とある理科の教科書（英語）の一部を載せてみた
In previous book and articles we have examined the use of mechanical power in manu- facturing in Japan after the 1880s.1 We argued that the rapid development of modern sources of power, steam and electricity, played a significant role in the rapid industrialization of Japan. In this paper we examine the utilization of the water wheel, a traditional source of mechanical power, in Japanese manufacturing prior to industrialization. Water wheels were the primary source of mechanical power in preindustrial Japan. In fact, before the end of the Edo period (1603-1867), the water wheel was the only widely used source of power other than human power, Unlike in the West, animal and wind power were rarely utilized.2 Steam engines were first introduced to Japan during the late Edo period, but water power remained dominant over steam power until the 1880s.3 As late as 1884-86 water wheels comprised 56.1~~ of total horsepower compared with the 43.9~~ * Professor (Kyo~,ju) of the Institute of Economic Research, Hitotsubashi
tMany thanks are due to Professors Shigeru Ishikavva, Mataji Umemura and Yoshiro Matsuda at the Hitotsubashi University for valuable comments. I am indebted also to Mrs. Donna Vandenbrink at the University of lllinois for editing the paper and the English. 1 Rydshin Minami. D5ryoku Kakumei to Gljutsu Shinpo : Senzen-Ki Seiz,~-Gyd no Bunseki (Power revolution and technological progress : A study of prewar manufacturing) (Tokyo, 1976); "The Introduction of Electric Power and Its Impact on the Manufacturing Industries : With Special Reference to Smaller Scale Plants," in Hugh T Patrick ed Japanese Industrialization and Its Social Consequences (Berkeley, Calif. 1976) ; "Me- charncal Power in the Industrialization of Japan," Journal of Economic History, 37 (Dec. 1977); and "Me- chanical Power and Printing Technology in Pre-W.W. 11 Japan," Technology and Culture forthcoming. It rs generally agreed that the industrialization of Japan began in the mid-1880s. The modern cotton-spin- ning industry, the frst modern ind
p1' Ibid., p. 93. 1* Mitsukuni Yoshida, Kikai (Machinery) (Tokyo, 1974), pp. 10-11. la The agricultural History Research Association, ed.. History, vol. 2, p. 93. 15 Shigeo Miwa. Ishi-Usu no Nazo : Sangy,~ Ko~ko-Gaku e no Michi (Puzzles on stone mills : An approach to industrial archaeology) (Tokyo, 1975), p. 226. 1' Yoshida, Machinery, p. 11. only on the condition that the mill be closed during the growing season, from spring to autumn. Struggles between millers and farmers in this district became more serious with the construction of another four water wheels between 1751 and 1763. In 1764 the millers finally had to agree to remove the water wheels within the subsequent five years. Through- out Japan during the Edo period millers were restricted from operating water wheels during the rainy season from spring to summer. And in winter, when they were allowed to operate, sufficient water was not available. This competition with agricultural interests for access to water disturbed the development of water-pow
On the other hand, in addition to the restrictions it imposed, the importance of water in Japanese agriculture also created a situation favorable to the diffusion of water power for these manufacturing activities. The highly developed irrigation system, which extended to most of the arable land, made it possible to install water wheels throughout the country. In other words, grain processors were not limited to locating near principle rivers or streams. This positive impact is largely neglected in appraising the early development of Japanese manufacturing.18 Water Wheels in the Meiji Era During the Meiji period (1868-1911) the importance of water wheels for grain proces- sing gradually diminished as new sources of power became available;9 It is possible to trace this trend in some detail because quantitative statistics on water wheels are available for the Meiji period.
pNationwide statistics on water wheels used for rice cleaning for the year 1822 to 1910 are available from the Ch5hatsu Bukken lchiran Hy5 (Tables for goods and resources for provision).ao This survey was conducted by the Rikugun Sho (Ministry of War) to deter- mine the national capability to provide goods and resources in time of emergency. It re- cords the number of water-powered rice-cleaning mills in operation, a figure which can also be taken as an estimate of the number of water wheels employed for this activity. Column 1 of Table I , which is taken from this survey, shows that a great number of water wheels were in operation and their number was rather stable for the period. (The decline between 1896 and 1900 is attributable to a change in the coverage of the survey.) The significance of water power in rice cleaning during the Meiji period can be judged further by comparing the capacity of water-powered mills with total rice production. Col- umns 2 and 3 of Table I present estimates of rice-cleaning c
In spite of the diminishing importance of water wheels in rice cleaning, an interesting innovation appeared in this field. It was rice cleaning in a boat on the Tenryn River, which commenced in 1893-94 and became most developed during the late 1910s. Six to twelve (usually eight) querns were run by water wheels installed on both sides of a boat. Tamura, History, pp. 735-736. This technique calls to mind the "floating mill" which was widely seen in Europe. T. K. Derry and Trevor I. Williams, A Short History of Technology: From the Earliest Times to A.D. 1900 (Oxford, 1960), p. 252.
