite probable that this and several other local expansions of the old river troughs are due to certain structural weaknesses of the rocks which made erosion easy.
Ohio River.-The stream which entered the area along the line of the present Ohio probably headed near Manchester, Ohio, not more than seventy miles above Cincinnati. If this stream be called the pre-glacial or ancestral Ohio it must be understood that it is so called purely on account of its course in this vicinity, and with the understanding that there was no Ohio River then in the sense of a stream having the same drainage basin as the present Ohio. This stream. turned northeast at California through the valley now occupied by the Little Miami. South of Madisonville it was joined by East Fork and together these waters flowed west through the Norwood trough to what is now Mill Creek Valley.
After this stream from the east had united near Elmwood Place with the Licking from the south, the combined stream flowed north through what is now Mill Creek Valley and, turning west in the same valley, joined the Miami about three miles south of Hamilton. Passing southwest from there past Venice, the master stream flowed through the broad valley described in Chapter I as the New Haven trough. This valley is the natural continuation of the Miami, not only in direction but in dimensions and in the style of its bluffs, whereas the narrow trenches past New Baltimore and Miamitown agree in none of these. Near Harrison this master stream turned and flowed south through what is now the Whitewater Valley, a valley which plainly was made by a larger stream than the one which now occupies it.
Todd's Fork.-Still another large stream crossed the northeast comer of the area joining the Miami at or near Middletown. This stream made and occupied the large valley described in Chapter I as the Union Village trough. When traced headward this trough is seen to be continuous with the valley of the Little Miami from Morrow to South Lebanon, and, east of Morrow, with the valley of Todd's -Fork. In pre-glacial time a continuous stream flowed along this line from the valley of Todd's Fork past Morrow, South Lebanon, Camp Hageman, and Union Village, joining the Miami at Middletown.
Tributary Streams.-Most of the principal tributary streams occupied the same valleys then as now. This is true of Sevenmile, Fourmile, and Indian creeks. All these streams, small and large, flowed at levels from 50 to 150 feet below their present channels, as indicated by the depth of the bed rock surface below the glacial drift and alluvium.
A number of tributary streams, very important in the history of the drainage, but no longer in existence, are shown on the map (Fig. 48). One of these flowed north from Foster at the eastern edge of the map. It followed the line of the present Little Miami (reversed) and discharged into the preglacial Todd's Fork near King's Mills. Another flowed south from Foster to Milford along the present course of the Little Miami, but it was then a small tributary of the East Fork. Near Foster, perhaps just south of it, the north-flowing and the south-flowing streams were in headwater opposition and had cut a notch or col in the upland.
Another pair of tributaries, which later became of great importance, flowed east and west from near Dayton, Ky., along the line of the present Ohio; the one being tributary to the former Ohio, the other to the Licking. There was also a sag or col or notch in the upland between their heads.
In similar manner another pair of small streams flowed north and south from the col one and one-half miles south of Miamitown. The south-flowing stream followed the course of the present Miami and jcined the master stream near the present mouth of the Whitewater. The north-flowing stream flowed eight or ten miles to join the master stream between Shandon and Fernald. It probably received a small eastern tributary from near New Baltimore, and it is certain that another small stream flowed east and north from New Baltimore along the line of the Miami (reversed) to join the master stream.
The Dry Fork of the Whitewater, which now flows south on the western edge of the area past Shaker Village and New Haven, formerly turned east to Shandon through the valley now followed by the C. & 0. Railroad, and then south through what is now the valley of Paddy's Run to the master stream.
A small stream heading near Mount Hope Church, two and one-half miles north of Miamitown, flowed northwest joining the master stream west of New Haven.
Another pair of tributaries, which later came into great importance, flowed east and west from a divide just east of Andersons Ferry. The east-flowing one joined Licking River near Ludlow, Ky. The westflowing stream joined the master stream west of Cleves, following the present course of the Ohio only so far as North Bend, there turning north and then west in a course not very different from that of the Big Four Railroad. Thus it joined the main stream near Valley Junction on the north side of the isolated upland. Where the Ohio now skirts the southeast edge of this upland there was then a minor divide where two small streams headed, one flowing northeast to join the stream just described at North Bend, the other flowing--southwest to join the main stream near Lawrenceburg. The stream which flowed from Andersons Ferry past North Bend received, west of Cleves, the northern tributary described above as heading at the divide south of Miamitown.
The evidences by which it is known that the streams were as here described, naturally divide themselves into two classes, those which go to show that the large streams did not have their present courses and those which go to show that they did have the courses described. Speaking first of the former:
Narrows at Andersons Ferry.-A group of features showing that the Ohio has not followed its present course very long is found at or near Andersons Ferry. The bottom of the narrow trench at that place is scarcely wider than the river in high water (see Plate. I-A)- The bluffs here are exceptionally steep and, as pointed out in Chapter I, the south bluff for three miles east of the Ferry is almost free from ravines or even gullies. These features indicate a very young gorge. The trench of the Ohio elsewhere has at least from two to four times the width of the river. Even this is not old, but the actual time required to cut a trench twice the width of the river is much more than twice, perhaps many times the time required to cut a gorge as wide as the river. A large stream may cut a gorge of its own width in a relatively short time; this is all done by down-cutting, but the widening of a trench is by very different processes which do not work so rapidly.
