¹ One can demonstrate a clear continuity with distinguished work-e.g., that by Schumpeter, Ripley, Healy, Kirkland, Taylor, and Goodrich. But a new sharpness and direction have clearly been given by Fishlow's, AlbertAmerican Railroads (Cambridge: Harvard University Press, 1965)Google Scholar and Fogel's, RobertRailroads (Baltimore: Johns Hopkins Press, 1964)Google Scholar . Cost-benefit studies (e.g., those by Eckstein, Meyer, Peck) and research in transport problems of underdeveloped nations (e.g., by Fromm, Heymann, Wilson) have complemented such work in pointing to this direction.

² Fogel, pp. 47, 223. According to Fishlow (pp. 52, 54, 57) the all-product figure was “at least 15 per cent” for 1890.

³ The monetary costs of interregional transport of farm products are computed by Fogel at $88 million (p. 42). Add to that his figure for imputed value of nonmone-tary advantages-some $121 million (p. 47). Together they amount to 1.7 per cent of his $1.2 billion income estimate for 1890.

⁴ Viner, Jacob, “Stability and Progress: The Poorer Countries' Problem,” in Hague, D. C., ed., Stability ana Progress in the World Economy (London: Macmillan 1958), p. 57.Google Scholar

⁵ Tredgold, Thomas, A Practical Treatise on Rail-roads and Carriages (London, 1835), p. 414Google Scholar , viewed the risk of expending capital on railways as equal to that for canals.

⁶ Laws of the State of New York in Relation to the Erie and Champlain Canals… (Albany, 1825), II, 376Google Scholar ; Ohio State Archaeological and Historical Society, History of the Ohio Canals (1905), p. 69Google Scholar.

⁷ “There is an economy of at least 3 per cent per annum in employment of money borrowed on the credit of the State, at less than 5 per cent interest per annum, instead of investing in fixed property the active money of active business men, to whom active money is worth at least 8 per cent a year”; Proceedings of the Western RailRoad Corporation (Boston, 1838), p. 9Google Scholar . This, and most of the railroad reports cited in later footnotes, may be found at the Bureau of Railway Economics in Washington, D.C.

⁸ Fogel applies as per cent interest rate to $293 millions of government waterway subsidy, allocable to 24, 791 million ton miles (pp. 42, 46–47), or .0007 per ton mile. But the New York Report of the Committee on Canals, 1899, pp. 151-52, 181, indicates that by 1882, $79 million had been spent to build New York State canals. Allocation to the 940 million ton-miles they carried in 1890, also at 6 per cent, gives .0053 per ton mile—or over seven times Fogel's allowance. Even at this, a further allowance is requisite for repairs and maintenance costs on waterways-no less a subsidy, albeit not a capital charge.

⁹ Fogel, p. 82.

¹⁰ To $49 million for monetary costs one must add $121 million of nonmonetary costs-the $170 million then contrasting with the actual $88 million payments for actual 1890 transportation of interregional farm products (Fogel, pp. 42, 47). Since actual transportation involved shipping over one fourth of all grain by water, or by water and rail, the ratio of rail advantage per se would appear to have been even greater by this measure.

¹¹ The above data relate to freight. Early cost estimates treat fifteen passengers and their baggage as equivalent to one ton of freight. The ratio of advantage for railroad versus canal would probably have been greater for passengers than for freight, however, since cubage requirements were greater for passengers than for an equivalent mass of coal, wheat, or lumber.

¹² McDougall, Congressman, Hunt's Merchants Magazine (1854), pp. 633–35Google Scholar . Fogel's estimate, cited in n.2, suggests the same conclusion.

¹³ Dupuit, J., “On the Measurement of the Utility of Public Works” (1844), reprinted in International Economic Papers, II (1952), p. 94.Google Scholar

¹⁴ We are not concerned here with the very substantial advantage to the individual entrepreneur of delivery certainty. For him, to preclude the “fear of not being able o t sell the inventory by the end of the finite planning horizon” is worth a tangible premium in more than one industry. The phrase is that of Nevins, Arthur, “Some Effects of Uncertainty,” Quarterly Journal of Economics, LXXX (02. 1966), p. 82Google Scholar . Thus, cotton was frequently carted by road, rather than canal, at much higher rates, “because speed and certainty as to delivery are of the first importance” ; Booth, Henry, An Account of the Liverpool and Manchester Railway (Philadelphia, 1831), p. 128Google Scholar.

