It was the construction in the 1830's of lateral canals, such as the Genesee and the Chenango Canal, leading off to the south over high summit levels, which brought about two innovations in American Water Development- the construction of artificial impoundments to serve as reservoirs for water-supply, and the collection and analyses of hydrologic acid, albeit on a spot or ad hoc rather than a systematic basis. These reservoirs were needed to augment the low flows of the small streams at the summit levels which were over 1000 feet in elevation. Having definite measurable capacity, a reservoir would naturally enough arouse questions as to whether the entering stream flow was enough to fill it. These reservoirs were maintained through 1905 as part of the feeders to the Erie Canal. Kingsley Brook dam was washed out by a flood in April 1843. It was repaired and restored to service. Inconsistencies, in the published account make it impossible to judge the adequacy of these reservoirs, but the matter must have troubled Jervis because in 1835 he instituted what is believed to be the first set of observations of rainfall and runoff in the United States. In later years he described this work.
The estimates of water supply would need to be based on precipitation, for which Jervis' measurements were available. The question was how? Talcott expressed disagreement with Jervis' use of estimating runoff from precipitation by multiplying the letter by a coefficient, stating (p. 54) "Nor does it follow, either from the experiments quoted above, or from any we have ever seen, that there exists any constant ratio between the fall and drainage." It was his judgment that the difference between precipitation and runoff, which is taken as the "demands," is the more uniform, citing as evidence the data for Eaton and Madison Brooks (the latter as corrected), as follows:
June to December, differences between precipitation and runoff
Eaton Brook 12.05 inches Madison Brook 13.30 inches
He then noted (p. 54) "The demands (that is, the differences between precipitation and runoff, the evapotranspiration in modern terms) are much more uniform than the falling water and since they must be supplied first, it is evident that the drainage (that is, the runoff) will depend altogether upon the excess of fall (that is, the precipitation) over what is required for their supply." He added (p. 55) "We do not claim that these facts are sufficient to establish any general rule, although they seem to confirm the views herein taken of this question."
This is a classic argument, of some importance, because the estimate of the runoff in a dry year would be proportional to the precipitation according to the coefficient method; the estimate of the runoff calculated as a residual would be much less.
Talcott then calculated the annual drainage (= runoff) in the Genesee Valley during a dry year by the residual method as follows (p. 56)
Precipitation. ("falling water") ---------------- 28.36 inches Natural Consumption - -------------------------- 21.63 inches Drainage = Difference ---------------------- 6.73 inches
Examination of long-term records of Stream flow in the Genesee River drainage area indicates that the annual lows are quite variable, ranging from 5.2 inches for Canadice Lake outlet near Hemlock to about 10 inches for the Genesee River itself.
A few years later, Henry Tracy (1850, Sen. Doc. 40, p. 17) assembled the following available data on annual rainfall and runoff:
Rain & Water Evaporation Ratio of Drainage Snow that ran from surface (Inches) off (in.) of ground(in.)
1835 Madison Brook------------- 35.26 15.83 19.43 0.449
1837 Long Pond (Mass.)--------- 26.65 11.70 14.95 .439
1838 Long Pond (Mass.)--------- 38.11 16.62 21.49 .436
which tended to show the ratio to be more uniform than the difference between rainfall and runoff. Tracy's estimate of inflow to Hemlock Lake, south of Rochester, which it was proposed to control as a storage reservoir, employs the coefficient method (1850 Sen. Doc. 40, p. 29). Hemlock Lake, 1,544 acres surface area, 29,525 acres of contributing area Least annual fall of rain, 22 inches Ratio of drainage, 0.4, or 8.8 inches inflow to lake=943 million cubic feet Loss from the lake (1,544 acres) Total evaporation per year= 49 inches offset by rainfall on the lake of 22 inches making net loss of 27 inches. Hence annual loss from lake= 151 million cubic feet giving 943-151 = 792 million cubic feet as the amount available "in the year of least fall of rain."