It is a general principle that tributaries soon become numerous and the upland soon dissected in the immediate vicinity of a great down-cutting river, but the opposite is the case here. Immediately south of the river lies a broad upland sloping from the river, and almost the same feature is seen on the north side. The undissected character goes to show that until recently these uplands were remote from master streams. It is this exceptional circumstance which accounts for the commanding character of the sites occupied by the two convents on opposite bluffs, Villa Madonna on the south and St. Joseph Mothers' Home on the north.
If the character of this gorge be contrasted with that of the Miami trench or that of the Little Miami below Milford or with Mill Creek Valley (see Figs. 6.-12), it will be seen that, these latter are not only from three to six times as wide but that their bluffs are much eroded and there are no broad uplands near these valleys. The actual time necessary to cut these valleys is not three nor six, but many times that which would be necessary to cut the gorge at Andersons Ferry.
The evidences of youth are of the same class, though generally less pronounced, at all the other places which were described above as cols or divides from which two former small streams were said to have flowed in opposite directions along the line of the present large streams. The narrows between Dayton, Ky., and Walnut Hills, and the similar narrows below North Bend are less striking in character. So also is the one at New Baltimore on the Miami, but the one south of Miamitown almost rivals the gorge at Sedamsville.
Narrows of the Little Miami. - The evidence of this character derived from the Little Miami is very striking. The features of the gorge at Foster are essentially those described at Andersons Ferry, though of smaller dimensions. No observer of valleys could possibly fail to note the peculiarity of the Little Miami, flowing as it does from its spacious valley at South Lebanon into the gorge south of Kings Mills, and later into the broad flat south of Milford.
Direction of Tributaries. - Another class of evidence going to show that the large streams did not follow their present courses is found in the direction of some of their tributaries. Streams commonly join at an acute angle. It is exceptional or abnormal if the angle between the two streams above the jundion is obtuse. In that case the tributary must to a considerable degree turn back on its own course. Such an exception or abnormality is very striking between Andersons Ferry and Ludlow, Ky. Tributary streams on both sides of the Ohio enter at very obtuse angles and turn back on their own courses (Fig. 49), indicating that they were normally tributary to an east-flowing stream. West of Andersons Ferry all tributaries join normally.
Character of the Old Valleys.-Having now considered the features which indicate that the large streams did not have their present courses in pre-glacial time, it remains to point out those features which indicate that the streams were as shown in figure 48. Of the evidence that the former large streams did have the courses described above, the most striking consists in the great valleys themselves. This is especially convincing where the great valleys are not in use at present by any through-flowing stream, for example, the Norwood trough, the New Haven trough, the Mill Creek Valley west of Flockton and the Union Village trough. (See Fig. 4) That such great valleys should be directly connected with present river valleys, continuous as to direction, dimensions, and character, is strong evidence that they were made at the same time and in the same way. This is especially convincing where the stream in leaving the continuous valley turns into a narrow and evidently young trench.
Slope of the Bed Rock Surface.-In determining the direction of former drainage, the final appeal must be to the attitude of the bed rock surface along the lines in question. If rivers flowed in the courses assumed, and made the valleys in question, the valley floors must have sloped in the direction indicated. If filled later by loose material the slope should still be preserved on the surface of the underlying rock. It is difficult to ascertain this in detail. Our knowledge of the form of the bed rock surface comes largely from deep wells, piers, etc. From many deep wells drilled in and near Cincinnati, it appears that the surface of the underlying rock between the bluffs is not far from flat. No part is known to be more than seventy feet below the level of low water, possibly seventy-three feet in one case or 357 feet above sea level. The well in which bed rock was presumably found at this depth is located at the gas works on the river front. A well at Lockland (Stearns and Foster Paper MR) struck rock at level of only 356 feet above the sea. At Hamilton one of the wells of the city water works struck rock at a depth of 208 feet or 360 feet above the sea. The above named depths are so much alike that it cannot be stated with certainty from this evidence which way the rock surface slopes between Hamilton and Cincinnati. Future wells at either place might reveal lower depths, but this is more likely at or near Hamilton than near Cincinnati, because the large amount of drilling already done at the latter place leaves less chance of discovering the rock surface at greater depths.
North of Hamilton, there is no question that the rock surface inclines southward almost as much as the river falls. The lowest surface reached by wells at Middletown is 410 feet above the sea and the lowest at Dayton 508 feet. At St. Paris, in Champaign County, a well drilled for gas went through 530 feet of drift without reaching bed rock, but even at that depth the bottom of the well was more than 700 feet above the sea. The conclusion may therefore be accepted that the drainage in pre-glacial times was southward toward Hamilton.
Unfortunately, for many miles down the Miami below Hamilton, there have been few deep wells drilled, and none deep enough to reach the rock. At Lawrenceburg, Ind., rock was encountered at an altitude of 365 feet above the sea, but that was not in the middle of the valley. The rock surface in the axis of the valley may be considerably deeper. So far as it concerns the area here treated, the chief conclusions to be drawn from the slopes of the rock surface are that such evidence is nowhere adverse to the plan of pre-glacial drainage as determined by topographic features, and that north of Hamilton the evidence is clear and positive.