¹⁵ This adjustment is probably excessive. In 1840 the Boston and Worcester reported only three days lost because of snow, ice, or other interruptions; Annual Reports of the Rail-Road Corporations in the State of Massachusetts for 1840 (Boston, 1841), p. 15Google Scholar ; Tredgold, p. 142, stipulates 312 working days in a year.

That freight trains did not operate on Sundays is suggested by the annual reports of the Boston and Worcester (p. 64) and Charlestown Branch (p. 79) in Annual Reports of the Rail-Road Corporations…Massachusetts…1843 (Boston. 1844)Google Scholar . However , The Report upon the Merits of the Great Western Railroad, Canada West, by a Committee of Its American Friends (Boston, 1851), p. 13Google Scholar , estimates its potential operation at 360 days a year. In Massachusetts Senate, No. 98, Report and Bill on the Petition of the Seekonk Rail-Road Branch Company, April 18, 1838, p. 64, the wages of brakemen are estimated at $365. We take this to imply a 365-day work year, the prevailing wage often being estimated at $1 a day.

¹⁶ Contrasting transit times, the engineer for the New York Commission on Steam on the Canals (second Annual Report [Feb. 1873]), p. 139, stipulated 313 days annual running time for trains—clearly subtracting 52 Sundays from the year's total

¹⁷ As found in the following sources: for the , Schuylkill, Niles (03 11, 1843), LXIV, 24Google Scholar ; also Hunt's (June 1843), VIII, 548; for , Erie, Annual Report of the Commissioners of the State of New York (1871), p. 221Google Scholar , and Andrews, Frank, Grain Movement in the Great Lakes Region, U.S. Department of Agriculture Bulletin No 81 (1910), p. 58Google Scholar ; Sault Ste Marie, U.S. Congress, Senate, 43d Cong., 1st sess., S.R 307, Part 1, Report on Transportation to the Seaboard, Appendix, p. 203; Ste Mary's Falls and Welland, Andrews; Delaware and Hudson, Illinois and Michican, Welland, and Rideau, U.S. Congress, House, 54th Cong., 2d sess., H.R. 192 , Report of the U.S. Deep Waterways Commission (1897), pp. 201 ffGoogle Scholar.

¹⁸ Report of the [Pennsylvania] Canal Commissioners… Dec. 7, 1837 (Harris burg, 1837), p. 26Google Scholar , Table 1, p. 87. The period closed for ice averaged 112 days, ranging from 88 days on French Creek to 123 days on the West Branch Division.

¹⁹ Annual Report of the Canal Commissioners of New York (Albany, 1871), pp 8–9Google Scholar , 20, 31, 87, 222. Floods in April 1870 did not suspend navigation, but only be cause navigation did not even begin until May 10.

²⁰ U.S. Congress, House, 20th Cong., 1st sess., H.R. 141, Chesapeake and Ohio Canal (1828), p. 112Google Scholar ; Bernard and Totten, estimating canal costs, “assumed the day as of 12 hours length, whereas, in Summer, they extend to 14 and 16 hours.” We use the same figure for canals and railroads, so that the level per se need not be precise.

²¹ This is according to New York State Report 113 (Feb. 24, 1829), Report of Commissioners…Relative to…Railroad from…Boston to the Hudson River (Albany, 1829), p. 17Google Scholar . In 1832 the Board of Internal Improvement forecast a ½ mile rate from Georgetown to Pittsburgh (on the Chesapeake and Ohio Canal); U.S. Congress, House, 22d Cong., 1st sess., H.D. 101, Report on Steam Carriages (1832), p. 309Google Scholar.

