Primary Decisions

Less in the public view but, as things turned out, of greater import was an argument of a technological nature. The now classic argument between the merits of inland seasonal navigation and those of railroads began even before rail lines existed. The controversy emerged with a published letter from John Stevens (1812), the Hoboken inventor, to Governer Morris, Chairman of the Commissioners of 1811, advising that a relatively small research investment in a steam railroad would forestall the early obsolescence of the canal. Stevens explained the principles involved - for example, that the square law of resistance does not apply to motion on rails as it does to motion through water, and that a rail line is more flexible in location of route and is usable all year. The proposal was not taken seriously because canals, towpaths, and horses were known from long British experience to be a proven technology, whereas rails were still only an untested concept. Anticipation of technological change to this time is not a part of water planning in the United States (White, 1969).

The clear and present advantages of building a canal seemed to be greater than the political and technologic uncertainties of the future, so the project moved toward approval. The Commissioners of 1816 appointed to design a canal, needed only to address those engineering matters that were necessary to give physical bounds to their proposal-to adapt the project to the terrain. Thus they proceeded to resolve such fundamentals of water engineering as choosing between an Ontario canal or an Erie canal and between a ship or barge canal; selecting the profile and the route location; and, finally, the cross section-deciding on the width and depth of the canal

AN ONTARIO CANAL OR AN ERIE CANAL

Far from determining the form the project eventually took, the geography actually seemed to dictate an entirely different and seemingly easier and less expensive alternative (Goodrich and others, 1961). From the earliest times, inland navigation meant the use of navigable rivers and lakes. And so it has generally been in this country where rivers, especially the great continental streams and the Great Lakes, were main avenues of settlement and commerce. It was therefore natural enough to view the geography of New York State in that way-the immediate object being the transport of goods from tidewater in the Hudson River at Albany, across the topographic saddle at Rome, to the Great Lakes region. However, navigation on the upland streams, such as the Mohawk River, encountered many difficulties in passing from pool to riffle, and as the river levels varied between flood to drought. Altering such rivers to improve navigability I-early al- ways involved the construction of lift (double gate) locks to avoid natural falls, riffles, or other obstruction's to navigation. River water above the falls was diverted into a canal or sluice, and thence through a set of lift locks leading to a sluice that returned to the river below the falls. Such improvements had been made in the 1790's to permit boats to avoid Little Falls on the Mohawk River and Great Falls on the Potomac River near Washington, D.C. (Civil Engineering, 1972).

A look at a map of New York State (see fig. 2) would therefore suggest a low-cost scheme that would use the natural watercourses -the rivers and lakes. One would put sluices and locks around local obstructions to navigation along the Mohawk River, across the Rome summit, thence by way of Wood Creek, Oneida Lake, and Oswego River to Lake Ontario. Sluices and locks would also be built around Niagara River and Falls between Lakes Erie and Ontario. But the prospects of substantial savings in cost were lessened by the very considerable changes in elevation required, 150 feet down from the Rome summit to Lake Ontario, and then 326 feet up from Lake Ontario to Lake Erie, in addition to the 420 feet up the Mohawk River to the Rome summit.

Moreover, experience in Great Britain and on the Mohawk and Potomac Rivers in the United States in the 1790's proved that river improvements were thoroughly unsuccessful. The navigability of a river in its natural state between sluices is highly variable, being subject to disruption by flood, drought, and sedimentation. This experience in Great Britain had led to the adoption of independent off-river canals. Their success was proven by the Bridgewater canal, built in the 1760's for the transport of coal to the factories of Manchester. The advantages of independent canals were conveyed in a letter from Benjamin Franklin, written from London on August 22, 1772, to the mayor of Philadelphia (Ringwalt, 1888).

*** Franklin's description fit that of the Bridgewater and other English canals (Hadfield, 1968, p. 39) that followed the contours along the valley sides, passing from one river drainage to another by crossing low saddle divides that separated one catchment from another. These schemes worked.

Convinced by the evidence that to use "the beds of rivers for internal navigation" was impractical and inefficient, the Commissioners of 1811 rejected that practice right at the outset of their report (p. 3-4). This decision was correct then and for the reasons given. With this decision for an independent canal, the course, profile, dimensions, and flow of the State's inland rivers no longer governed, and a choice of routes became possible. This is the kind of decision that might have been costed out; but policy issues prevailed. The route by way of Lake Ontario was rejected. A direct route to Lake Erie was adopted to avoid the feared possibility that shipping afloat on Lake Ontario might be diverted by way of the St. Lawrence River to market at Montreal. An independent canal as a waterway across the State then created a need for numerous subsidiary yet still major decisions, especially those concerning size, profile, and route.