The description above given of topography and streams depicts the country at a time just previous to the first ice invasion. The first glacier to reach this area was not the first which invaded the United States. It may have been the third. It is called the Illinoian because it practically covered Illinois, and throughout most of that state the drift sheet which it left was not itself covered by that of a later glacial stage. This Illinoian ice sheet came, in a general way, from the north, but the exact direction cannot be told because few scratches or "striations" made at this time are left on the bed rock of this region. Near Newkirk, on the line of the Chesapeake & Ohio Railroad, about ten miles west from Hamilton, the bed rock is found smoothed and striated. Specimens were collected from this locality by Dr. George Twitchell of Cincinnati, who reports their direction as decidedly east of south. Dr. Twitchell also reports smoothed and striated rock surfaces on the bluff south of Excello. It will be observed that the bluff at this place is in the form of a wedge pointing north, and was, therefore, peculiarly exposed to the force of the southward moving ice.
Effect of Outwash Before the Glacier Arrived.-With the advance of the ice sheet from the north, the system of drainage above outlined began to be disturbed. Even before the ice reached Middletown, when it entered the upper basin of the Miami, the effect began to be felt in this area. No doubt the Miami quickly became an overloaded stream like other streams which receive drainage from glaciers. Its valley began to fill up with sand and gravel. Some of the sand and gravel now resting on the bed rock beneath the present streams may have been laid down at that time and never since disturbed. The partial filling of the Miami Valley raised the outlets and checked the currents of all its lower tributaries. They must then have begun to deposit sediment. If the deposits of each stage could be separated from all others and recognized, it would doubtless be found that some of the sand and clay passed through in drilling wells in Mill Creek Valley, the Norwood trough, and the Union Village trough were dropped at that time from the slackened waters of the ancestral Ohio and Todd's Fork.
Changes in Todd's Fork and Little Miami(See note 13)., --When the glacier reached the bluff at Excello, south of Middletown, the current of Todd's Fork was stopped, and its waters backed up in a lake. On the bottom of this lake there settled mud brought down by the stream, and mud with coarser sediments washed out from the ice. These lacustrine sediments covered the alluvial sediments deposited when the current was first slackened. It is not now possible to separate them. Later, however, when the ice advanced farther, a sheet of bowlder clay was laid down on top of the alluvial and lacustrine sediments. A well one mile south of Union Village pierced twenty-eight feet of such bowlder clay between the depths of 102 and 130 feet. Between this bowlder clay and the underlying limestone were found three feet of gravel. No alluvial or lacustrine clay was identified at this place.
The ponding and rising of the water in this valley caused it to back up its tributaries and at length to overflow the divide at Foster. The whole Little Miami was later buried by ice, but before the ice again disappeared from the mouth of this valley at Middletown the new gorge at Foster had been cut so deep that the waters could not resume their old course. Thus Todd's Fork became the upper course of the Little Miami, or, it may be said, the Little Miami came into existence at that time. The lower valley of Todd's Fork was left without a stream. During the retreat of the ice front when the glacier again blocked the outlet near Middletown, there may again have been a lake in this valley, and in this lake clay would have been laid down on top of the sheet of till. Such clay is found in great abundance. The well mentioned above is said to have passed through nothing but gummy clay between the depths of 6 to 102 feet. A well at the canning factory at Camp Hageman went sixty feet through the same "gummy clay," finding gravel below. It is not possible to say how much of this was laid down in the lake which existed during the retreat of the first ice sheet and how much in the similar lake which may have existed during the advance of the second ice sheet. It is quite probable also that somewhere in that ninety-six feet of "gummy clay" is a sheet of till not detected by the drillers, which was laid down by the second ice sheet, only to be again covered by lacustrine clays laid down during the wasting of that sheet, but before it was melted away from the mouth of the valley below Middletown. At Oakland there are seventeen feet of till just beneath the soil, and near Middletown several square miles of the valley floor are underlain by till. These sheets, at least, were deposited during the later ice stage, and rose above the level of any lakes which may have existed during the last retreat of the ice.
It may occur to the reader that if the drainage changes resulting in the formation of the Little Miami were thoroughly established before the retreat of the first ice sheet, there was no occasion for any later lake in this valley, or at least not in that portion of it which drained toward the newly formed Little Miami. Yet the large amount of fine clay which forms the larger part of the valley floor and, in part at least, overlies the later drift, shows that the valley must have held standing water, even as late as the retreat of the second ice sheet. The reason for this last lake is no doubt found in the great deposit of kame gravels which cross the valley near its southeast end. If lakes also existed during the retreat of the first ice sheet and the advance of the second, these may be explained by assuming that the valley was similarly obstructed by deposits of the first glacial stage.
The Cut-Off at Andersons Ferry. - As the ice advanced nearer and nearer to the junction of the former Miami and former Ohio, just south of Hamilton, the current of the latter stream was more and more checked by detrital filling at its mouth. Eventually when the ice itself had crossed its mouth, perhaps after advancing a considerable distance to the south, the current of the former Ohio was stopped and its waters rose in a lake. As in all similar cases there were sediments in this lake derived partly from the inflowing stream and (near the ice) from glacial outwash. The latter would be distinguished by some admixture of feldspar and other minerals from the igneous rocks of Canada, but, except for that, it will probably remain impossible to differentiate these lacustrine sediments from those river sediments which were deposited during the slackening of the current. In any case, clay, sand, and some gravel would be deposited during the advance of the ice, and these would later be covered by a sheet of till when the ice passed over. This agrees with the data obtained from drilled wells.