The earlier version of this survey by the Rikugun Sh6, the Kydbu Sei Hy6, also contains statistics on the number of water powered rice-cleaning mills. There were 9,200 in 1878; 9,516 in 1879; and 10,010 in 1880. The average number of water wheels in these three years, 9.575, was only about one seventh of the number recorded in the Chdhatsu Bukken lchiran Hyd for 1882 (Table 1, column l). One of the reasons for this underenumeration in the former survey is that this survey covers only towns and cities with 100 or more in-habitants.
1982] Yea r 1882 1883 1884 1885 1886 1887 1889 1 890 1891 1896 l 900 l 902 l 904 1906 l 908 1910 WATER WHEELS IN THE PREINDUSTRIAL ECONOMY OF JAPAN TABLE I NUMBER OF WATER WHEELS FOR RICE-CLEANING AND THEIR CAPACITY COMPARED WITH RICE PRODUCTION : 1882-1910 Number of Water Powered Mills (1) 62,879 56,507 55,692 59,264 47,048 48,649 52,924 47,776 52, 1 30 52,710 38,983- 42,823- 43,047- 44,100e 43,705* 43,876- Capacity for Rice Cleaning (2) (3) (1,000 koku) Rice Production (4) (1,000 koku) 27,481 23,583 25,673 35,797 30,121 41,082 39,656 38,261 41,441 37,278 49,773 46,231 51,322 46,316 Rat io (2)/(4) (3)/(4) 8,803 7,91 1 7,797 8,297 6,587 6,81 1 7,409 6,689 7,298 7,379 5,458 5,995 6,027 6, 1 74 6, 1 19 6, 143
1 3,205 ll,866 l I ,695 l 2,445 9,880 l0,216 11,114 10,033 1 0,947 l I ,069 8,186 8,993 9,040 9,261 9,178 9,213 a Estimation under the assumption t at the rice-c]eaning capacity o a water wheel is 140 koku. b Estimation under the assumption it is 210 koku. c Water mills processing one koku and more a day. Source: See Text. that a water wheel can process 140 or 210 koku of rice.21 Total rice production is shown in column 4, and the ratio of capacity to production is indicated in column 5 or 6 depending on the capacity assumption. Two important observations can be made from these statistics. The first is the overall low level of dependence on water power in rice cleaning in the Meiji period. As shown in columns 5 and 6, only 30+50~~ of total rice production in 1883-85 was cleaned by water power.
One koku is equivalent to 180 Iiters. Yoshiyuki Sueo made similar estimates of rice cleaning capacity for Nara Prefecture. His estimates were derived from the number of water wheels in the Suisha Shirabe (Survey on water wheels) in 1 881. Our assumptions about processing capacity of a water wheel employed in Table I are taken from his estimates. Suiryoku Kaihatsu Riy(~ no Rekishi Chiri (History and geography of development and utilization of water power) (Tokyo, 1980), p. 94. 22 According to Sueo only one-sixth of rice produced in Nara Prefecture was cleaned by the water whee]. lbid., p. 94. (5) 28.7 33.l 32.3 18.4
24.6 16.3 1 8.4 19.3 13.2 16.1 1 2. l 13.4 ll.9 13.3 (5~) (6) 43.2 49.6 48 . 5 27.6 36.9 24.4 27.6 28.9 19.8 24. 1 1 8.2 20.0 17.9 19.9 TABLE 2 WHEAT FLOUR CONSUMPTION BY SOURCE: 1878 1912 [February (~) Year 1878 1884 1896 1897 1898 1 899 l 900 1901 1 902 1903 1904 1905 1906 1907 l 908 1909 1910 1911 1912 Imports Processed by : Water Wheels Other Engines 97 97 89 89 88 87 16 78 12 79 16 76 35 56 33 57 lO 32 56 12 29 59 12 20 47 33 10 41 49 37 58 41 54 43 52 36 60
Sources: The Japan Engineering Association, Metji Ko~gyd-Shi (History of manufactures in Meiji) Kikai Hen (Volume on machinery) (Tokyo, 1930), pp. 199-200, 222-223. bustion engines, and electric motors-for the water wheel in rice cleaning.23 This substitu- tion was closely associated with the development of rice cleaning as a commercial activity in urban areas. It is more difficult to examine the utilization of the water wheel for milling because there is no nationwide statistical data. Nevertheless, some indication of the importance of water-powered milling in the early Meiji period can be gained from the Suislla Sllirabe in Nara Prefecture in 1881.
