^{page 90 note 1} Table 1, p. 94, summarizes the component elements of this Sanyo group (Sanyo A), as well as the later group appointed between 11 June 1868 and 15 August 1869 (Sanyo B), and the Sangi of 1869–85. Details of appointments are most conveniently to be found in the tables given vol. VI of lshin-shi (6 vols., Tokyo, 1939–1).

^{page 90 note 2} The Gijō in this period comprised in all 17 kuge (including Imperial princes) and 13 daimyō including former daimyō and heirs to fiefs). Of the 11 fiefs named above, only Tottori was not represented. The others each provided one member, except Hizen, which had two. Apart from these, there were only the daimyō of Tokushima and Tsuwano (the latter promoted from Sanyo, being included in the eight individual appointments noted under that head).

^{page 91 note 1} Of a total of 22 appointments, 9 were of kuge, 13 of daimyō. The only difference in fief representation was that in the new list Tsuwano dropped out and was replaced by Tottori.

^{page 92 note 1} On this subject see Hakaji, Fukaya, Kashizoku chilsuroku shobun no kenkyū, Tokyo, 1941, 41–89.Google Scholar Fukaya's account can best be understood if supplemented by more detailed study of some fiefs; see, for example, the various works dealing with Tosa, Satsuma, and Chōshū listed below, n. 1, p. 98, n. 3, p. 98, n. 2, p. 98.

^{page 92 note 2} The koku (about 5 bushels capacity) was the measure used in assessing feudal rice revenue. A family's holding, whether fief or stipend, was usually listed by its kokudaka, i.e. the estimated total product of the land in rice. The actual revenue received, of course, varied with the rate of dues payable by the cultivator, as well as other factors. It rarely exceeded 40 per cent of the kokndaka and by the nineteenth century was in some fiefs as low as 20 per cent, at least for samurai stipends. In this article figures given in koku are, unless otherwise stated, always kokudaka, not actual revenue.

^{page 92 note 3} Originally those whose place in battle was around their lord's person and banner. Umamawari is only the most common of several designations used.

^{page 92 note 4} Fukuzawa Yukichi, in his Kyūhanjō, has described the distinctions which existed in the Okudaira fief of Nakatsu, from which he came; see Fukuzawa Zenshū (10 vols., Tokyo, 1925–1926), VI, 671–96.Google Scholar

^{page 92 note 5} Economically gōshi need to be distinguished from the others, since in some fiefs they might hold as much as 300 koku, compared with a maximum of 10 to 20 koku for junshi or ashigaru. However, they did not rank as full samurai, though they came closer to it in Satsuma than elsewhere.

^{page 93 note 1} This term is used very loosely in English as a rule. Even the various Japanese equivalents, such as kakyū-bushi, are used differently by different Japanese scholars. To many they connote all those of heishi rank and below, including men of semi-feudal status, while some—Fukuzawa Yukichi, for example—use them only of the latter, excluding heishi.

^{page 93 note 2} Most of the relevant biographies and memoirs are listed in Kōshi, Takanashi, Ishin shiseki kaidai (denki-hen), Tokyo, 1935.Google Scholar To avoid unnecessary footnotes, bibliographical information will only be given below where it seems necessary to supplement Takanashi's entries or give more precise references. The two most useful general biographical works for the period are Kinnō resshi-den, Tokyo, 1906,Google Scholar and Zōi shoken-den, 2 vols., Tokyo, 1927.Google Scholar Both have been extensively used. However, detailed references to them will not be given, since both books are relatively easy to use, the biographical entries being arranged geographically by provinces in the former, alphabetically (gojū-on) by persons in the latter.

^{page 93 note 3} The results of this analysis are summarized in Table 2, p. 94. For the other group of Sanyo (Sanyo B) compare also Table 3, p. 94.

^{page 95 note 1} As exceptions to the general rule, there wqre five daimyū who were granted rank as vassalsin-chief even though they also served as karō to the three senior branch houses of the Tokugawa family. Two were attached to Kishū, two to Owari (these including Xaruse's Inuyama fief), and one to Mito. All were of either 35,000 or 38,000 koku.