²² Bogart, E. L., Internal Improvements in Ohio (London: Longmans, 1924), p. 27.Google Scholar

²³ Dame, Lorin L., “The Middlesex Canal,” Contributions of the Old Residents Historical Association, Lowell, Massachusetts, III, 277.Google Scholar

²⁴ Roberts, Christopher, The Middlesex Canal, 1793–1860 (Cambridge: Harvard University Press, 1938), pp. 147Google Scholar , 141. The twenty-year average, Boston to Concord, N. H., was five days up and four down.

²⁵ Quoted in Niles (Jan. 17, 1835), XLVII, 342. Sixty to seventy miles daily could be made with the aid of relays.

²⁶ Report of the Pennsylvania Canal Commissioners, Dec. 1831, quoted in Report on Steam Carriages, p. 244.

²⁷ General Convention of Agriculturists and Manufacturers and Others Friendly to the Encouragement and Support of the Domestic Industry of the United States, Monday, July 30, 1827 (n.p.), p. 41. This is the so-called Harrisburg Convention.

²⁸ Report of the Commissioners of the State of Massachusetts on the Route of Canals from Boston Harbour to Connecticut and Hudson Rivers (Boston, 1826).Google Scholar

²⁹ Report on Steam Carriages (cited n. 21), p. 170.

³⁰ Massachusetts Agricultural Repository (Boston, 1827), p. 77.Google Scholar

³¹ Quoted in Report on Steam Carriages, p. 244.

³² Derrick, Samuel M., Centennial History of South Carolina Rail-Road (1930), p. 48.Google Scholar

³³ Niles (Nov. 15, 1831), XLI, 198; B & O Annual Report (Baltimore, 1832), p. 120.Google Scholar

³⁴ Poussin, G. T., Chemins de Fer Americains (Paris, 1836), p. 206.Google Scholar

³⁵ Annual Report, Petersburg Railroad (Petersburg, 1841), p. 13.Google Scholar

³⁸ Bird, Henry, “Report,” in Annual Report of the Petersburg Rail-Road Company (Petersburg, 1836), p. 16Google Scholar . Bird reports a combined load of freight and passengers making 60 miles in five hours, including all stoppages-suggesting that a ten-mile average applied to this road as well, before its rails began giving way.

³⁷ Annual Reports of the Rail-Road Corporations in the State of Massachusetts for 1846 (Boston, 1847), pp. 7Google Scholar , 11, 15, 29, 107. “Average rate of speed adopted for freight transit including stops” was ten miles an hour for the major Western Rail Road Corporation, eight for Boston and Maine, twelve for Boston and Providence, fifteen for Connecticut River. The Annual Report for 1850 shows corresponding rates of fifteen, ten, fourteen, and ten.

³⁸ Inspection and Condition of the South-Carolina Rail-Road, by the Committee of Six (Charleston, 1848), p. 25Google Scholar . In March 1848 the Columbia to Charleston run of 130 miles took 33 hours.

³⁹ Laws of the State of New York in Relation to the Erie and Champlain Canals together with the Annual Reports of the Canal Commissioners… (Albany, 1825), p. 261.Google Scholar

⁴⁰ Report by , Douglass, in Second Report of the Pennsylvania Canal Commissioners… (Harrisburg, 1827), pp. 29–30Google Scholar . These times were for boats following each other.

⁴¹ Ibid.

⁴² U.S. Congress, House, 20th Cong., 1st sess., H.R. 141, Chesapeake and Ohio Canal (1828), pp. 111–12.Google Scholar

⁴³ New York State, Annual Report of the Canal Commissioners (New York, 1871). The initial profile chart in this report indicates that the Erie had 71 locks, with a total lockage of 675 feet, or ½ feet per lock. The report (p. 73) quotes a statement that the new steamboats “can go through a lock in 6 minutes…or about one half the time a loaded horse boat takes.” Cf. also p. 72.Google Scholar

⁴⁴ Letter to Gouverneur Morris, in Colden, Cadwallader, The Life of Robert Fulton (New York, 1817), p. 287.Google Scholar

⁴⁵ Armroyd, A Connected View of the Whole Internal Navigation of the United States. by a Citizen of the United States (Philadelphia, 1830), pp. 158, 297.Google Scholar

⁴⁶ U.S. Congress, House, 44th Cong., 2d sess., H.D. 46, Part 2, Internal Commerce (1877), p. 41Google Scholar ; and U.S. Congress, Senate, 43d Cong., 1st sess., Serial Set 1588, S.R. 307, Part 1 , Report…Transportation Routes (1874), p. 144Google Scholar . On limited portions of the area where the rail-canal choice was being made, of course, the gradients were steeper (farther West) or milder (Midwest).