According to this survey, of the 556 water wheels operating in Nara Prefecture in 1881 only 38 (6.8~;) were employed for milling wheat compared with the 296 (53.2~;) in use in rice cleaning establishments.24 Although relative use of water wheels for milling and rice cleaning certainly differed among prefectures, it would be safe to con- clude that throughout Japan milling wheels were far outnumbered by rice-cleaning wheels during the Meiji period, just as they had been during the Edo period.25 Not only was fiour milling a relatively insignificant activity, but also, the dependence on water power in milling decreased over the Meiji period with the establishment of large- scale mills employing steam engines. As shown in Table 2, the percentage of total flour consumption processed by water wheels fell rapidly from its 97~; share in 1878. By 1912 water wheels processed only 36~~ of wheat fiour compared with the 60~; processed by other engines.
During the Meiji period grain processing came to rely less and less on the water wheel " Although there is no nationwide data on the number of water wheels after 1910, it is believed that it decreased rapidly since the 1920s depending on the substitution by internal combustion engines and electric motors. Toshiro Kuroda. Masami Tamaoki and Kiyoshi Maeda. Nipp0'1 no Suisha (Water wheels in Japan)
History before Late Edo Era Water power has been applied to various manufacturing uses outside of grain process- ing, dating back to the eighteenth century. Until the late Edo period, water wheels were, with few exceptions, found only in lower levels of manufacturing processes. But the benefits which mechanization by water power offered were in a range of industries. Sake brewing was a typical industry which early benefited from the introduction of water wheels. Sake brewing consists of two processes, rice cleaning and brewing itself. Although brewing, which required technical skill and long experience, could not be mech- anized, Iabor-saving mechanical devices could be applied to rice cleaning, which was a very simple process.
Brewers in Nadagog6 were the first to successfully employ water wheels for rice-cleaning at the beginning of the eighteenth century. Nadame, a section of Nada- gog~, situated on the so-called fall-1ine starting at the Rokko Dislocation, was in a partic- ularly favorable location for water power utilization. The number of water wheels employed by sake brewers increased from 4 in 1718 to 40 in 1788 and to 66 in 1810. Introducing water power had considerable positive impacts for sake brewers. Not only did labor costs decrease and output increase, but also output quality improved remarkably. Tar~zaemon Yamamuro utilized water wheels to produce a special brand of sake from rice cleaned continuously for 72 hours.
There are numerous other examples of the advantages of water power to manufacturing coming from the early Edo period. In the eighteenth century oil manufactures in Settsu, Kdchi, and lzumi introduced water wheels to grain toasted rape and cotton seeds. Setyu Roku (Record of oil manufactures) written by Nagatsune Okura in 1836 documents that grinding efficiency doubled with this innovation. Ground seeds were then pressed by the 2: The Agricultural History Association, ed., History, voi. 6 (Tokyo, 1955), Table 15 ofch. 3, pt. 3. Yasuto Hashimoto, Komugi Selfun to Seimeu (Wheat milling and noodle making) (Tokyo, 1937), pp. 41 ~,6.
28 See n. 26. 29 Sei-ichi Kawamura, "ShuzO Manufacture to Suisha" (Sake brewing manufacture and water wheels), sa (wedge press) and nakagi (lever press) to release their oil. However no animal or mech- anical power was utilized in this latter process.30 Technical progress occurred in sugar cane processing first with the introduction of animal power but then also with the use of water power.31 Substitution of cows for men to operate the sakuslla (crushing machine) increased the amount of sugar cane processed from 135-136 kan to 25O-1,000 kan.32 Then, in 1717 Sabuni Tabata in Amami Oshima introduced a water wheel, and output of raw sugar, which had been I cho per day by animal power, doubled to 2 chd.33
Ceramic manufactures and ore processors also realized the advantages of water power for crushing materials. It is known, for example, that a water wheel installed on a boat on the Nishiki River at lwakuni was being used to crush potter's clay as early as the beginning of the nineteenth century. And at Arita and Seto, the centers of porcelain production in the Edo period, water wheels were being used to crush subsoil at least since the late eighteenth century.34 During the Bunka period (1801~1817), a gold mine in Sado employed a water wheel with nine pounders to crush ore into much finer powder than was possible by human power using iron hammers.35 The industrial uses to which water power was put during the early Edo period all showed the characteristics that they were at the primary level of manufacturing.
As the previous example show, utilization of water power was most common in activities which involved processing or crushing materials (rice, wheat, rape seeds, sugar cane, soil, and ore). Water wheels were used only rarely at higher levels of production processes. One example of the more advanced application of water wheels occurred in yarn production. Water power had been used in twisting silk yarn at Ashikaga from about 1800. And Shige-ichir~ Numaga at Fujitsuka Village in the Usui District ran several sets of zakuri (sedentary reeling machine) by water power. This unique structure was washed away by a flood in 1861.36 In spite of the advantages of water wheels, water power was not widely applied to in- dustrial activities until the end of the Edo period.