SIZE

As a first approximation of size, the Commissioners of 1811 considered the possibility of a "sloop" (that is, ship) canal, by which was meant prism dimensions sufficient to accommodate vessels capable of navigating upon the Great Lakes and the Hudson River, thus avoiding the costs- of cargo transfer to and from small canal barges. However, it was evident to the Commissioners of 1811 that

"If the passage were only a few miles, the propriety of bringing vessels of 8 feet drought of water across, if practicable, would be readily admitted; but it may well be questioned whether to save the expense of lading and unloading at each end of a canal three hundred miles long, the expense of cutting two yards deeper than would otherwise be necessary, ought to be ,encountered." (p. 20)

The canal was therefore to be designed for barges. As will be described, the ship-canal idea remained viable during the 19th century and into the 20th century.

PROFILE

A canal leading directly to Lake Erie, some 570 feet above the level of the Hudson River, offered the prospect of a bold, single stroke resolution of decisions as to profile and water supply. Such indeed had been recommended by the Commissioners of 1811 in the following terms:

"The difference in level (from Lake Erie to the Hudson) being upwards of five hundred feet, all the descent which can prudently be obtained by an inclined plane, is so much saved in the expense of lockage; and in all human probability, the transportation for centuries to come, will be of so much greater burden from the interior country than back from the sea, that a current from the lake is more to be desired than avoided, more especially as it will, in some degree, counteract the effect of frost." (p. 21)

With depth and width considered in a qualitative way so that the friction would counteract the tendency for velocity to accelerate, and with preliminary caution, the Commissioners assumed hypothetically that a canal should have an average descent of 6 inches per mile (p. 24). According to their levels, the grade would pass above Cayuga Lake outlet by some 150 feet and clear above the Rome summit by 42 feet. The canal would follow the hillsides above the river valleys (especially the Mohawk) and when it was necessary to cross major river valleys, as, for example, the outlets of the Finger Lakes, or the Schoharie, a major tributary of the Mohawk, rather large embankments and aqueducts would be required. This grade would leave about 350 feet to be accomplished in locks or an inclined "railway" in the reach between Schenectady and the Hudson where (p. 35) "water used for machinery would probably yield a rent sufficient to keep the canal in repair."

They rejected the plan of a locked canal fed by rivers along the route in order to avoid any dependence of the canal upon rivers for supply and especially to avoid the effects of the anticipated decrease in the flow of the streams to result from deforestation and farming (p. 20). Therefore, in further explanation of a graded canal, they state:

" In a word, if, on due examination, a thing of this sort should be found practicable, instead of depriving the country of water, every drop of which, is needed by its inhabitants, they will gain a great addition from the canal."

Thus, in 1811 the notion was already afoot to use the canal to convey water as well as traffic.

To save lockage, the Commissioners took the greatest slope they considered wise, and in a canal 4 feet deep, at a grade of 6 inches per mile, the velocity would be about 1.5 feet per second (1.0 mph)-somewhat large for a horse-drawn barge. Had the Commissioners included locks to accomplish part of the fall, their profile could have been significantly lower and still have cleared the Rome summit. Another factor also made the scheme impractical at that time. One can now calculate that to supply the water needed for leakage over a 360-mile length from Lake Erie would have required a cross section at its western end sufficient to convey a flow of the order of 600 cubic feet per second. If velocity were to be kept below I mile per hour, a section about 5 times larger than that actually built would have been necessary. The Commissioners had little comprehension of the hydraulics of their scheme.

Although the Commissioners of 1811 repeatedly emphasized the need for more detailed study and examination and modified their profile in their report of the following year, the proposal became the subject of considerable ridicule that put the whole project in jeopardy. The Commissioners of 1816 (1816-17 As. Jour. 40th sess., p. 313-355) therefore adopted the more practical scheme of a locked canal that closely followed the terrain. The profile shown on figure 4 follows the general land surface and, although generally downward from west to east, has a pronounced sag in the central part of its profile. This dip prevented the flow of water from Lake Erie eastward, and created the need to develop feeders from rivers along the route in order to cope with the never resolved water-supply problem, as will be described.