Unfortunately, probably following British practice of the times, the idea was left there-that the yield of streams could be accurately inferred from rainfall, whether by the ratio or residual method. For Rafter (1897, p. 174), noting the lack of attention by canal authorities to measuring the flow of streams, added that Jervis' "results, while covering too short a period of time to furnish safe averages, have still been, as regards the yield of small streams in the State, the handy stock in trade of the New York Canal Department from that day to this." The writer recalls, an interview in 1935 with Frederick Stuart Greene, then Superintendent of Public Works, who said that he still favored the estimation of Stream flow from rainfall records. He emphasized the difference in cost between rainfall and flow measurements. Talcott's note of inadequate time and resources to collect hydrologic data was perhaps the first appearance of an apology that has been written again and again in project reports in this country for over a century and still appears today.
In view of the high cost of error in modern public-works practice, major reliance is placed upon a national, systematic network of continuous records of Stream flow. The controversy regarding the estimate of annual runoff by the coefficient or the residual method is now moot. Precipitation data are used to extend records of daily flow or to synthesize the daily flow of ungaged streams by the use of complex hydrologic formulas or models that depend for their calibration on flow data from the systematic network.
HYDRAULIC COMPUTATIONS
The hydraulic difficulties that developed during the operations of the original canal were to be corrected in the enlargements that were begun in 1836. By that time formulas for the velocity of water in open channels became generally known from European research, which were then applied in canal design, especially as greater quantities of water had to be conveyed. The hydraulic problem was the following. To design a channel section and profile that will convey a flow of water sufficient to maintain a prescribed navigable depth, with a known-rate of channel loss per mile and to supply lockages, and not exceed a maximum mean velocity of 1 mile per hour. This is a standard hydraulic problem, easily solvable with a number of possible combinations of depth, area, and slope, provided the channel is stable-that is, the channel does not react to the flows carried. The prescription of a limiting velocity of 1 mile per hour assures that this condition of stable bed and bank is met. 0.W. Childs (1848, As - Doc. 16) and Henry Tracy (1850, Sen. Doc. 41) carried out these analyses for the western division. A hydraulic problem was created in that division by the desire to obtain as much water as possible from Lake Erie to supply the canal all the way to the sag point in the profile at the Montezuma marshes. The key to the problem was a velocity formula, and at the time, the best known (Rouse and Ince, 1957) was the Prony formula. av+bv,=DS where v = mean velocity, in feet per second, D =hydraulic mean depth, in feet, S= slope of the water surface, and a and b are constants determined by experiment. According to Childs (p. 150), Prony gave the following values for the constants: a= 0.0000444499 b = .0000942772
and as defined later by Eytelwein:
a = 0.0000242651 b = .0001114155
(The 6-figure precision of the constants probably came about in the conversion of the formula from metric to English units. The general concept of reporting calculations with regard to their real accuracy did not seem to be part of the mid-19th century engineering practice.)
One may note that because of the decrease in the a term and an increase in the b term, the later Eytelwein version approaches the Chezy formula in which the product DS varies only as vl, em- bodied in the Manning formula of present practice.
lt is interesting to compare the Prony-Eytelwein formula with the Manning formula now in common practice:
v = 1.5 D 2/3 S ´ n
where v, D, and S are as before and n is a coefficient of roughness. (The constant 1.5 arises because of the conversion of the formula from metric to English units, and some relic of false precision is found even today where it is usually reported as 1.486, the cube root of the number of feet in a meter).
For slopes and depth encountered in the Erie Canal, the Prony Eytelwein formula gives about the same results as the Manning formula with a value of the roughness factor n of about 0.025, not far above what is now considered applicable to a smooth artificial channel in good condition.
Measurements in the improved outlet channel of Onondaga Lake were used (1848, As - Doc. 16, p. 172) as a check. The Prony formula (with Eytelwein coefficients) gave a velocity of 1.82 feet per second. The velocity as measured was reported as follows:
Surface Velocity 1.45 Feet per second Reduced to the Mean 1.13 Feet per second
The difference was to be accounted for by the irregular and rough section of the outlet.
Based on the Prony formula (with Eytelwein coefficients) Tracy (1850, Sen. Doc. 41, p. 10) presented four plans for the canal between Lockport and Rochester.
