¹ Lamb, Robert B.. The Mule in Southern Agriculture (Berkeley. 1963).Google Scholar

² Ibid.. pp. 26. 83.

³ Genovese, Eugene. The Political Economy of Slavery: Studies in the Economy and Society of the Slave South (New York. 1963). p. 107.Google Scholar

⁴ Kirby, Jack T.. Rural Worlds Lost: the American South. 1920–1960 (Baton Rouge. 1987), p. 198.Google Scholar

⁵ See. for example. Shannon, Fred A.. The Farmers Last Frontier: Agriculture, 1860–1897 (New York. 1945). pp. 48–50:Google Scholar and Howard, Robert W.. The Horse in America (Chicago. 1965). p. 218.Google Scholar

⁶ Jones, Robert L.. “The Horse and Mule Industry in Ohio to 1865,” Mississippi Valley Historical Review. 33 (06 1946). p. 83.CrossRefGoogle Scholar

⁷ Burkhart, J. A.. “The Reluctant Mule,” Farm Quarterly. 5 (09 1950). pp. 30–33;Google Scholar and Johnson, William. “How to Cope with Mules,” Country Gentlemen, 87 (12 16, 1922), p. 9.Google Scholar

⁸ Smith, Alfred G.. “Our Migratory Mules,” Country Gentlemen, 88 (02 17, 1923), p. 9.Google Scholar

⁹ Wentworth, G. E.. Breeders Gazette. 73 (05 9, 1918). p. 977.Google Scholar

¹⁰ Skinner, John. “Address to the Agricultural Society of Maryland,” American Farmer, 2 (10 20, 1820). p. 239:Google Scholar and Hood, James. “The Mule,” American Farmer. 12 (04 16, 1830), p. 33,Google Scholar quoted in Lamb, The Mule in Southern Agriculture. p. 27.Google Scholar

¹¹ Jones, J. L., “The Mule,” U.S. Bureau of Animal Industry, Annual Report, 8th and 9th, 1891–1892 (Washington, DC, 1893).Google Scholar quoted in “In Praise of Mules.” Living Age, 291 (12 1916), p. 429.Google Scholar

¹² Chittenden, D. W.. “Horses and Mules for Farm Power,” Breeders Gazette, 84 (06 17, 1923). p. 725:Google Scholar and Moore, John H.. The Emergence of the Cotton Kingdom in the Old Southwest (Baton Rouge. 1988). pp. 51–52. While the average work life of the mule may have exceeded that of the horse, it has been reported that under a heavy work regime in the Mississippi Delta around the turn of the century. mules worked from six to eight years.Google ScholarSmith, “Our Migratory Mules,” p.9. I am not aware of data on the horse under a heavy work regime.Google Scholar

¹³ Although there are no data on the relative incidence of disease between horses and mules in southern agriculture, during World War I the ratio of illness in horses to mules was six to one. See Fraser, Alexander G.. “The Draft Mule in the Field in Mexico,” American Veterinary Medical Association Journal. 52 (12 1917). pp. 357–61. For a discussion of working mule behavior on warm days.Google Scholar see “Marvin Wolfe-Mules,” Loblolly, 5 (Winter 1977). pp. 49–51.Google Scholar Both of these notions are corroborated by Moore. The Emergence of the Cotton Kingdom, pp. 51–52;Google Scholar and Olson, John F.. “The Occupational Structure of Plantation Slave Labor in the Late Antebellum Era” (Ph.D. diss., University of Rochester, 1983). p. 70.Google Scholar

¹⁴ Olson. “The Occupational Structure,” p. 70.Google Scholar

¹⁵ It may be this characteristic that has led several commentators to conclude that mules were preferred to horses in the slave South because horses could not bear the treatment they received from slaves. There are many similar statements in southern literature. For example. see Olmsted, Frederick Law. A Journey in the Seaboard Slave States in the Years 1853–54, with, Remarks on Their Economy (New York. 1904). p. 51:Google Scholar and Gray, Lewis C.. History of Agriculture in the United States to 1860 (Gloucester. 1958). pp. 851–52.Google Scholar

¹⁶ Warder, J. T.. “Mule Raising,” in U.S. Patent Office. Report of the Commissioner of Agriculture for the Year 1863. 38th Cong., 1st sess.. H. Ex. Doc. 91. serial no. 1196(Washington, DC, 1864). p. 185:Google Scholar and Renner, G. K.. “The Mule in Southern Agriculture 1821–1950,” Missouri Historical Review, 74 (07 1980). p. 443.Google Scholar

¹⁷ There is an additional economic factor suggesting the mule’s superior ability to perform draft work compared with the horse. Although there was some fluctuation over time, the price of a mule was always 10 to 15 percent higher than that of a draft horse in both the North and South. Not unexpectedly, the price of mules varied with cotton prices. John Moore reported that in 1849 plow horses from Kentucky and Tennessee Cost $100 while the price of a mule ranged from $120 to $150. Moore, The Emergence of the Cotton Kingdom. pp. 51–52. For a comparison of the price of mules from 1860 to 1920.Google Scholar see Wright, Gavin. Old South, New South: Revolutions in the Southern Economy Since the Civil War (New York. 1986). p. 120:Google Scholar and for a comparison of the price of mules with horses and cotton from 1870 to 1940. see Lamb. The Mule in Southern Agriculture. p. 24.Google Scholar