It is possible that by the time the ice reached Lockland the valley had been filled to a height of 500 feet above the sea, for no till is well authenticated at a lower depth. Farther north the glacier arrived earlier, before the valley had been so deeply filled. The wells of the Glendale Waterworks encountered some till down to the level of 472 feet above the sea; beneath that level was found mainly sand.
The impounded waters in this valley necessarily rose until they found an outlet. Probably for a considerable time they flowed along the edge of the ice when it stood against the bluff south of Symmes Corners. This probably accounts for the fine steep.bluff west of that point. This bluff has almost no gullies and bears all the appearance of youth. Its youthful appearance may have been again renewed in like manner during the later glacial stage.
When the water could no longer escape by following the edge of the ice, it rose until it crossed the divide between the headwaters of the small streams at Andersons Ferry. When the ice finally retired, leaving the old valley partly filled with drift, this notch was already cut so deep that the stream failed to resume its former course.
Changes in the Miami.- Miami River can scarcely be said to have existed while the glacier covered Butler County, its entire drainage basin being covered by ice. The drainage from this area no doubt gave rise to many streams running over the ice and leaving its front at different points. Some drainage may also have followed the present river valley beneath the ice. After the glacier filled the valley west of Venice, the immediate drainage from the ice necessarily chose the lowest course it could find among the hills to the south. For a time this lay along the present course of the Miami through the narrows past New Baltimore and south of Miamitown. This line was marked out by the small pre-glacial streams shown in figure 48. On the disappearance of the ice this line of escape for glacial waters became the permanent channel for the Miami, because the old valley west of Venice had been partly filled by glacial drift.
It is not probable, however, that the new channel past New Baltimore and Miamitown was cut down entirely by waters coming from the glacier. As pointed out above, it is probable that when the ice front stood some four or five miles south of Hamilton, the waters of the ancestral Ohio escaped westward for some time between the ice on the north and the bluff on the south. This condition may well have continued for some time after the Miami Valley further west was completely obstructed. If so, the entire volume of the master stream was available for some time to assist the cutting of the new valley past New Baltimore and Miamitown. Judging from the size of the present Miami Valley in that section it is quite probable that its cutting was thus helped by the stream from the east, at a time when the drainage of the Miami basin was dissipated and relatively ineffective.
Other Cut-Offs by the Ohio.- Later still the glacier reached the highland west of North Bend and again dammed the main stream, which by this time was flowing past Andersons Ferry, but turning north at North Bend and then west at Cleves. By this obstruction the Ohio was again ponded, and its waters rose until they crossed the notch between headwaters southwest of North Bend along the line of the present Ohio.
Whether before or after this event, the ice reached the bluff south of St. Bernard. Previous to that it is probable that all the water of the Ohio and Little Miami flowed around to the north of Cincinnati through what is now the Norwood trough. When the ice reached the upland of Cincinnati the water rose above the col or notch opposite Dayton, Ky., and flowed directly west, thus completing the present course of the Ohio.
Changes in Dry Fork.-At an early stage in the advance of the ice the former valley of the Dry Fork of the Whitewater was obstructed west and south of Shandon. That stream then crossed the highland straight to the south toward the Shaker Village. That portion of the valley west of Shandon was fffled with till to a height of more than 660 feet above the sea. It'is not known how much' of this belongs to the first glacial stage and how, much to the second. This filling is at least 112 feet deep, as known from a well near the center of the valley. It may extend much deeper, since bed rock was not reached. A well farther west, and belonging to Mr. T. Schradin, is only about one-eighth mile east of the present stream, and started thirty feet above the stream level. The record of material passed through in drilling this well corroborates the account of drainage changes here given. The material is shown in TABLE III.
The till at the top is that of the later glacial stage, and shows that the fine clay was deposited as the result of an earlier obstruction. The fine clay represents lake filling, and the peat may have accumulated either on the bottom of the lake or (more probably) in a depression on an old flood plain.
Where the stream crossed the upland to the south it now flows in a narrow steep valley more than 120 feet deep, and possibly 100 yards wide at the bottom. At places its lower ten or twelve feet are cut in rock. How much of a notch the stream found in the upland along this new course cannot be told. It may have been enough so that the drift alone in the valley to the east was sufficient to turn the stream in this direction. Otherwise, it would be necessary to assume that the ice either blocked the old valley before it reached the new or remained in the old valley longer.
Thickness of the Till.-The deposit of till on the uplands varies from almost nothing to a thickness of ten or fifteen feet. Thicker deposits are generally confined to valleys and ravines. The average thickness, aside from the broad valleys, may be from one to five feet near the Ohio, and from eight to twelve feet near the Hamilton-Butler County line. This applies to the area not covered by the later drift sheet. Farther north where it is overlain by the younger till sheet, its thickness is hard to estimate because of the difficulty in distinguishing the two formations. In the more hilly sections bordering the Miami Valley from Venice to Lawrenceburg, this drift is almost as thin as on the bluffs of the Ohio.