⁴⁷ Cranmer, Jerome, “Canal Investment, 1815-1860,” in Trends in the American Economy in the Nineteenth Century (New York: NBER, 1960), p. 564Google Scholar . Fogel, pp. 93-96, outlines a system with a gradient below three feet, selectively serving territory between Michigan and mid Nebraska.

⁴⁸ The state-sponsored Pennsylvania Canal is a marked exception. The failure of the Chesapeake and Ohio Canal to be completed along the steep slope in competition with the B & O railroad is more typical.

⁴⁹ A trial on the Erie, over the stretch from Syracuse to Utica in the 1870's, required ten hours of lockage over the 56 miles, or ten minutes per mile, although the average slope was far under ten feet per mile. Cf. Serial Set 1588 (n.46), p. 152.

⁵⁰ Walker, James, Report to the Directors of the Liverpool and Manchester Railway… (Philadelphia, 1831), p. 30.Google Scholar

⁵¹ The language, not totally clear, permits the road to start trains from Charlestown every hour on the hour, also ten minutes before such times, and permits their starting “5 minutes before the regular stated period”; Annual Reports of the Railroad Corpo rations ‘hellip;Massachusetts, for 1843 (Boston, 1844), pp. 72–73.Google Scholar

⁵² Serial Set 1588 (n.46), Part 2, pp. 48, 66, testimony of Cassatt for th e PRR. Cf. also Whitcomb, chief engineer of the C and O (ibid., p. 144) and Senator Davis ( ibid., p. 24). The general manager of the Blue Line Fast Freight reported tha t trains could follow one another fifteen minutes apart (ibid., p. 14).

Some years later, Albert Fink reported the PRR moving 92 freight and 19 passenger trains in each direction in six hours on its Philadelphia Division. Fink then estimated 36 trains a day as feasible each way on a single-track road-i.e., twenty-minute intervals with actual capacity as “25% more”; Internal Commerce (n.46), pp. 41, 42.

⁵³ Canal, Erie, 30 tons; Laws (n.39), II, 259Google Scholar ; cf. also Report of the Committee on Canals of New York State 1899, Greene, F. V., Chrmn (Albany, 1900), p. 56Google Scholar . In 1825 the president of the canal commissioners stated that boats had “a burthen of 30 to 40 tons”; Report of the Commissioners… Massachusetts (n.28), Appendix 37. Nicholas Wood, A Practical Treatise on Railroads (Philadelphia, 1832), p. 457Google Scholar , stipulates a 25-ton boat, presumably as typical, for his cost estimates. Niles (May 26, 1827), p. 224, states Erie boats carry 40 tons. Morris Canal, 25 tons; Report of Commissioners … Massachusetts, Appendix 55. Delaware and Hudson Canal, 25 tons; D and H Canal Company, Annual Report, 1828, p. 17. Schuylkill Navigation, 25 tons, and Osweco Canal, 20 tons; 1880 Census, U.S., “Report on the Agencies of Transportation,” pp. 734–35Google Scholar . Middlesex Canal, 20 tons ; , Stark, “Navi —gation of the Merrimack River,” in Contributions of the Old Residents Historical Association, Lowell, III, 294Google Scholar , 277. The Chesapeake and delaware was projected for 100-ton boats; Armroyd (n.45), p. 158.

⁵⁴ ERIE: An average load of 41 tons is quoted from the Auditor's Report in Hunt's (Jan. 1860), p. 118. However, New York State Assembly Document 277, Common Canals and Internal Improvement Society of New York (Albany, 02 28, 1840)Google Scholar states that the canal was not adapted to boats over 31 tons. We reject this assertion on the basis of General Ransallaer's 1825 letter citing 30 to 40 tons (see Report of Commissioners … Massachusetts [n.28]) and the report in Hunt's.