ROUTE LOCATION

After the decision to build an independent locked canal from the Hudson River to Lake Erie had been made, the job was to fit the route to the terrain. The main features of the topography to affect the choice of location were the narrowness of the Mohawk River valley leading from the Hudson River at tide level to the summit level at Rome (elev 420 ft msl) ; the sag in the profile to about 360 feet elevation in the middle or lake division; the north-facing Niagara escarpment in the western region; and finally the elevation of Lake Erie at about 570 feet above msl. (See fig. 4.)

The route itself was examined in detail by the Commissioners of 1816, who resurveyed and marked the whole route and sank-test pits at a numberof places to ascertain the nature of required ex- cavation. The Commissioners report laid out section by section the yardages of excavation and fill, the kind of excavation (earth, marble, rock), and gave directions for many details such as location of the towpath (north or south) and stone work for the culverts. Warnings were given about possible problems with floods and stability of embankments, and of the possible need for feeders (water supply) along the route in addition to that to be fed at summit levels.

Among the many difficulties presented by the terrain to the construction of the canal, there were three of major proportions-the "Noses" or promontories along the Mohawk River where the river flowed against the steep rock wall; the extensive swamps in the middle or lake division; and the deep rock cut in the western or dry division. The broad, flat saddle composed of alluvial deposits at the Rome summit was recognized at the start as the most favorable feature of the terrain. The strategy was to begin there, and so quiet the residual opposition to the canal project by rapid progress.

MOHAWK VALLEY

In keeping with the decision for an independent canal, a route was sought along the narrow valley separate from the river. The canal followed the south or right side of the river for it offered, the fewest difficulties. The main challenge was to find space above the flood levels of the river. In 1817, when construction began, the Mohawk River valley was subject to extensive inundation "as if, to indicate at the commencement, by the height, impetuosity, arid durability of the greatest floods, the exact dimensions and strengths of the works necessary to discharge or resist them." (1818 As. Jour. 41st sess., p. 68). In modern terms, the, flood of 1817 became the design flood. In adhering to the south or right bank, the canal had to pass by rock spurs such as the Noses that crowded against the river. In these places, the canal had to be built up along the river upon a foundation laid on the river bed and protected by stone work against cutting by the river. In such exposed locations the canal was also subject to damage by wash of loose rock and overburden from the steep valley sides.

In retrospect, the problems encountered in the Mohawk Valley inspired the Commissioners to observe in their report for 1824 (As. Jour. 47th sess., p. 547), that

Had this section been commenced originally while information on the subject of constructing canals was merely theoretical, it is probable that the attempt to complete it would either have been entirely abortive, or so imperfectly executed as to have defeated, and perhaps postponed for a century, the accomplishment of the great work of internal improvement

MIDDLE (LAKE) DIVISION

For some 60 miles west of the long summit level where the canal crossed the saddle from the Mohawk at an elevation of about 420 feet, the valley bottoms lay below 400 feet elevation and below the levels of the Finger Lakes whose outlets flowed northward through extensive swamps. This region is a sag in the generally uphill profile from the Hudson River to Lake Erie. It is poorly drained. The valley bottoms are composed of marl-a limy clay - of postglacial origin. "The marl varies in color from a pure white to a yellowish white. It is handled readily when dry or plastic, but becomes very slippery when wet." (Landreth, 1900, p. 576.) Coupled with standing water, this material was troublesome and the builders of the original canal wisely tended to follow a sidehill position at the edge of the swamps, closely hugging the hills composed of glacial sands and gravels (fig. 5). The Commissioners made repeated reference to the difficulties encountered in excavation along these wet lowlands along the Montezuma-Cayuga Swamp, and along those that border the Seneca River which receives the outlets of Finger Lakes and is the master stream of the region.

The difficulty of cutting through 11- 12 miles of wet meadows from Ninemile Creek (outlet of Otisco Lake) to Skaneateles Outlet induced the builders to raise the level of the canal in that reach, a decision that required the introduction of a short summit level called the Jordan level (see fig. 4) (1819 As. Jour. 42d sess., p.201-203). The Jordan summit level - a local summit in a sag ill the profile-was retained throughout the 19th century period of operation of the canal. Soil and water problems did not end with construction because difficulties were created by bank slides and rising bottoms. Continued maintenance was required for navigation.