Character of the Till.-The Illinoian till is generally a hard blue clay (yellow near the surface), containing many fragments of limestone. These fragments are generally small, many good exposures showing no fragments larger than one or two inches. However, fragments the size of a hand are common, and the dimensions may reach a foot or more. The majority of these fragments are of blue limestone, being from the same formations as those which underlie the area. Many others are of light-colored limestone, such as underlies western Ohio and eastern Indiana farther north. A very few, perhaps one or two per cent, are of igneous rocks, granite, basalt, etc., and metamorphic rocks, gneiss, quartzite, etc., derived from north of the Great Lakes. But although such stone are rare in the drift, the few large bowlders which are found on the surface are nearly all of this character.
This till or bowlder clay is generally very hard. It cannot be shoveled, even where free from large stones, and offers considerable resistance to a pick. It is generally very calcareous, effervescing freely with weak muriatic acid. In such tests, the limestones must of course be avoided. This calcareous quality is absent only within a foot or so of the surface, where the lime has been leached out in the process of weathering.
Another surface phenomenon in the till is a change of color from blue (generally very pronounced) to yellow. This is due to the complete oxidation of the iron as in case of our local limestones. This oxidation generally goes down from five to ten feet, and may go farther. The oxidized till retains its calcareous nature except very near the surface.
Another feature of this till consists in the prominent cracks which are sometimes seen to traverse it in fortuitous directions where freshly exposed in the bank of a stream. In many cases water is seen oozing from these cracks, and along all of them the clay is completely oxidized for an inch or more on both sides.
Southern Limit of the Drift.-The exact southern boundary of the drift is hard to determine because of its thinning and gradual disappearance. It is thin or absent on the steeper bluffs of both the Ohio and the Miami. There are few good exposures in Kentucky, but, on the other hand, there are few large areas without some trace of glacial deposit. Wright reports twelve feet of till one mile east of Hebron(See note 14), and there are small exposures in the cuts along the Fort Mitchell car line southwest from Covington. Gullies from two to five feet deep are regrettably common on the steeper slopes, and in the bottoms of these it is not uncommon to find a few igneous or metamorphic stones of Canadian origin. Where a few of such are found it is safe to assume that many more limestones were also left by the glacier, but are not always distinguishable from the residual fragments in the mantle rock.
An easier but less accurate method of determining the limits of the former ice sheet is by means of foreign bowlders found in creek beds. It is a fair assumption that such bowlders came from the drainage basin of the stream in which they are found. If the stream is flowing toward the north, then the local limit of glaciation is at least as far south as the bowlders. Practically all of the small streams and ravines in the southwestern part of this area, west of Banklick Creek, Ky., contain some foreign bowlders. Gunpowder Creek at and beyond the southern margin is no exception. Some of these may have been carried south by this creek beyond the point where they were dropped by the ice, but that could not have been the case with the bowlders found in the small tributary ravines. The bed of Bullock Pen Creek, which flows eastward to the Banklick from Florence, Ky., just south of the limits of this area, has many large crystalline bowlders, some of them three or four feet in diaraater. Probably the ice reached at least that far south. The area between Banklick Creek and the Licking has furnished a few foreign stones as much as five or six inches long. It is not improbable that the ice covered at least its northern part. Similar evidence seems to indicate that the ice invaded the northern end of Campbell County, Ky., perhaps as far south as Fort Thomas. In the small streams and ravines of the south bluff of the Ohio east of Fourmile Creek, foreign stones are somewhat plentiful.
Topographic Effect.-South of the Ohio and for a few miles to the north, no topographic changes due to glacial agency can be detected except such as result from the drainage changes already described. Even near the Hamilton-Butler County line the present valleys and even the larger ravines seem all to be of pre-glacial origin, indicating that the drift deposit was too thin to obscure them. The. general surface seems to have been coated with drift over hill and valley alike and it remains so to the present except in the channels of some streams and on some of the steeper and more exposed slopes. Inferences as to Behavior of the Ice Sheet
From this description of the character and distribution of the drift, and the topographic effects of the ice invasion, some things may be inferred concerning the general character "of the glacier and its behavior.
Thickness of the Ice.-It is a fair inference that the ice on the uplands was everywhere thin. Otherwise there would have been at least local erosion and striation of the underlying rocks. Till was deposited without erosion, not only on weathered rocks but on laminated sands and clays in the valleys. This can be seen in the fine section along the Pennsylvania Railroad between Claire Station and Shademore, and again in the bluff one mile north of California. It may well be that in these cases the ground over which the ice advanced was frozen, but the facts are none the less noteworthy as showing a contrast in behavior between the ice which moved over our soil-mantled hills and that which plowed the great furrows in the massive limestone of Kelleys Island.
Another inference is that the ice was not only thin, but thinning as it advanced southward. This is inferred from the fact that there remained no drift to be deposited as terminal moraine at the limit of advance. Another inference to account for the same fact is that the ice did not remain long at the southern limit of its advance. These two assumptions support each other. The greater the stress placed on one, the less dependence need be placed on the other.
Derivation of Its Load.-It is clear also that the ice was handling chiefly local material. Probably one-half of all that was laid down in this area was also picked up within its limits (if that can be said to be picked up, which is merely dragged along between the ice and the earth). Probably nine-tenths of the material came from the states of Ohio and Indiana, and ninety-eight or ninety-nine per cent from south of the latitude of Lake Superior, that is, from the area of unmetamorphosed sedimentary rocks. Some of the drift near the Ohio looks as though it were residual mantle rock, slightly disrupted by the glacier, and with an occasional foreign stone dropped or pressed into it.