Delaware and Hudson: An average of 30 tons in 1828-1843 is noted in the 1880 Census, “Report on…Transportation”; and a 28-33 ton range is noted in Reports on the Rear Mountain Railroad, by Johnson, Edwin F. and Casey, William R. (New York, 1845), p. 54.Google Scholar

⁵⁵ Their 1831 report, quoted in Report on Steam Carriages (n.21), estimates only four tons of freight per horse on a road with a rise and fall of thirty feet per mile. However, reference to Report of the [Pennsylvania] Canal Commissioners (1828), I, 142–43Google Scholar , indicate s that they estimated 4.8 tons for their Susquehanna to Schuylkill railroad (with its ½ feet of ascent per mile) and 10 tons “on a level railroad.”

⁵⁶ Report on Steam Carriages, p. 163. These dat a compare tons for Mauch ½ Chunk with ½ tons for the B & O, while p. 160 indicates the Mauch Chun k calculation is based on an eight-hour day. The 227½ figure is reported by th e B & O superintendent of transportation and “was performed chiefly with the chilled-box car.” On pp. 194 and 199, ibid., tables excerpted from Wood's Treatise lead to a figure of 6 to 6½ tons, at 2½ mph pe r horse. Though these figures are not reconciled by the Report, we assume that Wood's data refer t o English experience, whereas the B & O's experimentation with improved freight cars probably led to higher rates.

⁵⁷ U.S. Congress, House, 24th Cong., 1st sess., H.D. 169, Road-Portsmouth and Linville (1836), p. 27.Google Scholar

⁵⁸ “Minimum,” since it actually applies to road construction.

⁵⁹ Report on Steam Carriages, pp. 158, 153-54. The data shown on p. 152 imply 15 tons at a power of 100 pounds, which is below the average 125-150 pound capacity of a horse. We do not adjust upward to that average because it generally applied to a horse working 6 hours a day (at 3 mph), when its use of capacity is not a maximum. Since the B & O dat a refer to 20 miles a day (i.e., 8 hours at 2½ mph), we adop t the lesser figure; apparend y usage favored the eight-hour, rather than a six-hour, day.

⁶⁰ Statement of Knight, J., in Report on Steam Carriages, p. 153Google Scholar . Knight converts a 17-ton load plus freight car, at 10 mph up an ascent of 1 in 96, to 53 on a level.

⁶¹ Ibid., p. 155. Knight states that in 1828 a locomotive could make the specified ascent and speed with no load; and by 1830, coul d carry a 17-ton load. If we assume that the locomotive weight was 10 tons, as do Walker and Rastrick, then die capability rose from 10 to 27 tons. By this ratio, Knight would hav e implie da 10-ton load in 1828.

⁶² Niles (Nov. 5, 1831), quoting B & O Annual Report. Since the engine weighed only 3½ tons, the real advance was substantial; the amount of coal needed to overcome its friction would have been reduced by well over one half-Walker's estimate having been 10½ tons for engine, tender, and water.

⁶³ Wood, Nicholas, A Practical Treatise on Rail-Roads (2d ed.; Philadelphia, 1838), p. 593.Google Scholar

⁶⁴ Report of a Committee of Directors of the Boston and Worcester Railroad to Reduce the Rates of Fare and Freight on Twelve Roads (Boston, 1840),. p. 29Google Scholar . Down freight averaged 6, 81½, 12½, 1254.

⁶⁵ Tredgold (n.5), pp. 12, 81. From p. 12, we can estimate special coal wagons averaging about 30 per cent of the gross weight, a ratio we apply to his 72-ton gross load from p. 81.

⁶⁶ Wood, p. 737. Each engine hauled from 24 to 28 wagons, each with 53 cwt. of coal-or, say, 26 cars of 3 tons.

⁶⁷ Report of the [Pennsylvania] Canal Commissioners… (1837), pp. 43Google Scholar , 51, 113; cf . American Railway Journal (09. 30, 1837), p. 581Google Scholar.

⁶⁸ Locomotive Engine Catalog of Baldwin, Vail and Hufty (Philadelphia, 1840Google Scholar ; reprint, n.p., 1946), pp. 1-9.