WESTERN DIVISION

The Commissioners of 1816 had laid out a route from a connection with Lake Erie at Buffalo, thence along the shore of the lake and of Niagara River to Tonawanda Creek. The proposed route then went up Tonawanda Creek (see fig. 2), and crossed the divide between that creek and the drainage basin of the Genesee River at an elevation 75 feet above Lake Erie. This route would introduce a summit at that level with additional locks up and down. It was recommended in lieu of an alternate route to the north that avoided the intermediate summit and its added lockage, presumably to gain the advantages of shorter length, easier construction, and lesser cost. In addition, the route passed through the lands offered by the Holland Land Company, owners of a large tract in the western part of the State.

Water supply at the summit level was, however, a critical matter as the Commissioners had warned. They accepted the measurements of flow made by Joseph Ellicott, one of the Commissioners, with a vested interest as a sub-agent of the Holland Land Company (Whitford, p. 79). The Commissioners reported that Ellicott

"had the sources of this supply gauged, with great care, during the driest part of the last season, which has been more remarkable for severe drought than ever before experienced in that part of the State. Independently of waters deemed sufficient to repair the waste occasioned by evaporation and soakage, these sources consist of ten streams naturally flowing or capable of being conducted to -the summit level." (1816-18 As. Jour., 40th sess., p. 315.)

He found the flow of these 10 streams (not named) totaled 253,435 cubic feet per hour (70 efs), an amount sufficient to "fill 673 locks every day." It was also suggested that the summit pound (see glossary), which covered 1,000 acres, would provide a reservoir of water to supplement the natural flow. The natural flow plus draft on the reservoir was then judged to be adequate for lockages (no account at that time was made of that required for leakage). The southern route was retained in the plan until 1820, when experience with the canal section already built at the Rome summit gave some indications that "more water has been wasted, in it, by evaporation, soakage and leakage than we had anticipated.

And this discovery we deem, in itself, 'sufficient to settle the question, between the two routes." In further support of this decision, they added their fears that the flow of the stream available as feeders will diminish as the land is cleared for farming (1821 As. Jour., 44th sess., P. 868).

Choice of the northern route required crossing of the "mountain ridge" as the prominent north-facing Niagara escarpment was then called (why "mountain" is not clear-the escarpment is of the order of 60-70 feet). Still, there remained the question of levels. For the next year the Commissioners noted that

"ideas about raising level in approach to the mountain ridge to save rock excavation and to have bottom of the canal at the same level as the surface of Lake Erie were abandoned when local feeders were deemed inadequate * * * after much pains, however, to gauge the streams during the last autumn (1921) we determined to adopt the lowest level and to construct the canal in the first place, so as to receive its supply of water from the lake." (Canal Commissioners' report dated 27 Feb. 1822, p. 11-12.)

Hence, a northern route was adopted that "nowhere rose above the level of Lake Erie," thus to maintain the lake as a source of water for the western division; but as explained in section "Hydraulic Computations," the flat slope from the lake to the escarpment limited the amount of water that could be drawn from the lake.

From Tonawanda Creek the northern route headed for a low point in the ridge, the notch made by a stream. that the surveyors had named "Eighteen Mile Creek." With some 60 feet of fall to be taken in one step at the locks descending the ridge, of the 68 feet between the lake and Rochester, not much fall remained to give the necessary slope to the canal in order that, as will be explained in further detail, it could convey the water needed to maintain navigable depth. The remaining slope was literally to be measured in fractional inches per mile. Since lock lifts were then limited by hand-operated timber miter gates to about 12 feet, a tier of 5 chambers in timber was required to descend the escarpment. (See fig. 6.) Called "combined" locks, the upper gate of the lower chamber serves as the lower gate of the chamber next above. To save time and water (as will be explained under water supply), a double tier of combined locks was built at this point. The Lockport combines soon became one of the marvels of the canal.

Together with the rock cut for the approach, their construction involved the largest amount of rock excavation on the route, an important consideration in the days of hand drilling and black powder explosives. The removal of about 600,000 cubic yards of rock was the last excavation in the construction. The connection with the lake was made on the lake itself rather than along the Niagara River which was closer and which had been urged by the citizens of Black Rock (now incorporated with the city of Buffalo). This decision gained an extra 5 feet of elevation and so reduced the amount of rock excavation required into the hard Lockport limestone in the approach to the "mountain ridge." Even so, the Deep Cut, as it was known, extended for some 2 miles west of the Lockport combines. It became part of the tourist attractions at that point.