The water concession of farmers, which was discussed earlier as a restriction on the application of water wheels to grain processing, may have been a factor limiting the use of water power in other manufacturing activities as well. But it does not seem to have been the decisive factor, because it does not explain the expansion of water power utilization to industrial uses which occurred from the end of the Edo era. It is our hypothesis that the demand for mechanical power played a significant role in the diffusion of water power. That is, before the end of Edo, manufacturing indus- tries had not been developed enough to utilize mechanical power. History in Late Edo and Early Meiji Over the latter part of th,e Edo and early Meiji periods water power was increasingly
applied to a number of more sophisticated manufacturing processes. The spread of water 30 The Pre-Meiji Japan History of Science Association, ed., History, pp. 9-1 1. 31 Ibid., pp. 84-92 and the Japan Engineering Association, ed., History of tnanufacture. Kagaku Ko~gyd Hen (Volume on the chemical industry) (Tokyo, 1925), pp. 777-78. 32 One kan is equivalent to 3.76 kg. *3 A ch,~ is a measure of a barrel of sugar, the size of which is unknown. 34 Yoshida, Macllinery, pp. 1 8-20. 35 Ibid., p. 16. 3e The Pre-Meiji Japan History of Science Association, ed., History, pp. 12-13. power to these industrial activities occurred both with the introduction of Western technology and with the wider use of the traditional Japanese water wheel.
Modern technology imported from the West in the late Edo offered advantages to a number of industries. The Saga Clan, installed a water wheel in 1852 and used it to bore a cannon in the next year. By this iunovation they saved on labor inputs and increased the accuracy of the cannon. In 1853 the Satsuma Clan made the same innovation.37 Water power was also used to operate bellows at the Kamaishi lron Mine in 1857, and in the middle 1850s the Tokugawa Shogunate established several gun powder plants powered by water wheels.38 A notable example of water power utilization was the Suisha Kan (water-powered factory) at Tagami Village of Satsuma Clan. Here imported weaving machines were run by water power for about ten years starting in 1857.39 Water wheels became a substantial source of power for manufacturing in the early Meiji period.
The energy policy of the government to encourage water power utilization in place of steam power was somewhat responsible for the spread of water wheels. A govern- ment report on the International Exhibition in Vienna in 1873, was a declaration of the policy. The report emphasized the advantages of water power over steam power in Japan a country with abundant water resources. The government encouraged the use of the Western water wheel and water-powered turbine. These engines were widely adopted in modern industries like spinning, weaving, paper manufacturing, and so forth.40 However, substantial growih in water power utilization also occurred with the application of the traditional Japanese water wheel in traditional industries.41 The textile industry is the foremost example of this phenomenon. The traditional water wheel was the primary source of power in the silk-reeling industry, which developed in Nagano Prefecture during the early Meiji period (1877-1886).
This technique gradually spread to silk-reeling establishments in other regions.42 Water power was also used in connection with traditional technology in cotton spinning. For example gara-spinners, a development of the older hand spinners, were run by water wheels from the second decade of the Meiji.43 The significance of the application of water power to the traditional textile industry can be understood by comparing power capacity in this industry with that of other manu- facturing industries in the early Meiji period. Table 3 shows the mean horsepower of water wheels and water turbines in various manufacturing industries for 188l~90. (These machines were largely of the Japanese type and turbines were not yet widely used.) A very great proportion of total water power in all manufacturing was used in textiles (87.6~), especially in silk reeling (73.9~)・
Silk reeling was the foremost user of water power in Japan during the 1880s. Spinning employed a much smaller, though comparatively quite large, share of total water power (13,1~~), because spinning was accomplished to a large extent in modern B7 Ibid., pp. 8-9, 14. 38 Ibid., p. 9. 3D Ibid., pp. 13-14 and Hirone Saigusa, Saigusa Hirone Chosaku-Shti (Collected papers of Hirone Saigusa), vol. 10 (Tokyo, 1973), pp. lO(~l08, 40 Yoshida, Machinery, pp. 2~~29. il lbid., p. 23 and the History of Modern Japanese Industries Research Group, ed., Gendai Nippon Sang:y(~ Hattatsu-Shi (History of the development of modern Japanese industry), vol. 29, Sdron, Jy5 (General con-siderations, Part 1) (Tokyo, 1967), p. I 14. 42 Minami, "Mechanical Power," pp. 947~8. 43 Ibid., pp. 949-50.