¹⁸ Carver, Thomas N.. Principles of Rural Economics (Boston. 1932). p. 275.Google Scholar

¹⁹ In southern cities horses outnumbered mules approximately four to one. See U. S. Bureau of the Census. Thirteenth Census of the United States, 1910. vols. 6–7. Agriculture (Washington, DC, 1912–1914).Google Scholar

²⁰ Lamb. The Mule in Southern Agriculture, tables 2. 3. 4.Google Scholar

²¹ Data on the stock of draft animals are available from the Census of Agriculture. 1910. This census is especially appropriate: it precedes the introduction of mechanized draft power, and the crop year was typical of southern agricultural production for over a century. Data were obtained for the following states: the cotton South of Alabama. Arkansas. Georgia. Louisiana. Mississippi, North Carolina. Oklahoma. South Carolina. Tennessee. and Texas: and the southern border states of Florida. Kentucky. West Virginia. and Virginia. Missouri is also included because of its importance as a mule-raising state. In addition. data were obtained for four agricultural midwestern states including Illinois. Indiana. Iowa. and Kansas. These states are included in order to have a basis of comparison for the ratio of mules to horses on farms for nonsouthern states. The data included all counties within each state with the exception of those that produced less than $250,000 in agricultural output and those with more mules in cities than on farms. The data include 88 percent of all countries in the 19 states, and approximately one-half of those excluded are Texas counties that produced less than $250.000 in agricultural output.Google Scholar

²² By 1890 the mule-raising industry was concentrated in Missouri. Yet, even in the 10 counties in the state that raised the largest number of mules, horses outnumbered mules two to one. It is thus apparent that most Missouri farmers raised mules for cash income, not for draft work. Renner, in “The Mule in Missouri Agriculture 1821–1950.” p. 448. estimated that 40 percent of Missouri farmers raised mules solely for cash income.Google Scholar

²³ Although cotton was a major cash crop in the South. the proportion of improved acres devoted to cotton was never very large. For example, in 1910 in the cotton states alone, only 27 percent of the 114.8 million acres of improved farm land was devoted to cotton. Regression analysis, however, clearly suggests the influence of cotton and the ratio of share tenants on the use of mules in the South. John Moore and John Olson have noted the relationship between mules and cotton, and Robert Lamb has demonstrated a correlation between mules and cotton and mules and sharecroppers. although the correlation did not strike him as being important. See Moore, The Emergence of the Cotton Kingdom. pp. 51–52: Olson. “The Occupational Structure,” pp. 72–73; and Lamb, The Mule in Southern Agriculture. pp. 79–80.Google Scholar

²⁴ It is interesting to note that after the passage of the Agriculture Adjustment Act in the spring of 1933 and farmers were required to plow up a portion of their crop. “It was said that where single-mule plows were used, farmers had a difficult time getting the animals to walk on the row. A New York Times editor wrote that mules were being punished ‘for refusing to trample down the stalks’ which they had been ‘taught to revere’.” Fite, Gilbert C.. Cotton Fields No More: Southern Agriculture 1865–1980 (Lexington. 1984). p. 130.Google Scholar

²⁵ For examples of the variety of contracts between landowners and share tenants, see Reid, Joseph D. Jr., “Sharecropping As An Undesirable Market Response: The Post-Bellum South,” this Journal. 33 (03 1973). pp. 106–30.Google Scholar

²⁶ There is an interesting anomaly in the use of the mule by nonsoutherners. By the time of the American Revolution, the Pennsylvania Germans had developed the Conestoga. the only widely known native draft horse produced in America. The breed died out around 1850, which is interesting and of great regret because in some ways it was superior to any of the imported draft animals. See Thompson, James Westfall. “A History of Livestock Raising in the U.S.: 1607–1860,” U.S. Bureau of Agricultural Economics. Agricultural History Series, no. 5 (11 1942). p. 51. It is also interesting that the Amish-Mennonites who developed the breed in eastern Pennsylvania replaced the Conestoga with the mule. By 1910 the proportion of mules to horses on farms in the southeastern Pennsylvania counties with relatively large populations of Amish-Mennonite families ranged between 19 and 33 percent. while in abutting counties with small numbers of families of German descent the proportion ranged from 2.5 to 6.5 percent. Moreover, crop mix within the region could not account for the disparity in the ratios.Google Scholar See Hostetler, John A.. Amish Society (Baltimore. 1980). p. 131.Google Scholar

²⁷ The inclusion of all 19 states produces larger coefficients and each coefficient is significant at the 0.0001 level with the correct sign.Google Scholar

²⁸ The exclusion of these three states from the regression produces almost identical results.Google Scholar

²⁹ Examination of the regression’s residuals indicated the possibility of heteroskedasticity associated with only one variable, the ratio of share tenants to owner-operated farms. Application of generalized least squares to adjust for this yielded results that were essentially the same.Google Scholar

³⁰ The adjusted R ² equals 0.6269 with N equal to 1.204. The omitted state dummy is Missouri.

³¹ Although it is known that the mule was used widely in tobacco production. the regression analysis does not substantiate the hypothesis. The most probable explanation is that tobacco acreage was always a very small proportion of cultivated acreage on farms, and because horses were used for transportation as well as a consumption good. the proportion of mules to horses on tobacco farms will not be large enough to have a significant sign. In addition. Olson. in “The Occupational Structure.” p. 107. found that smaller farming units were less likely to have mules and had fewer mules than horses on averageGoogle Scholar.