CHANGES

Aside from some canal shortening and reduction in the number of locks, the 1817-25 choices of route and profile remained throughout the 19th century. With independently powered barge and greatly advanced construction engineering, the 20th century alignment was changed to follow a canalized Mohawk River, and thence across Oneida Lake. The only change in profile was the elimination of the Jordan level. Canal building in New York State, during the 19th century is outlined in the section "New York Canals in the 19th Century" along with a record of the decline of canal traffic for commerce and its growth for recreation,.

DIMENSIONS OF THE CANAL OPTIMALITY

Included among the decisions on route arid profile there lay the issue of size how wide and how deep? Should the canal be built with a small prism to wind easily about the hills and valleys and so to require only relatively low investment, or should the canal be built more substantially to profit by the economies of scale and to be more attractive to shippers?

At the time, canals in Great Britain were either "narrow" or "broad." Narrow canals were built for the passage of 30-ton barges, 7-foot beam by 70 feet long and 3 feet 6 inches draft; broad canals were for the passage of 100-ton barges of 13 feet 6 inches beam, 90 feet long and 3 feet 6 inches draft. The Commissioners of 1811 (p. 32) thought in terms of a canal prism 15 yards wide and 2 yards deep. The Middlesex Canal, built in 1793-1804 between the Merrimack River and the Charles River in Massachusetts, was designed for boats of 9.5-foot beam, 70 feet long, and 2-foot draft.

With this information available to them, the Commissioners of 1816 faced the problem of optimization among elements of costs that were relatively certain and expected revenues that were uncertain. The principles are classic: capital cost as well as maintenance cost increases with size of the prism; traffic capacity increases with size of prism as does the delay and uncertainty of actually putting that capacity to use. Estimates of cost were made at several reports in the range of $5 to $6 million, and in 1812, the Commissioners compared their estimated cost of $6 million with annual revenues expected to be $1.25 million and reasoned that income would be sufficient to cover costs. Some rough computations or at least amputations of how costs would vary with prism size were surely possible and may have been sketched by the design Commissioners of 1816. Nothing of that sort is referred to in their report, but one can piece out a schedule of costs, from the record, as follows:

Estimate Estimate Amortiz of annual Prism Annual tion operation Total (width by capacity of construc- at 6 and annual depth, (million tion cost percent mainten cost in ft) tons) (million for 25 yrs ance million dollars) (million (million dollars) dollars) dollars)

30 X 3 0.5 3.5 0.27 0.03 0.3 40 X 4 1.5 6.0 .47 .06 .53 50 x 5 3.0 12.5 .98 .12 1.1

1 Estimated at 1 percent of construction cost.

The costs were to be paid from revenues collected as tolls which, at the planning stage, would have been highly uncertain. The 1816 estimate of anticipated revenue of $1,250,000 per year would have been adequate for a 5OX5-foot canal. But, recognizing the uncertainty of that estimate, comparisons would need to be made using conditions of traffic significantly smaller and larger than first assumed.

Prism Annual Revenue Revenue (Width x Cost (Million dollars) (million dollars) depth) (Million 50 % smaller 50% larger in feet dollars) Gross Net Gross Net 30 X 3 0.30 0.62 + 0.32 11.25 +0.95 40 X 4 .55 .62 +.07 1.62 +1.05 50 x 5 1.1 .62 -.48 1.62 +.52

1 Limited by capacity to 500,000 tons per year at $2.50 per ton.

According to these results (1) a 50 x 5-foot canal would have been risky; (2) a 30 x 3-foot canal with a low traffic capacity, would have been fiscally "safe"; and (3) a 40 x 4-foot canal would have been fiscally solvent over a wide range of variation in revenues and had potential to accommodate a growing traffic if that might materialize. Some analysis such as this may have preceded the following recommendations of the Commissioners of 1816.