The glacier did indeed carry some bowlders from the far north, either in the ice or on the surface, brought there by the melting of the ice above. Because these were carried in or on the ice they traveled to its edge, and were there deposited in somewhat larger proportion than along the way. Probably no creek bed in the area of the Illinoian drift has so many large foreign bowlders as the Bullock Pen, just beyond its southern margin.
A study of the deposits left by the Illinoian ice sheet in the Mill Creek Valley, and other great valleys of the region, indicates that its behavior in these valleys differed materially from its behavior on the uplands. Three remarkable features of the drift in these great valleys are its thickness, the interstratification of till and water-laid sediments, and the terrace-like or flat-surfaced form of the deposit.
Topography of the Valley Drift - Generally the glacial deposits of this age form terraces in the wide valleys, the surfaces of these terraces being approximately flat, but sloping a little from the bluffs toward the axis of the valley. Usually they abut against the bluff in a fairly abrupt manner. Locally, where the bluff is not steep or well defined, its slope merges with that of the deep glacial deposits in the valley. Where the terrace is well defined the altitude of its outer and upper edge is generally not much above 640 feet, or below 620 feet. Where the deposits are preserved in the middle of the broad valley, as in Mill Creek Valley north of St. Bernard, the slope toward the axis may carry the altitude below 600 feet, even as low as 580 feet.
Interstratified Materials.- At the edges of this terrace the material is locally exposed in section as low as 500 feet above the sea. As far down as exposed, bowlder clay is seen to be interstratified with the sand, gravel, and clay. At places almost the entire thickness seems to be of bowlder clay; elsewhere it is almost entirely water-laid. material, but there are few places where bowlder clay does not appear at the top just beneath the soil.
The water-laid deposits in these terraces are perhaps equal in volume to the till, and hence very important. They consist of gravel, sand, and clay. The gravel is composed almost entirely of limestone pebbles, generally not more than two or three inches in diameter. The few foreign pebbles present have more importance in indicating the history of the gravel than in making its bulk.
The large amount of limestone in these gravels, and the favorable conditions for percolating water, have favored the cementation of the gravel into conglomerate. This is more especially true at the edge of a gravel terrace where ground water escapes by seepage or in springs. There the water evaporates, precipitating its carbonate of lime. In terrace fronts, therefore, these gravels are generally, at least partially, cemented, and not infrequently consolidated into a firm conglomerate. The edge of the terrace east of California furnishes abundant illustrations. The sand beds are also locally cemented, especially in concretionary forms.
The reason for the relatively flat surface of these valley deposits is suggested by the nature of the materials. Deposits by overloaded streams necessarily make a nearly level surface. The terrace-like character and striking uniformity of altitude of the drift deposits in the larger valleys are, therefore, due to the large part played by water in their making. Glacial lobes acting alone would indeed have made thick deposits in these valleys, but the surfaces of such deposits would have been more rolling. The fact that all the till in these valleys was laid down on a nearly level floor of alluvium tended to give it a level surface. The further fact that, in most places, the till was laid down in several sheets, the surface in each interval being leveled up by waterlaid sediments, tended still further to give to the final till deposit a nearly horizontal surface.
Glacial Lobe in Mill Creek Valley.-The great thickness of till in the broad valleys, when compared with the thin deposits on the upland, suggests not only that the ice was thicker and more powerful in these valleys than over the intervening uplands, but also that it worked for a longer time in the valleys than on the uplands. From theoretical considerations it was seen that the former fact is certain, and probably the latter also.
Mill Creek Valley, whose axis is from 300 to 400 feet below the adjacent uplands, and which lies in the direction of the ice movement, was especially favorable to the formation of a glacial lobe. A tongue of ice way even have pushed down this valley in advance of the ice when it first covered the area, but it is much more probable that the ice survived in this valley as an active part of the glacier after it had melted away from the uplands.
Behavior Inferred from Typical Sections.-The effects of greater power, exercised for a longer time, is seen in the thick accumulation of bowlder clay at the bridge over MiU Creek, west of Carthage, on the road to College Hill. The west bank of the creek is here more than fifty feet high, showing the following section (TABLE IV). It is not known how deep the bowlder clay extends beyond the base of the section, and it is probable that the top of this section does not indicate the upper limit of the deposit before erosion.
There is no indication in this section that the deposition of bowider clay was interrupted, except at the time when the five feet of laminated clay was being deposited. Such clays may be deposited in quiet pools beneath the ice, but such pools would necessarily be very local. It is more probable that the ice front had temporarily receded. As seen below, the water-laid beds shown in other sections are too abundant and continuous to be easily explained without assuming that the ice occasionally melted back a considerable distance.
A quarter of a mile f arther north another section exposed in the same bank is shown in the section of TABLE V. An effort to correlate these two sections shows how irregular are the deposits made by glacial waters near the edge of the ice. Making suitable allowance for the slumping of such deposits, and for error in measurement of altitudes, it may be that numbers (1) and (2) are identical in the two sections, that is, that they represent continuous beds, the clay below showing a time of ice disappearance, and the till ab6ve.representing the final advance. It is certain, however, that the fifteen feet of sand (No. 4 in the northern section) is not represented in the other section one-fourth mile to the south. This sand may have accumulated beneath the ice, or it may have been laid down in a depression in the till surface during a temporary recession, or it may have been laid down at both places during such a recession, being covered up by till and preserved in the northern locality, and eroded away by the readvancing ice (or by water in the meantime) in the southern locality. The chief importance of these two sections consists in showing, first, the general constitution of the terraces of Illinoian age and, second, the great local variation. The various assumptions here made as to the behavior of the ice and the circumstances of deposit have widespread application.