1. Constant width, vary depth and slope so that mean velocity = 0.5 miles per hour. 2. Constant depth (7 ft), vary width and slope. 3. Depth at 8 ft for first 40 mi west of Lockport, thence gradually decreasing to Rochester. 4. Level bottom to have such depth at Lockport as to give sufficient slope to the water surface.
(In each case, flows were to be as follows: Pendleton, 31,000 cfm; Lockport, 29,600 cfm; and Rochester, 17,000 cfm; depth at Rochester was kept at 7 ft.)
Tracy carried out a series of calculations based on these conditions, and for a low and high water level. He favored plan number 4 mainly because the level bottom would not exclude the possibility of supplying the canal from the Genesee River in an emergency.
The channel as built on the 62.5 mile pound from Lockport to Rochester (Searles, 1877) conformed more nearly to Tracy's plan number 2, with depth nearly uniform.
Section below Lockport combines
Surface width ------------ 96 feet Depth ------------------- 8 feet Slope -------------------- .068 foot per mile
Section midway
Surface width ------------ 87 feet Depth ------------------- 7.6 feet Slope -------------------- .050 foot per mile
Section at Rochester
Surface width ------------ 70 feet Depth ------------------- 7.8 feet Slope -------------------- .028 foot per mile
The standard section east of Rochester at the time was 70 feet wide by 7 feet deep.
Because of the continuous leakage from the canal, the flow decreases and therefore the hydraulic capacity must decrease. Of the indicated 50 percent decrease in hydraulic capacity, from Lockport to Rochester, one-third was made by reduction in width and two-thirds by reduction in slope. As in natural rivers, the profile was concave.
Lake Erie Levels.-Variations in the level of Lake Erie would necessarily affect the flow into the canal. Considering the low gradient from the lake to the "mountain ridge" (total fall, at average -lake levels, of 4 feet) even small changes in level and therefore in slope would significantly affect the flow down the canal.
The Commissioners of 1811 (p. 21) suggested "that it is impossible there should ever be a considerable variation in the surface of Niagara River. - - - Indeed, we know from experience, that a greater difference of elevation at the mouth of Lake Erie is occasioned by a change of wind, than by any variation of seasons."
They had ascribed an elevation of 525 feet to Lake Erie; those of 1816 made it 564.85 feet; more accurate levels run during the construction placed its level very close to the presently adopted mean level of about 572.5 feet. Nevertheless, the lake was soon known to change its levels, seasonally and from wind. These levels are also subject to secular changes owing to climatic fluc- tuations over the contributory area of the Great Lakes. Isostatic readjustment taking place following retreat of the glaciers (Flint, 1957, p. 250) would not affect the canal, first, because it emerges on the lake near its outlet; thus the canal and lake levels would maintain their same relative position, and, secondly, because it is small in the Lake Erie region (about I mm per year).
Continuous records of lake levels are available only since 1860, but Horton and Grunsky (1927, p. 276) show a historic high of 575.1 feet in June of 1838. The levels had receded considerably by 1841 when the Commissioners of the canal gave the low levels some attention because of the resulting difficulty. They reported (1842 As. Doc,. 24, p. 53, 60) that the level reached in 1838 was 5 feet 1 inch higher than in November 1820, when the lake was at its lowest known stage. This would make the 1820 level 570 feet. Commissioners go on to say that the level in November 1841 was only 9 inches higher than in 1820, or 570.7 feet. At that time the lake was lower than the level of Tonawanda Creek and water in the canal flowed toward the lake rather than out of it. Water in the western division of the canal was again low in the 1890's. (1892 As. Doe. 15, p. 133.)
MEASUREMENT OF WATER FLOW
The Erie Canal was begun on faith that water supply would be sufficient for the need. There was, of course, superficial justification for this faith in the generality-the Mohawk River and Lake Erie seemed to offer quantities of water far in excess of that needed and, in addition the canal crossed other streams that could be used for incremental supplies along the way. Accordingly, few measurements of flow were made before construction began. However, soon after sections of the canal were filled, measurements of flow became commonplace as problems of leakage and water supplies became manifest, particularly in the lateral canals which were built in terrains and at a time less favorable than that of the Erie Canal. The streams available for water supply were smaller and had become subject to claims b mills which had developed them for waterpower.