"The dimensions of the Erie canal and locks, ought, in the opinion of the commissioners, to he as follows, viz. width on the water surface, forty feet, at the bottom, twenty-eight feet, and depth of water, four feet; the length of a lock, ninety feet, width, twelve feet, in the clear. Vessels carrying one hundred tons, may navigate a canal of this size; and all the lumber produced in the country and required for market, may be transported upon it." (1816-17 As. Jour., 40th sess., p. 313-314)

As shown on figure 7, the canal was built to these specifications, except that the locks were built 15 feet wide. (All locks except those in the Lockport combines were single.) In 1824 the Canal commissioners (1825 Sen. Jour., 48th sess., p. 289-291) projected tolls to reach $1 million by 1836 and $2 million by 1846 with a potential annual revenue of $9 million. Enlargements were planned in 1836 when tolls had exceeded the projected $1 million and when traffic carried was about 700,000 tons, 50 percent of its capacity, and growing at an annual rate of about 45,000 tons. Thus encouraged, the Commissioners in 1841 projected revenues to increase from the then current $1.8 million to $3.5 million by 1852 (1842 As. Doe. 18, p. 9). Using the latter figure as the anticipated revenue for the enlargements, and with due regard for the higher costs of operations, one can set the following comparisons among alternate canal prisms based on data in the record.

Prism (width X depth) in feet 40 x 4 50 x 5 60 x 6 70 x 7

Annual capacity million tons 1.5 3 8 12

Enlargement cost million dollars 0 13 32 48

Amortization 25 years at 6% million dollars 0 1 2.4 3.7

Operations and maintenance million dollars .35 .55 .75 1

Total annual cost million dollars .35 1.55 3.15 4.7

Anticipated revenue million dollars 3 3.5 3.5 3.5

Net revenue million dollars 2.65 1.95 .35 -1.2

The optimal economics would have been to retain the 40 x 4-foot canal. But optimism, one may judge, led to a decision to build a canal at the upper level of solvency-not the optimal net revenue, but the largest canal for which projected revenues would defray the costs. The results in the table may explain the decision to build a 70 x 7-foot canal.

This analysis omits consideration of the "secondary" benefits - that is those accruing to shippers and others benefiting from lower costs of transport. At a toll of $2 per ton (about 1 cent per ton mile) anal revenues captured only a small part of the benefits. Before the canal was built, wagon transport in the region ran about 20 to 70 cents per ton mile (Goodrich and others, 1961, p. 227-228). It was probably this latter point that showed in the Political rhetoric of the day, but could not be included in any optimization scheme. The canal lost this competitive margin as railroads came into operation. Revenue from tolls averaged about $1.75 per ton until about 1850 ,it decreased to $1.17 by 1860, and to $0.75 in 1870. Tolls were Eliminated in 1882.

As a result, the enlargement was not fiscally solvent as manifest in table I showing totals of revenues and costs (in millions of dollars) for the period through 1882 when tolls were eliminated from the State canal system, and when tonnage carried had reached its historic maximum.

The upshot seems to be that the 40 x 4 foot original canal was efficiently sized considering the revenue constraints; that of the enlargement was designed at the margin of anticipated revenue.

TABLE l.-Revenues and costs for the original and enlarged canal (millions of dollars)

Revenue Construction and improvements 1 Interest on canal debt Operation, maintenance and repairs Net

Original canal (1817-50) midway through enlargement 42 9 7.5 7.5 +18

Enlarged canal (1850-82) 79 40.5 18.5 21.8 -2

Total 121 49.5 26 29.3 +16

1 Includes extraordinary repairs Source: 1883 As. Doc. 4.

SUMMARY

The clear and present advantages of a canal to the general economy surmounted sectional opposition and a prediction of technological obsolescence. The weight of experience that inland rivers had proved unsatisfactory for dependable navigation led to a decision to build an independent canal which freed the location from the constraints of the river system and made possible a route directly to Lake Erie. With a connection to Lake Erie in prospect, it was natural enough to propose a canal profile as an inclined plane leading directly from Lake Erie at elevation 570 feet to the summit at Rome at elevation 420 feet and thence down the Mohawk Valley to the Hudson River at tide level. In concept, the scheme had two advantages; it would tap a copious supply of water for the canal which could serve as a source of water for future development of the State; it would avoid dependence of the canal for water supply upon the rivers along the route whose flow was expected to diminish as a result of deforestation. Far-beyond the engineering capacity of the times, the inclined-plane scheme was rejected in favor of a profile that followed the terrain, thus introducing a sag in the lake or middle division, and requiring a continuous development of new feeders from the rivers to maintain navigation.