⁶⁹ Ibid.; and Warner, Paul T., “The 4–4–0 (American) Type of Locomotive,” Railway and Locomotive Historical Society, Bulletin (1934), XXXV, 12–15.Google Scholar

⁷⁰ U.S. Congress, House, 29th Cong., 1st sess., H.D. 173, Memorial…Roadway (1846), p. 11.Google Scholar

⁷¹ Cranmer (n.47), p. 564.

⁷² In 1832 a House Committee, having reviewed costs for the Erie and Pennsylvania canals, forecast a $25,000 average; Report on Steam Carriages, pp. 181-82.

⁷³ Message from the Governor, Accompanied with the Report of the Canal Commissioners… (Harrisburg, 1828), pp. 228–52Google Scholar . The single-track road is estimated using iron plates ½ by ½ double-track, 2 by ½

⁷⁴ Gerstner's, F. A. von data show one locomotive for every seven miles of road; “Letters from the United States,” Journal of the Franklin Institute (10. 1840), p. 306Google Scholar . The average locomotive cost about $7,000 and horses $75. Taking one seventh of both locomotive and horse costs per mile yields a difference of nearly $1,000. This figure is clearly overgenerous; horses worked only six to eight hours daily, so that the reserve stock of horses would have had to be proportionately greater than that for engines. Moreover, the useful life of a horse was near five years in such work, whereas that for engines was near ten.

⁷⁵ Poor's data, quoted by Wicker, E. R., “Railroad Investment before the Civil War,” in Trends (n.47), p. 516Google Scholar . By 1850 the region's average had risen only to $40,000. Cf. U.S. Congress, Senate, 32d Cong., 1st sess., S.D. 112 , Andrews Report (1853), p. 389Google Scholar.

⁷⁶ B & O, Annual Report, 1852, Table P, estimates $71 thousand for Boston and Lowell, $65 for Boston and Providence, $71 for Boston and Worcester-as well as (p. 76) $33 for the B & O west of Cumberland, $43 for the New York and Erie (excluding cars and stations). The Third Annual Report of Directors of the Michigan Central Railroad Company (06 1849), pp. 15–17Google Scholar , reports $29 for the Michigan Central, $40 for Utica and Schenectady, $37 for Syracuse and Utica, $40 for Albany and Schenectady. The Exhibit of the Central Ohio Railroad Co. (03 1854), p. 6Google Scholar , estimates $27.6 midway in the work . The Exhibit of the Cleveland, Columbus and Cincinnati RR Co. (06 1849), p. 6, estimates $18.77Wicker, p. 518Google Scholar.

⁷⁸ U.S. Congress, House, 37th Cong., 2d sess., H.R. 86, Permanent Fortifications (1862), p. 258.Google Scholar

⁷⁹ , Broadsheet, Columbia Telescope (Columbia, S. C., 09. 19, 1839)Google Scholar . McNeil foun d that while the terrain wa s easier in the South, costs were not proportionately lower; “Whenever mechanical skill is to b e brought into requisition, the y [i.e., th e North] having more of it, becaus e of a more constant demand for it, can more easily command it, and at a cheaper rate.”

⁸⁰ “Railroads in the United States,” Journal of the Franklin Institute, n.s., XXVI (1840), pp. 89–102, 227-34, 301-6.Google Scholar

⁸¹ In 1835 Tredgold asserted tha t there would be no “considerable difference” between “the risk of expending capital on railways, including the time before a dividend can be made… [and] the risk of expending it on canals”; Treatise (n.5), p. 141.

⁸² Knight estimated $500 per mile for a railroad to and across the Alleghanies, and $528 for “a canal that should be compare d with it”; Report on Steam Carriages, p. 187.

⁸³ Annual Report of the Canal Commissioners…New York… (Albany, 1871), p. 217Google Scholar . An average for the five years ending in 1836 would b e abou t the same . Annual Report of the [Pennsylvania[ Canal Commissioners (1840); we tabulate data for the individual canals from pp. 10–34 and mileage data from p. 5Google Scholar.