The first set of measurements was used to resolve the critical question of the water supply available for the proposed Tonawanda Creek summit-the Commissioners of 1816 stated that the flow of 10 streams "gauged with great care," totaled 253,435, cubic feet per hour, then judged to be sufficient. Additional measurements (no data given) were made in 1820 (1820-21 As. Jour., 44th sess., p. 866), as were measurements of leakage of water from the middle section of the canal that had already been filled. As already mentioned, these measurements led to the abandonment of the Tonawanda route in favor of a route to the north. Leakage from the canal was an unanticipated problem that led to measurements of flow (leakage being the difference in rates of flow over a suitably long reach of the canal), and it was natural enough that reference be made to the results of these measurements in the design of the new canals. In his report on the proposed Chenango Canal, D. S. Bates (1830 As. Doc. 47, p. 31) referred to his measurements of canal leakage made in 1824 in the several reaches of the western division of the Erie Canal (see table I for results). John B. Jervis' report on the same proposed Chenango Canal (1834 As - Doc. 55, p. 54) referred to his measurements of canal leakage in the eastern division, particularly the section from Amsterdam to Schenectady.
Bates (1830) used floats to measure flows in the canal, but details are not known as field notes are not available. A report by D. S. Bates, quoted by Sherman (1932) on flow measurements made in Ohio in 1823 in connection with canal proposals inspired by the Erie, states that he used weirs with rectangular notches on the smaller creeks and floats in the larger streams. In that report Bates mentions the results of his measurements in New York as follows (Sherman, 1932, p. 157) : "Quantity of water expended in the New York canals has been found to be 100 cubic feet per minute 11 per mile in ordinary cases," a result based on his gagings in 1823 and 1824 (Bates, in 1830 As - Doc. 47, p. 31).
The Chenango Canal, built in 1834-36, one of the new canals that entailed serious questions about water supply, followed the valley of the Oriskany Creek from Utica, reaching a summit level at an elevation of 1,050 feet, where it crossed the saddle between Oriskany Creek and the Chenango River. The canal then followed down the valley of that river to Binghamton. Uncertainties about the water supply came about because of the high summit and because water could not be taken from Oriskany Creek owing to objections from the established mill owners. Hence, all water north of the summit had to be supplied from the streams at the summit, making it necessary for the promoters of the canal to store the high runoff of the spring season for use during the low water period. A scheme of storage reservoirs was proposed for six of the small streams that otherwise flowed south to the Chenango River. Jervis, the engineer, arranged for the measurements of precipitation and of the flow of Eaton and Madison Brooks that were described in the section on "Hydrologic Data and Analyses."
Assessment of the adequacy of streams to be used as feeders was carried out by ad hoc spot gaging during the summer or autumn season of low flow. Thus, in 1841 when "streams were unusually low," 0. W. Childs (1843 As. Doc. 25, p. 40-41) was directed "to make accurate gauges of streams to be used as feeders for the enlarged canal, from the lock at Geddes to the Seneca River." Because the results indicated insufficient flows it was decided to lower the bed of Skaneateles Lake outlet by 5 feet -in order that the lake could be drawn down by that amount through a gated bulkhead.
Jervis (1834 As - Doc. 55, p. 56) observed that 1833 was too wet a season to gauge for low flow determinations and "sluices were put in all of the streams to be gauged and every opportunity was improved to procure a measurement of the lowest water." No account was given of what the "sluices" were like. He expressed a correct view; the "lowest gauge (i.e. measurement) was the one in which there was the best evidence of the regular flow of the stream" or, in modern terms, when the streams had receded to base flow. He then added "various opinions were expressed in relation to the comparative condition of the streams when gauged and at the lowest state of the water. It is believed, however, that a deduction of 25 percent from the gauged quantity would not essentially vary from the minimum flow."