^{page 95 note 2} There are several references to the fact that Totoki was fief karō, but the only information on his family's kokndaka is to be found in Chikugo kokushi (3 vols., Fukuoka, 1926–1927), II, 28–9.Google Scholar This reprint of a Tokugawa period work entitled Chikugo shōshi gundan, the original preface to which is dated 1853, gives some notes on the history of the Totoki family. The last member of it to be mentioned—the dating is not clear, but the order is apparently chronological—was a karō of 1,000 koku. This does not appear to be the Totoki Settsu named here, but it is probable that his kokudaka was of about the same order.

^{page 95 note 3} On Kumamoto, apart from the general works already cited, see Kaitei Higo-han kokuji shiryō, 10 vols., Kumamoto, 1932,Google Scholar and Higo jimmei jisho, Kumamoto, 1936.Google Scholar There are biographies of Nagaoka and Yokoi, for details of which see Takanashi, op. cit.

^{page 95 note 4} Kaitei Higo-han kokuji shiryō, X, 904–7.Google Scholar The office of bugyō was one of the three most senior in the Kumamoto fief, limited as a rule to upper samurai. Hence Kimura is so classified, despite lack of other evidence.

^{page 96 note 1} There are biographies of Ōkuma Shigenobu and Soejima Taneomi, for details of which see Takanashi, op. cit. Information on Ōki Takatō has to be sought in obituary notices, etc., some of which have been collected in a volume entitled Ōki Takatō kinen (unpublished, no date), to be found in the library of the Shiryō-hensanjo, Tokyo. Unfortunately, none of these works gives precise information on feudal status, beyond the statement that Ōkuma's family once held 400 koku. None the less, it is fairly clear from general inference that they were heishi.

^{page 96 note 2} These were the two fiefs which had closest connexions with the Bakufu, Owari being one of the three senior Tokugawa branch houses and Echizen a kamon or collateral branch. For supplementary biographical information on Owari see Owari no kinnō, Tokyo, 1925,Google Scholar and Aichi-ken kiyō, Nagoya, 1913;Google Scholar on Echizen see Echizen jimbutsu-shi, 3 vols., Tokyo, 1910.Google Scholar There are several works on Yuri Kimimasa, for details of which see Takanashi, op. cit.

^{page 97 note 1} Aichi-ken kiyō, 280–1, states that Tamiya eventually reached 1,500 koku, though it does not make clear whether this was before or after the Restoration. The figure of 500 koku given here is his hereditary family stipend.

^{page 97 note 2} Tanaka's father is described (Owari no kinnō, appendix) as obscure and of low birth, but was apparently a samurai. Niwa's father was a fief official with a stipend of 300 koku, who had in some measure introduced his son to politics; but since the father was still alive, Niwa Ken had not himself succeeded to the family stipend.

^{page 97 note 3} Their age alone is evidence of this. Niwa was 21 and Tanaka 22 at the time of the Restoration.

^{page 98 note 1} Details of rank, descent, and kokudaka of these three men are given in O-samurai-chū senzogaki keizu-chō, a manuscript held in the prefectural central library at Kochi (Kōchi-ken Chūō Toshokan). Fukuoka Kōtei's rank was low for a karō branch house, compared with other fiefs, but this was not unusual in Tosa. The information concerning the appointment of Kōyama's father is taken from Tosa shidan, No. 44, 1933, 169,Google Scholar where it is given in the lists of Tosa fief officials published serially in that periodical, Nos. 42–4, 1933; 48–9, 1934; 51–2, 1935. The two umamawari-kumigashira, with stipends of 400 koku, were appointed to control the umamawari from within that group's own ranks: see the account of Tosa fief administration by Michio, Hirao in Tosa shidan, No. 55, 1936, 205–7;Google Scholar on the same general subject see also Kōchiken shiyō, Kochi, 1924, 264–6,Google Scholar which includes in addition a useful summary of samurai class divisions in the fief. For biographical information on the Tosa samurai, apart from works cited by Takanashi, see Zoku Tosa ijin-den, Kochi, 1923.Google Scholar