⁸⁴ The Pennsylvania average for 1839 was only $500 a mile, suggesting deferred maintenance made up in 1840. We assume th e figure encompasses depreciation; Pennsylvania cost detail indicates that lock replacement and work on aqueducts and bridges were included.

⁸⁵ Gerstner (n.74), p. 298. Cf. also , MassachusettsAnnual Report for 1840Google Scholar ; Central of Georgia Railroad, Engineer's Report (01. 7, 1842), pp. 75–77Google Scholar ; Annual Report of the [Pennsylvania] Canal Commissioners… (1840), pp. 53Google Scholar , 62; Mobile and Ohio, 1858 Annual Report, Table 12, p. 17.

⁸⁶ Colden (n.44), p. 279.

⁸⁷ , Wood, Practical Treatise (1st ed., 1832), p. 457Google Scholar . U.S. Congress, House, 19th Cong., 1st sess., H.R. 228 , C & O Canal (1826), pp. 112–13Google Scholar.

⁸⁸ Report by the State Commission on Cheap Railway Transportation between Boston and Lake Ontario, to the Legislature of Massachusetts (Boston, 1870), p. 121.Google Scholar

⁸⁹ U.S. Congress, Senate, 60th Cong., 1s t sess., S.D. 325, Preliminary Report of the Inland Waterways Commission (1908), p. 219.Google Scholar

⁹⁰ , Fulton, as quoted by Colden, estimates $500 in 1814. Report on Steam Carriages, p. 236Google Scholar , estimate d the 75-ton Lehigh boats had a three-year life, cost $700. We reduce somewhat less than proportionately to get an estimate for an average 40-ton boat. Susquehanna arks, carrying 40 t o 50 tons, “cost at least sixty-five dollars,” were salable for $15 at their destination ; Second Annual Report of the Pennsylvania Canal Commissioners (Harrisburg, 1827), p. 64Google Scholar.

⁹¹ Gerstner.

⁹² Stephenson, Robert and Locke, Joseph, Observations on the Comparative Merits of Locomotives (Philadelphia, 1831), p. 77Google Scholar ; , Wood, Practical Treatise (2d ed., 1838), p. 628Google Scholar.

⁹³ Reports by the President of the South Carolina Railroad Company and of a Committee of Seven (Charleston, 1846), p. 93.Google Scholar

⁹⁴ Report on Steam Carriages, p. 154.

⁹⁵ State of Massachusetts, Senate, No. 98, Report and Bill on the Petition of See-konk Branch Railroad Company, April 1838, p. 64; Report by the State Commission on Cheap Railway Transportation (n.88), p. 122.

The state Report gives data for the Boston and Providence averaging $410; for the Boston and Worcester, $375; the N & L, $325. The first (1843) report for the largest state road, the Western, came to $333.

⁹⁶ Fishlow, pp. 351, 398; he estimates a twenty-year life for rolling stock.

⁹⁷ U.S. Congress, House, 34th Cong., 1st sess., H.R. 274, Pacific Railway and Telegraph (1856), pp. 57, 61, 63.Google Scholar

⁹⁸ , Wood, Practical Treatise (1st ed., 1832), p. 457Google Scholar . Fishlow estimates 7 per cent of steam road costs for equipment. Gerstner (cited in n.74) indicates that engines came to $7,000-8,000 and entire trains from $16,000 to $20,000; p. 299. We therefore use 4 pe r cent of pur $25,000 horse-road investment figure as that for equipment.

⁹⁹ Gerstner, p. 219.

¹⁰⁰ An employment total of 7,000 is derived in the writer's “Labor Force and Employment, 1800-1960,” in Studies in Income and Wealth , Output, Employment and Productivity in the United States after 1800 (New York: NBER, 1966), p. 118Google Scholar . (In the same volume, Albert Fishlow, by an alternative technique, estimates 5,000.) Gerstner, p. 306, reported 3,333 miles of road in operation.

¹⁰¹ Annual Report of [Pennsylvania] Canal Commissioners…1839 (Harrisburg, 1840), pp. 81Google Scholar , 105. The daily rate paid each individual was multiplied by the number of days he worked, thus giving a weighted average.