Little was published concerning methods of making measurements of flow, even though in contrast to the long-established methods of land surveys, flow measurement was still exploratory and uncertain. Timed surface floats were used. The following description of flow measurements is contained in a hearing on claims made by mill owners for damages caused by diversion of water of the Genesee River to canal purposes (1854 As. Doc. 63, P 84). The testimony was given by Daniel Marsh, a civil engineer previously employed by the State, then appearing for t claimants.
I found the width and depth, and multiplied these into the velocity; to find the velocity I Used a pine stick of lath 3 feet long to float upon the surface; I tried it within 18 inches of each side and so quite across the stream at about equal distances of each other * * * * The floats were timed over a reach, an assistant dropping t floats on signal at the upper end. Depths were measured at several places at equal distances. Marsh noted that "the velocity above stated is the surface velocity," and then, after some r marks that seem unclear, he added, "so I deducted one-tenth fro the surface velocity."
In the measurements of the flow of the outlet of Lake Ononda (1848 As - Doc. 16, p. 172) a coefficient of 0.78 apparently w used to reduce surface velocities to the mean. The shallow stream probably discouraged if not precluded the design of deep Flow that would move at the mean velocity in lieu of surface floats The use of current meters and systematic records lay in future. (1888 As. Doc. 25, p. 25; Rafter, 1889.)
FEEDERS, LOCKS, AND STREAM CROSSINGS
Feeders.-In 1825, when the canal as a whole was first in operation, water supply was obtained from (east to west)
Mohawk River at Johnsville (Minden)
Schoharie Creek
Mohawk River at Little Falls
Steeles Creek
Oriskany Creek
Mohawk River at Rome (Rome summit)
Wood Creek
Skaneateles Creek (Jordan summit)
Genesee River
Oak Orchard Creek (including diversion from upper Tonawanda (,reek)
Tonawanda Creek (lower)
Lake Erie
By 1862, the list of feeders or sources of supply was as follows (1863 As. Doc. 8, p. 442) :
Feeder Date Supply (cubic feet per minute)
Rexford Flats---------------------------------------- 1844 10,979 Schoharie Creek------------------------------------ 1845 6,800 Rocky rift------------------------------------------- 1856 10,602 Little Falls------------------------------------------ 1843 12,643 Ilion Creek------------------------------------------ 1838 800 Chenango Canal------------------------------------ 1836 750 Butts Creek----------------------------------------- 1838 1,400 Mohawk Feeder ay\t Rome------------------------ 1858 10,979 Black River Canal at Rome------------------------ ------ 708 Oneida Creek----------------------------------------- 1835 1500 Cowassolan Creek----------------------------------- 1858 320 Erieville Reservoir---------------------------------- 1850 2,130 Chittenango Creek Feeder------------------------- 1840 250 Cazenovia Lake Reservoir------------------------- 1857 2,631 De Ruyter Reservoir-------------------------------- 1863 3,972 Limestone Creek------------------------------------- 1852 210 Orville (Butternut Creek Feeder)------------------- 1858 450 Camillus Feeder--------------------------------------- 1843 1,500 Skaneateles Lake Reservoir------------------------- 1844 7,520 Genesee River Feeder--------------------------------- 1826 350 Genesee Valley Canal--------------------------------- 1842 861 Oak Orchard Creek------------------------------------ 1840 1400 Lake Erie, Buffalo------------------------------------- 1856 35,000
The list includes the inflow from the later-built Genesee and the Chenango Canals that led off to the south, reaching high summits en route. The lateral canals were continued as feeders after they were closed to traffic.
