examples of concise writing on the subject

are examples of concise writing on the subject. From around 1890, English textile machinery manufacturers started to improve the ring frame,

in part by using various types of spindle; Nasmith (1890) provides a good

account of this.

8. Calculated from, Nasmith (1890), pp.233, 259. This is a comparison of

the actual performance of the Platt Brothers' mule and Howard and

Bullough's Rabbeth-type ring frame. The spindle speed of 7500 r.p.m.

used in the comparison is a little slower than the standard speed of the

latter spinning machine at that time. These figures are adequately supported by Ivey (1912) and the Indian Textile Journal, September 1893

(p.240) and others. The standard output per spindle of 20s yarn was 0.38

lb in 10 hours for the ring frame, compared with 0.33 lb for the mule

(both for weft yarn).

9. Refer to Nasmith (1896), Chapter 12 on this point. Theoretically, there

is no great difference between the two, but by referring to available,

inferior data sources about their actual performance and taking into

consideration the technical design of the machines, it seems that the

mule was more suitable.

10. In the case of coarse and medium yarn the twisting quotient of mule

yarn was 3.75 for warp and 3.25 for weft, whereas that of ring yarn was

4.75 for warp and 3.50 for weft. Hence ring yarn was twisted more.

11. See Dainippon Menshi BOseki Dogyo Rengokai, Rengo Baseki Geppo,

2 (June 1889), pp.55-6 and 3 (July 1889), p.70. Also see NomukyokuKomukyoku (1885). The situation was the same in the Chinese market

(Indian Textile Journal, September 1893, p.240). Furthermore, it is extremely important when considering Japanese ring yarn exports to Asian

markets.

12. For more details on its operation, see Nasmith (1896), Chapter 9. As is

the case in Ivey (1912), Chapter 6, the difficulty of operating the mule

was frequently stressed in American writings on spinning technology.

13. For example see Wilkinson (1899). It is said that in the 'Model Ten'

spinning mills the number of stretches was only about 2, even though

12-16s yarn was being produced (Kawabe, 1937, p.55).

14. It is thought that such was the case in Japanese mule mills (e.g., Mie

Boseki) of the 1880s as well. Each mule had 500 spindles at that time

(see Sampei, 1941, p.377). On the other hand, in America one spinner

and one piecer attended one machine (see Copeland, 1917).

15. See Ivey (1912) and Copeland (1917). It was the same in India. In the

case of Japan, however, data of functional wage rates were not available,

so the wages have to be estimated from the male and female average

wage rates at each mill and the gender-specific grade wage tables from

Aichi BOseki, Mie Boseki and Osaka Boseki. At the very least, the estimated wage rates are not inconsistent with the functional wage system.

16. For example, in America and India, the prices were, respectively, $2.70

compared with $3.00 and 5s 8d compared with 8s Od, so the price of ring

frames was 10 per cent higher than the price of mules in America and

45 per cent higher in India. See Ivey (1912), p.47 and the Indian Textile

Journal, September 1893, p.240. In the case of Japan, data about the

price of the spinning machines themselves are not available and so the 

data has to be obtained by converting the price of the whole set of

equipment into price per spindle. According to such figures, the price of

a mule spindle was about 14 yen, whereas the price of a ring spindle was

between 17 and 19 yen (that is, the ring frame is 10-40 per cent more

expensive). See Takamura (1971), p.123. These figures coincide more or

less with the same type of data from India at that time (13.5 yen per

mule spindle, 15.75 per ring spindle; converted at an exchange rate of

0.45 yen to 1 rupee). Hence, the Indian data on the price of spinning

machines can probably be used as reference data. For these figures, see

Kawamura Ribei's communication to Osaka Boseki (Otani, 1926).

17. Calculated from Boshoku Geppo, 11 (May 1892), p.33. Also see Ivey

(1912), pp.47, 63, 72. In fact, the number of spindles that could be

operated depended on the speed of revolution.

18. This difference in machine productivity alone does not form the basis for

choosing the ring frame,-however. For example, if the ring frame technique was located at R' rather than R on Figure 4.3, M would have been

chosen instead.

19. Because its title was changed in July 1891 to Boshoku Geppo, the

February issue was actually Number 22 of Rengo Boseki Geppo. There

were occasional misunderstandings, as is shown by the entry called 'A

Thousand Questions on Spinning and Weaving' in the July 1891 issue,

where the use of the mule was advocated for warp spinning. On the

whole, however, the information on the latest developments was correct.

20. Articles discussing the pros and cons of the mule and the ring frame

appeared for a long time in the Indian Textile Journal (it was not the

official journal of the Indian Mill Owners' Association, but it did form

the nucleus for technological information). It also had many advertisements for mule and introductory articles about the mule. It is worth

noting that in contrast to the Indian case, there were no articles claiming

the superiority of the mule in Japanese journals.

21. From Boshoku Geppo, 3 (September 1891), pp.11-12. In other words, it

is necessary to take into consideration limiting factors such as power

sources and raw materials in the early period, rather than the differences

in efficiency of the two types of machine.

22. At that time Yamanobe purchased 30000 ring spindles, Saito (who went

to England in 1886) purchased 7000 mule spindles and 3000 ring spindles, Kikuchi bought 5000 ring spindles and Taniguchi bought 29 000

ring spindles and 2000 mule spindles. Also Hattori Shun'ichi of Owari

Boseki went to England in 1887 and bought 15 000 spindles, half of

which were ring spindles and half of which were mule spindles. So, although there was already a groundswell towards the ring frame at that

time, there was still somewhat of a lingering attachment to the mule. For

more detail see Kinugawa (1937-9) and Shibusawa (1956), pp.188-93. It

is worth noting that Kikuchi had already decided to adopt the ring frame

before going to England.

23. This is clear from Nasmith (1890), Chapter 12.

24. For more detail see Kenshi Orimono Toshikki Kyoshinkai (1885).

25. The fact that when raw cotton was imported from India it was not known

that the cotton could be used only after it had been processed by a 

willowing machine (there was a rush of orders for this machine), and the

fact that the word 'candy' (the Indian measuring unit for cotton) was not

understood both reflect this lack of knowledge. These facts also coincide

with the fact that before then, the raw cotton that was called 'Indian

cotton' and imported in small amounts via Hong Kong was actually

Annamese cotton, and that moreover it was not used for spinning. The

production of 20s yarn grew gradually after 1891 and it was not produced in quantity until after 1896. By 1889, each spinning mill had to

take a common test for 16s yarn since the main product was 16s yarn.

See Rengo Boseki Geppo, 2, (June 1889), pp.39-42. It is certain that one

of the impetuses for starting 20s yarn production was the 1890s recession

which encouraged the industry to produce 20s for export as well as for

import substitution.