¹⁰² The $1.53 all-employee average and a rate of $1.00 for watermen and wood-men may be compared with the 85-cent rate for the U.S. from the writer's Manpower in Economic Growth (New York: McGraw-Hill, 1964), pp. 263, 271, 541.Google Scholar

¹⁰³ Data cited ibid., p. 51, suggest a monthly hiring rate of $10 for Erie drivers.! We use a U.S. $10 rate for farm laborers. Smith's figures show the New York Canal Board paying $1.00 a day.

¹⁰⁴ Annual Report of [Pennsylvania[ Canal Commissioners …1839, p. 116. This gives a pay scale of $1.00 a day for hitchers, woodmen, and other unskilled workers. Hayward, James, Report on the Proposed Rail-Road between Boston and Ogdens-burgh (Boston, 1831), p. 26Google Scholar , estimates 75 cents per horse, including harness hire, and $1.25 a day for men to run the teams ; Facts and Arguments in Favour of Adopting Railways in Preference to Canals in the State of Pennsylvania (Philadelphia, 1825), p. 37Google Scholar , stipulates 75 cents a day each for horses and for men. Knight, the B & O's engineer, forecast in 1832 that the B & O could hire horses for 40 cents a day (including harness “and every other expense”) and drivers for 80 cents; Report on Steam Carriages, p. 154. Cf. also the Easton Canal rates of 90 cents for men and 85 cents for two horses and rope (ibid., p. 236) . The Report of the Board of Directors of Internal Improvement of the State of Massachusetts on … Railroad from Boston to Providence (Boston, 1829), p. 21Google Scholar , estimates 50 cents a day for horses and $1.00 for men. We reject these figures as impossibly low for 1840.

We take the 1.5 men per mile of horse railway to imply 1.2 horses per mile; and the 2,500 Erie boats to imply 3,000 horses, including reserve stock.

¹⁰⁵ xhe Reading Report (Jan. 1846), pp. 28-29, shows 1.7 cords of wood required to transport 53 passengers 92 miles; and 3.5 for 61 tons of freight over 92 miles. Fishlow, pp. 127-28 and 324, gives ton miles of freight, passenger miles, and the basis for a U.S. figure of $2.00 a cord. Gerstner reports 3,333 miles in operation. , Ellet (Journal of the Franklin Institute [01. 1844]Google Scholar ) gives costs per mile for water plus oil at about 10 per cent of wood costs.

¹⁰⁶ Philadelphia and Reading, Annual Reports (01. 1846Google Scholar , Jan. 1857) ; Journal of the Franklin Institute (11. 1843Google Scholar , Jan. 1844). Interest: calculated at 6 per cent for 32 additional trains, at $18,000 each; the latter figure based on Gerstner, p. 299. ROAD: from 1838, when it opened, to 1850, the Reading replaced 200 tons of iron (at $30 net per ton) on the average, per million tons of freight, having 83.2 tons per mile. Figuring labor at $500 per mile (in 1856 the P & R paid $1,078 for re-laying double track: Fishlow, p. 377, estimates $400) and multiplying this by 23.4 (million tons) gives $940 per mile. Equipment: the 295-ton loads of the P & R in 1845 generated repairs at 4.9 cents per engine mile, 5.8 per ton mile, for cars. PAYROLL: six employed per P & R coal train, $1.30 per man-day, or 8.5 cents for 295 tons moving one mile. The same figure assumed for 200 tons, adjusted to 23.4 million tons. P & R depot hands, etc., cost 29.3 per cent of crew cost in 1845, and that ratio gives $3,000 additional. WOOD: 13 cords per 192 miles on P & R 295-ton coal trains, reduced by 3 cords (P & R 61-ton freight trains, 92 miles) to give 92 miles of loaded coal train, adjusted to 117,000 miles run, at $2.00 a cord; the Philadelphia and Columbus paid $1.50 per cord in 1839, however, we use the $2.00 price information given in Fishlow, pp. 127-28. Oil, Tallow, Water: P & R, $8.35 for 295-ton coal trains, 186 miles, adjusted to 117,000 miles.

¹⁰⁷ Hazard (May 1840), p. 314, quoting the New York Canal Board.