Lift Locks on the Original Erie Canal 2 22 ---------- Up 2 22 ---------- Up 7 56 8 Up 2 15 ---------- Up 3 26 ---------- Up 4 32 ---------- Up 1 7 20 Up 1 35 ---------- Up 5 40 34 Schenectady (elev 225) Up 1 4 ---------- Up 1 6 ---------- Up 1 7 ---------- Up 1 6 60 Up
1 7 ---------- Up 1 8 ---------- Up 1 8 ---------- Up 1 8 ---------- Up 5 40 81 Up 1 8 ---------- Up 1 9 ---------- Up 5 40 ---------- Up Long Level ---- 110 ---- -- 165 Long Level Up (ele 414) ---- 2 20 ---------- Down 1 6 170 Syracuse Down Sag (elev 382) 6 ---------- up 11 175-182 jordan Up 11 ---------- Down 9 ---------- Down 9 ---------- Down 7 219 Montezuma Down. 8 ---------- Up 6 ---------- Up 7 ---------- Up 6 ---------- Up 1 15 ---------- Up 3 24 ---------- Up 2 20 235 Up 8 ---------- Up 5 37 ---------- Up -- -- 270 (Rochester) ---- -- -- 325 (elev 490) ----
5 60 Lockport elev 560 Up 81 604 Total up locklaage 62 Total Down Lockage 676 Total Locklage 542 Net
Stream Crossings.- Hundred, of streams had to be crossed either by bridging or fording along the route of the canal, and because this was to be an independent canal, some form of bridging in an aqueduct or culvert was to be preferred to fording the stream at grade. Aqueducts were viewed as structures for carrying the canal over a natural waterway; culverts, on the other hand, were viewed as structures that conveyed watercourses under the canal embankment. Aqueducts were bridges-the canal was carried in the material of which the bridge was built, stone or timber. Most aqueducts on the original canal were timbered' structures upon masonry piers, but a few were major stone arch bridges, as at Rochester. They were narrow, only as wide as a lock, that is, only wide enough for a single boat.
Fording where the canal crossed the stream at grade was used in those cases, where an aqueduct appeared too expensive. A low dam was built downstream from the' crossing impounding water to a navigable depth and trestles were built across the pool to carry a towpath. Guard gates or locks were usually installed along the canal on either side of the pool to isolate the canal in the event of flood in the stream. Because of its economy (the dam was usually of rock and brush construction,), the pool crossing method was advocated and adopted in several places in the original plan, for example, Tonawanda Creek, Oriskany Creek, and Schoharie River. One of the advantages was that the stream could 'serve also as a feeder but, being uncontrollable, the flow could work either way. In one case, Oriskany Creek, which at first formed part of the navigation, was disconnected from it by an aqueduct as early as 1822, because the connection enabled the mills and factories to draw water from the canal whenever the creek failed to give them their accustomed supply (1823 As. Jour. 46th sess., p. 503). Violating the primary rule of independence, such crossings other than on the Tonawanda were eliminated by converting the crossing to aqueducts and thus maintaining the independence, of the canal from the vagaries of river floods. The 12-mile reach on Tonawanda Creek remained the only natural channel on the canal after the improvements of 1835-62.
Some small streams, including nearly all those dry at the time of construction were admitted directly into the canal (Jervis, 1877, P. 55). Waste weirs were provided at intervals to permit the overflow of flood waters as well as the excessive flows diverted from the feeders.
CANAL CROSS SECTION
The trapezoidal cross section specified by the Canal Commissioners of 1816, as shown in -figure 7, is such that about 3 feet of excavation or cut would produce sufficient material or fill for the berm and towpath embankments. The banks were constructed by scrapers drawn by horses and cattle (oxen.) which, together with the hundreds of laborers, served to compact the fill. The side slopes of 1 ´, to 1 followed practice "often adopted in England" (1820 As - Jour. 43rd sess., p. 451).
There was no reason to expect erosion by flowing water since velocities were less than 1 foot per second (0-7 mile per hour). However, wave wash produced by moving boats was found to erode the earthen banks. This unanticipated erosion seriously narrowed the towpath and shoaled the bed. The response to this serious problem was of two sorts-administrative and engineering. (1) A speed limit of 4 mph (6 ft/sec) was enacted in 1822, and (2) the canal banks at the water surface were faced with stone, a program begun in 1824 (1825 Sen. Jour. 48th sess., p. 278) and continued from year to year throughout the length of the canal and adopted at the outset of the enlargement that was begun in 1835.