An exposure at the edge of the same terrace due east of Hartwell in the east bank of Mill Creek, but in the middle of the wide valley, is shown in TABLE VI. In this section, again, there is reason to think that Nos. (1) and (2) are identical with the corresponding beds in the other two sections two miles farther down stream, but correlation of the sands and till sheets below is mere guess work.
The several proportions of bowlder clay in the three sections are eleven-twelfths, five-eights, and one-third. The last named of the three sections is in the middle of the broad valley, and it is not improbable that the normal proportion of water work to ice work should be greater in the middle than at the sides. All three sections agree in showing that the outwash from the Illinoian glacier in this part of the Mill Creek Valley was limited to fine material, not coarser than sand.
In its southward movement the glacial lobe in this valley came against the high hills of Avondale. The deep deposits of till, with varying amounts of interbedded sediments, built the levels on which Bond Hill and St. Bernard are situated. South of St. Bernard, and reaching within half a mile of the Zoo, extends the same level, now sharply dissected by streams of later origin. Most of the natural exposures here are in bowlder clay. The somewhat higher land (approaching 700 feet) north and west of the Zoo has a deeper covering of till than is common for that altitude, because of the tendency of the ice lobe from the north to climb the slope which lay directly in its way.
Cumminaville Branch of Mill Creek Valley Lobe.- At this place the lobe divided. A part of it passed west to Cumminsville and perhaps thence south, down what is now Mill Creek Valley. Its deposits, ic-laid and water-laid, built the extensive (though now much dissected) terrace which includes Spring Grove Cemetery and Parker Woods-the latter now a part of the Cincinnati park system. The level of this terrace is marked by that of Hamilton Avenue to beyond Spring Lawn Avenue, and various remnants of the same level may be seen across the valley to the west. Down this far the terrace is largely, though not wholly, of bowlder clay. Farther south the only remnant is a shoulder or interruption, at the appropriate level in the slope of the west bluff north of Fairmount. This is made by a cemented gravel, the counterpart of that which makes this terrace elsewhere, to be described below. It is only to be expected that, as the end of the ice lobe is approached, the times during which it covered the ground should become shorter and shorter, and the intervening times of absence longer and longer. The proportion of water-laid to ice-laid material should be correspondingly increased.
Lobe in the Norwood Trough.-The branch of the Mill Creek Valley lobe which turned to the east was guided by the Norwood trough, the old valley in which had flowed the former Ohio. This valley, with its former axis not more than 360 feet above the sea, was filled partly (with glacial lake clays??) and the rest with glacial deposits and outwash in the same manner as Mill Creek Valley described above. The level of its axis was raised to nearly 600 feet. As far down as this filling is exposed it consists of the same materials as those named in the sections near Carthage and Hartwell. Where the Norfolk & Western Railway crosses the Duck Creek road, till is exposed to a depth of fifty feet. A short distance to the northeast is a similar exposure, but here the till is interbedded with water-laid materials. Duck Creek, a small stream developed at a later time, has cut down into the new surface of the valle exposing the materials to a depth of twenty to forty feet. Till is the chief substance seen in most of these exposures.
By far the best exposure of the deposits which fill the Norwood trough is seen at the east end where this trough borders the Little Miami Valley along the line of the Norfolk & Western and the Pennsylvania railroads. Along this line for a distance of about two miles. the Norwood trough ends in an escarpment which at its east end is more than 100 feet high, overlooking the flood plain of the Little Miami. This escarpment is primarily the bluff of the Little Miami and due to lateral corrosion by the stream, but it has been freshened and steepened by excavation for the railroads whose tracks are twenty feet or more above the stream. A section of about eighty-five feet above the railroad, and near the east end, shows the beds in the following order named from the top downward (TABLE VII).
The thickness of the uppermost sheet of till, a short distance away, is much greater than it is here. It seems to be fairly continuous, forming the floor of the present Norwood trough. The lower sheet is exposed for about one mile, its average thickness being about the same as that shown in this section.
A number of deep wells have been drilled in this trough for the Norwood waterworks, the material passed through being recorded as gravel, sand, and clay, but it has never been observed with sufficient care to determine whether some of the beds thus described are bowlder clay. It is highly probable that such is the case, and that the material for some distance below the level of 500 feet consists of alternating beds of the same nature as those here exposed.
The continuity of the beds in this exposure reinforces the impression made by the observations near Carthage and Hartwell, namely, that the ice alternately advanced and retreated considerable distances. It may be that the entire Mill Creek Valley and Norwood trough were several times covered by the ice, and completely laid bare in the intervals.