^{page 98 note 2} It is true that the Tosa leaders had more to lose by the eventual abolition of the feudal system than those of Satsuma or Chōshū, but the argument cannot be pressed too far. There were other reasons for the policies they followed. Their fief's influence was not based in the last resort on military strength, so it was logical that they should have sought to prevent complete Satsuma-Chōshū domination by the device of maintaining a loose fief alliance under the presidency of the Tokugawa. When that failed, they turned to ideas of constitutional government, playing an important part in the creation of political parties in Japan.

^{page 98 note 3} Accounts of Satsuma administration and feudal class structure will be found in Shigeno and Komaki, , Satsuhan shidan-shū, Tokyo, 1912, 513–56;Google Scholar Yoshihiko, Hayashi, Satsuhan no kyōiku to zaisei narabi gumbi (Kagoshima, 1939; 3 parts, separately paginated), II, 112–40;Google Scholar and Kagoshima-ken shi (5 vols., Kagoshima, 1939–1943), II, 18–26, 95–119.Google Scholar Of biographical works, the only useful one not listed by Takanashi is Shimonaka Yasaburō, Dai Saigō seiden, 3 vols., Tokyo, 1939–1940,Google Scholar though it is worth noting that ōkubo Toshimichi den, 3 vols., Tokyo, 1910–1911,Google Scholar contains much information on Satsuma samurai other than ōkubo, especially on Komatsu Tatewaki (at I, 177–81).

^{page 99 note 1} Inoue Nagaaki counts as unclassified for purposes of Table 2, but as a Shintō priest he probably ranked as heishi.

^{page 99 note 2} Some supplementary biographical information on the Sanyo from Chōshū, apart from that in works listed by Takanashi, will be found in Bōchō rekishi-reki, 4 vols., Yamaguehi, 1943–1944.Google Scholar For accounts of the fief's administrative and class structure, see: Bōchō kaiten shi (12 vols., Tokyo, 1911–1920), I, 35–68;Google Scholar and Yahachi, Tokiyama, Kōhon Mori noshigeri, Tokyo, 1916, 291–354.Google Scholar

^{page 100 note 1} Both fief doctors and Confucian scholars usually belonged to this group, though there were some who ranked rather higher among the heishi.

^{page 100 note 2} The whole question of the composition and significance of these forces is a difficult one. The latest works on the subject are Junya, Seki, Hansei kaikaku to Meiji ishin, Tokyo, 1956,Google Scholar and Tatsuya, Naramoto, Kinsei kōken shakai shiron, Tokyo, 1952.Google Scholar Takasugi himself died some months before the Restoration, on 17 May 1867.

^{page 101 note 1} Whereas the daimyō made Sanyo in the earlier period were from small and unimportant fiefs, these were men from the great fiefs, whom one might have expected to find among the Gijō. The analysis of this group of Sanyo is summarized in Table 3, p. 82.

^{page 102 note 1} The classification of Terajima as being of less than heishi rank may possibly be misleading. It is known only that he was a retainer of one of the Satsuma sub-vassals. As such he would rank formally lower than samurai of the Satsuma fief proper, but might have been considered equivalent to heishi among his own lord's followers.

^{page 102 note 2} Frequently it involved no more than a nodding acquaintance with things foreign—and ignorance masquerading as knowledge often led to errors, both comic and disastrous. Failure in execution, however, does not necessarily weaken the argument concerning the government's intentions.

^{page 102 note 3} In fact, Hizen's close connexion with Nagasaki, through which port alone Western knowledge entered Japan during most of the Tokugawa period, probably does more to explain the fief's influence than its size or political affiliations.

^{page 103 note 1} In Satsuma, for example, of the castle-town samurai proper, there were 132 families which in 1865 belonged to the upper samurai, as compared with 3,878 families of heishi; see Hayashi, , Satsuhan no kyōiku to zaisei narabi gumbi, II, 133–4.Google Scholar