A side-hill location above the valley bottoms was preferred; and the topography was such that this was usually along a north facing slope (see fig. 6),so that to the south, the land was nearly always higher, and this became the berm bank. The fill embankment was used as the towpath. Each bank had its characteristic problem. The berm bank was subject to erosion and soil wash from the higher ground to the south; the towpath bank to leakage. The section itself was above the regional water table to which the water in the canal drained. Small runners, normally dry, were allowed to fall directly into the canal and discharged water together with some sediment during wet weather. Also, the cut-bank intercepted some perched or wet-weather springs.
It is of some interest to note how conservative the width depth ratio remained. It was 10 to 1 in the two stages of the 19th century, and is 11 to 1 in those parts of the present canal that are in earthcut section. With barges customarily built in width-depth ratios of 3 or 4 to 1, it would me-an that a canal built to permit passing and with side slopes of l ´ to 1 would have a width-depth ratio of about 10 to 1 and so it remains.
NEW YORK CANALS IN THE 19th CENTURY
The early success of the Erie Canal induced not only its own enlargement (see table 7), but a canal building epoch in the country after that kind of transport facility had become obsolete. Several additional lines were built in New York, the following being those listed as in operation in 1853 in the State Engineers Report on the canals (1854 Sen. Doc. 60, p. 13) :
The main trunk of this system is the Erie Canal, occupying the valley of the Mohawk and the southern slopes of Lake Ontario, running east and west nearly through the center of the state, and connecting the chain of western lakes with the navigable waters of the Hudson.
Table 7:- Statistics of the original and enlargements of the Erie Canal
Original First En- Barge Largement Canal
Dates of Work--------------------------------------------- 1817-25 1836-42 1905-17 Canal Prism: Width at water surface (ft)------------ 40 70 133 Width at bottom (ft)- 28 521/2- 56 75-94 Depth (ft) 4 7 12 Length (Miles)---------------------------------------------- 363 351 340 Locks: Number-------------------------------------------- 83 72 34 Total Lockage (ft)-------------------------------- 676 655 680 Length (ft)----------------------------------------- 90 110 300 Width (ft)------------------------------------------ 15 18 44.5 Depth on sill (ft)---------------------------------- 4 7 12 Boats: Length (ft)------------------------------------------ 61;75 98 250 Beam (Ft)----------------------------------------- 7;12 17.5 40 Draft (Ft)------------------------------------------ 30;75 240 2000
Annual Traffic Capacity (million tons) 1.4 8 20
Of these New York canals, only the Erie, Oswego, Cayuga Seneca, and Champlain are now in operation as the New York State Barge Canal System.
The tonnage carried continued to increase until the 1880's after which it slowly decreased, despite the enlargement and the elimination of tolls in 1882. In order to stem the loss of traffic to the railroads, further efforts were made to improve the canal as, for example, by lengthening some locks so as to service two barges at one time; but the impact was gone. On completion of the canals in 1825, a boat carrying 30 tons pulled by a single horse or mule at the rate of 21/2 mph was a great improvement over a wagon hauled by a team carrying 1 ton at a speed, if that is the word, of I mile per hour, provided the road was dry. The improvements could not duplicate that impact. In 1882 the State auditor reported that the canals were viewed as antiquated and their continuance a subject for ridicule. (1883 As. Doc. 4, p. 16-17.) Less than 2 million tons were carried on the Erie division of the State Barge Canal system in 1972, about as much as was carried by the horse and mule drawn barges in the 1850's. With the decline in commercial use, the assets of the barge canal for recreational boating have been increasingly recognized, albeit reluctantly at first. In 1905, for example, it was reported (Whitford, 1906, p. 405) that pleasure craft had become a "problem"-over 1,000 permits for such boats had been granted, nearly twice the number of barges then operating for carrying freight. The attitude was then that canal cruising constituted a necessary evil-necessary to maintain the principle that the canal was to be open freely; evil because overpowered pleasure craft tended to disobey the 4-mile speed limit ( now 6 to 10 mph) with attendant bank wash, and required rather frequent opening of busy street bridges. Recreational boating is now (1974) actively encouraged. The state currently reports 100,000 lockages of pleasure craft annually, compared with about 30,000 commercial barges.