Terrain problems were difficult; the Mohawk Valley was narrow and offered little space for locating a canal above flood levels. The sag in the lake or middle division required excavation through swamps and wet ground. The western division involved either rock excavation through the 60-foot Niagara escarpment or a route up Tonawanda Creek introducing another summit in the profile which, in addition to eliminating Lake Erie as a source of water, created a new requirement for water. Faced with this choice, the decision was made to cut through the 60-foot rock escarpment, the last and hardest part of the 8-year construction project. The flat slope of the canal from the lake to the crest of the escarpment greatly limited the amount of water that could be drawn from the lake.

The decision on the dimensions of the canal-chiefly width and depth-involved a balance between a fear of building too small and thus not achieving the economic advantages sought, and a fear of building too expensively. The constraints proved effective and for the first part of its history the revenues collected were sufficient to repay all costs. So great was the economic advantage of the canal at the time that the rising trend in traffic soon induced an enlargement of the canal, this time with a new but riskier objective-build as large as the projected trend in toll revenue, would finance. The expected revenues did not materialize.

SEQUEL

The two imaginative proposals considered when the canal was in the early planning but rejected in the design, were not entirely without merit as they surfaced again during the ensuing century. The first of these was that for building the canal as an "inclined plane" from Lake Erie to the Hudson. The scheme as it was presented by the Commissioners of 1811 make the lake a source of water for the State. The same proposal was made again in 1883 by State Engineer Silas Seymour in his proposals to improve the canal, and then by his successor, Elnathan Sweet (1885 As. Doc. 38, p, 11). Its omission in the major alteration begun in 1905 was criticized in 1927 (Thompson, 1927) when the need for additional water supply for New York City was becoming a problem. The Commissioners of 1811 also considered whether to build the canal to convey ships or barges. In view of the limited traffic expected, they recommended a canal built only for barge traffic, by inference leaving the option open when added traffic might make the construction of a ship canal feasible. Later, when the canal authorities sought means to prolong the usefulness of a canal, the proposal was made again by State Engineer Elnathan Sweet in 1885, without reference to his, predecessors of 1811, as follows:

"It is clear to me that the Erie canal, to become the permanent highway of this commerce, must have sufficient capacity to float the largest vessels navigating the great lakes (sic) from Lake Erie to the deep waters of the Hudson I I "' (1885 As. Doc. 38, p. 11.)

The proposal was reviewed again by the Federal Board of Engineers on Deep Waterways in 1900 (H. Doc. 149, 50th Cong., 2nd sess., Dec. 7, 1900) and, still more recently, a news item in Civil Engineering for August 1974 states that the Great Lakes Commission is looking into the possibility of rebuilding the Erie Canal for ocean shipping and, for the same reasons the original canal was built, as a competitive alternative to the St. Lawrence route through Canada.

As will be described in the next section, water power competed with navigation for the water supply of the canal throughout the 19th century. All of these water powers were for direct mechanical drive at the mills. In a later time, when hydroelectricity became feasible, the potential for power generation was enhanced. It was then considered to combine a new and enlarged ship canal with power generation as a built-in facility, rather than a byproduct. In fact, Rafter (1905, p. 821) refers to a scheme for a "Great Eastern Canal" that would roughly occupy the route of the Erie on a graded profile (inclined plane) from Lake Erie to Schenectady, where a dam across the Mohawk would impound water to a high enough level that a direct diversion may be made into Tesopus and Rondout Creeks, where further impoundments would divert water into Wallkill River and thence into the Delaware River, which would be dammed at Easton, and leading by other impoundments would link together the Erie Canal, the Delaware, the Susquehanna, the Potomac, and finally the James River, with waterway connections to New York, Washington, Baltimore, Richmond, Philadelphia, Trenton, and other cities in the region. But the canal was also to be the vast headrace of a hydroelectric scheme to develop 15 million firm horsepower. The idea is an early precursor of the recently proposed NAWAPA scheme (Sewell and others, 1967) proposed for building dams and canals of continental, cope for water storage and water transport from Yukon Territory and British Columbia of western Canada to the central part of that country, to the western United States and to northern Mexico.

| Environmental Aspects of the Erie Canal | Operating the Canal |