Lobe in the Little Miami Valley. - The Little Miami Valley, being narrower and less straight than that of Mill Creek, did not offer such favorable conditions for a persistent lobe of the glacier. The tendency is clearly shown, however, as seen in the deep deposits at Milford, Miamiville, and elsewhere. Where composed largely of till, these deposits are rather less uniform in altitude and less distinctly terrace-like in character than those in the valleys above described. This should be expected in a crooked valley where a glacial lobe tends repeatedly to push up one bluff or the other, and must constantly readjust its movements. Since the Little Miami Valley is relatively narrow, and the stream much stronger than Mill Creek, the deposits of the Illinoian stage have been in large part carried away, but they remain in fragments as far north as Symmes Station. The bluff known as Cedar Bank, a mile east of Miamiville, is more than 100 feet high, its upper half being of till. The covering of talus and forest below may likewise conceal till. In the almost inaccessible upper half there seems to be some evidence of water work in several horizontal streaks of silt.
The high hill in the town of Milford is of the same nature, and shows that deposits of this epoch once filled the valley from bluff to bluff. The material of this hill is not well exposed, but a mile to the north, likewise in the middle of the valley, are two smaller hills rising to about the same altitude (about 650 feet). These are likewise remnants of the same deposits, and one of them, nearest the railroad, has been partly cut away, exposing a good section. The material there exposed is shown in TABLE VIII.
A short distance away, excavation has shown that the till at the
base goes at least ten feet deeper.
Four miles southeast of Milford, against the north bluff of the East Fork, is a terrace, which, according to the U. S. topographic map, rises above the level of 660 feet. It is composed of limestone gravel, which at the outcrop is firmly cemented. It may be that its substructure is in part composed of till. It indicates that the valley of the East Fork, at least near its mouth, was filled to the same depth as that of the valley at Milford. Probably the ice lobe divided at the bluff behind South Milford, one part going a short distance up the valley of East Fork.
The main ice lobe continued down the Little Miami Valley and joined that which came through the Norwood trough. The united stream of ice (and the waters which issued from it) aggraded this portion of the valley to about the same level as the Norwood trough. The proportion of water-laidd to ice-laid material increases to the south. In the east bluff of the Little Miami, near where the traction lines to California and Mount Washington cross the stream, is a good exposure of sand and silt, some of it minutely laminated. This is overlain by at least fifteen feet of till, representing a deposit which was no doubt thicker before erosion. This till now caps the ridges and remnants of a very much eroded terrace, similar to those farther north, but rising little above 600 feet. Other sheets of till may perhaps be interbedded with the sand and silt below the level of the exposure.
Lobe in the Ohio Valley.-The same glacial lobe continued south beyond Coney Island, making a large deposit of till at California, which caps a terrace three-fourths of a mile wide. This is the terrace in which the upper settling basins of the Cincinnati waterworks are dug. This sheet of till is locally at least twenty feet thick. It may be thicker, and there may be deeper beds of till not exposed, but the most southerly point at which two distinct and well separated beds of till are exposed is the bluff at Batavia Junction, south of Madisonville, described above. It may be that the ice tongue pushed forward but once to California, and that this event came late in the Illinoian stage when the great valleys had already been filled to an altitude of approximately 600 feet with outwashed sand, gravel, and silt.
Going southeast from California, thut is, up the Ohio, the last good exposure showing both till and gravel in the same section is on the road from Coney Island to Mount Washington, about one-half mile from the former. Here, in a gravel pit beside the road, twenty feet of till rest on forty feet of gravel and conglomerate. The till contains abrupt pockets of gravel, indicating the abundance of water beneath the rapidly wasting ice. The gravel consists mainly of limestone pebbles, the largest being three or four inches in size. (See Plate VI-A.)
Eastward, beyond this road, the terrace diminishes in width and clearness to FiveMile Creek, beyond which it is not recognized. It is composed largely of cemented gravel (Pleistocene conglomerate). Outcrops of such beds appear in the terrace edge at California. They are very conspicuous at Coney Island and at various points to the east.
The reasons for referring the till of this California terrace to a lobe of ice following the direction of the valley, rather than to the widespread sheet which is known to have reached the Ohio, and locally to have crossed it, are as follows:
(1) The adjacent bluffs are essentially without drift, and the uplands have very little. (2) The deep till is in terrace form, neatly limited to a certain altitude. This abutting of a horizontal surface against the steep bluff is a feature of the drift only, and does not belong to the underlying rock. (3) There are no similar thick deposits of bowlder clay in the Ohio trough farther southeast, or even in the Cincinnati basin, which was crossed by the ice sheet, but not favorably situated to be reached by valley lobes. (4) This interpretation is consistent with the evident history farther north.
In speaking of a glacial lobe occupying Mill Creek Valley, and advancing south through the Norwood trough and Little Miami Valley and up the Ohio, it is not intended to imply that a lobe extended this entire distance after the adjacent uplands were everywhere free from ice. How far the lobe at any one time extended forward from the main ice sheet is not determined. The significant points are that the main ice movements followed the great valleys, and that glaciation affected them in a manner quite distinct from the way in which the uplands were affected, and to a very much greater degree.
Behavior of the Ice in the Miami Valley. - The trough of the Miami below Lindenwald was less favorably situated to develop an independent glacial lobe, since the section between Hamilton and Harrison lies almost transverse to the direction of glacial movement. There is much till in the New Haven trough, but most of that which appears at the surface was deposited in the later glacial stage. The extent to which this part of the valley was filled by Illinoian drift has not been determined.
