^{note (1) page 40} J. Schumpeter, ‘Capitalism, socialism and democracy’, Harper & Row, 1942.

^{note (2) page 40} R. Solow, ‘Technical change and the aggregate production function’, Review of Economics and Statistics, August 1957; and Investment and technical progress’, Mathematical methods in the social sciences, Stanford, 1959.

^{note (3) page 40} This expression is not quite satisfactory since what is meant is a gap not in knowledge but in the application of existing knowledge. Some argue that the term ‘technical gap’ might be better. It would also then be distinct from the genuine ‘technological’ gap which exists in some developing countries where it can properly be said that there is a lack of techno logical knowledge.

^{note (1) page 41} Hence the name : LD = Linz-Donawitz.

^{note (2) page 41} For crude iron of more than 0.5 per cent phosphorus content.

^{note (3) page 41} Throughout the article ‘Germany’ means West Germany.

^{note (1) page 42} J. K. Stone, ‘Worldwide distribution of oxygen steel making plant’, Iron and Steel Engineer, November 1966.

^{note (1) page 43} The Swedish innovator was however very big by Swedish standards.

^{note (2) page 43} The Ajax process was developed and is still applied by a single major firm for the conversion of existing open hearth furnaces. As it uses pure oxygen to refine steel Ajax is an oxygen process, but it uses a furnace instead of a converter.

^{note (1) page 44} The relatively low percentage of oxygen steel in French output (table 1) is partly due to delays during negotiations on the ‘Plan professionnel de la sidérurgie’ in 1965-66.

^{note (2) page 44} Most of the information given in this section stems from H. Trenkler, ‘Ten Years LD-Process’, Vienna 1960. A number of other similar calculations have been published (for example, W. Adams and J. B. Dirlam, Big steel, invention and innovation’, The Quarterly Journal of Economics, May 1966).

^{note (1) page 45} Percentages of scrap in total crude steel output in 1966 were : Austria, 38; France, 37; Germany, 42; Italy, 63; Sweden, 58; UK, 52. For the six countries together the percentages used under the different processes were : open hearth, 63; electric, 96; oxygen, 22.

^{note (2) page 45} Large open hearth capacities had been installed in the late 1940s and in the 1950s in the industrial countries, even after the first commercial introduction of the LD process in Austria. These will only gradually be converted. A typical view was expressed by a spokesman of the US Steel Corporation, quoted in Forbes, August 1964 : ‘Nobody who has efficient open hearth furnaces is going to throw them out to buy oxygen furnaces. We waited until we needed to replace old capacity.’ Since the bulk of postwar open hearth capacity was built by large companies, this may explain why many of them were laggards in adopting LD.

^{note (1) page 46} On this point the inquiry confirms similar findings in the United States (E. Mansfield, The Economics of Technical Change’, New York 1968, page 109 and W. Adams and J. B. Dirlam, op. cit.).

^{note (2) page 46} For a very detailed and up-to-date description of CC, its history and development, and an extensive and thorough discussion of its economics see ‘Economic aspects of con tinuous casting of steel’, published by the United Nations (ECE) in 1968.

^{note (1) page 48} ECE, op. cit.

^{note (2) page 48} This is supported by ECE, op. cit.

^{note (3) page 48} The inventors were O. Brauns and L. Jordanson at the Central Laboratory of the Paper Industry in Sweden; further development was carried out by the Mead Corporation in the United States.

^{note (1) page 53} The first British-made NC machine was a contouring machine; it was sold to an industrial user in September 1956.

^{note (2) page 53} In the case of Germany the report on ‘Research and Economic Growth’ prepared by the IFO Institut for ECSC, was used. The information concerning France was exclusively based on the report’ Etude de certains aspects de la diffusion et des repercussions des innovations ‘prepared, also for the ECSC, by SODIC, Société pour la Conversion et Ie Développement Industriels, Paris.

^{note (1) page 54} One report in 1967 (‘The state of NC in Europe’; American Machinist, October 1967, vol. III, no. 2870, page 55) said that there were about 1,400 NC machines in England, 700 in Germany, 220 in Sweden, 120 in Italy, 100 in Switzer land, and 60 in France. The total for Western Europe was given as around 2,700 and that for the United States about 12,500. The French figure casts serious doubts on the validity of this estimate : the present inquiry counted 178 NC machines in France as early as 1966 (three times the American Machinist estimate, though it certainly covered not more than a part of total stock). The estimate for ‘England’, on the other hand, seems high even if England is meant to stand for the whole of the United Kingdom.

^{note (2) page 54} The figures for France, Germany, and Italy are estimated by SODIC, IFO, and ISCO/IFO respectively. Those for Sweden and the United Kingdom, which should be more reliable, are based respectively on our inquiry and statistics produced by the United Kingdom Ministry of Technology.

^{note (1) page 55} As the German metalworking industry is almost twice the size of the French, the suggestion that in 1966 the number of NC machine tools in the two countries was much the same, may seem surprising but could reflect the existence of a size able aircraft industry in France and the lack of its counterpart in Germany.

^{note (2) page 55} With a semi-logarithmic scale an exponential function is graphically represented by a straight line.

^{note (3) page 55} In view of the economic revival which has taken place meanwhile in both countries these predictions may need modification : the growth of NC stock may be faster and rising demand may delay the scrapping of older vintages.

^{note (1) page 56} The NCs produced in Germany have so far been mainly supplied to firms with over 500 employees. Of 32 French firms using NCs 24 had more than 1,000 employees. Each of the three Austrian firms employs more than 500. In 1966 over 90 per cent of the Swedish NC stock was in companies employing over 500. In the United Kingdom firms with employment exceeding 1,500 used 85 per cent of the surveyed stock.

^{note (1) page 57} The aerospace industry, together with its first- and second- line suppliers in various parts of engineering, probably accounted in 1967 for about three quarters of the French NC stock (and about one half of the British). NCs are particularly well suited to the complicated metalwork and to the materials (light alloys) used in this industry.

^{note (2) page 57} If an automatic transfer line or machine which now rigidly performs long runs of one operation alone can be made more flexible by NC (and it already can), it clearly cannot be said that NC is not applicable to automatic machines. Some suppliers already offer NC systems of interlinked machines, operating 24 hours, which can equally well manage any lengths of production runs with NC, and are flexible and profitable.

^{note (1) page 58} Most of the firms in Germany made the comparison with automatic equipment which could be reset. The saving in labour costs resulting from the use of NC machines was accordingly small. This however is not a true reflection of savings; NC tools and automatic machines are not strictly comparable, since NC machine tools generally still replace hand operations and not automatic machines, though there have been cases of NC machines replacing the latter as well.

^{note (2) page 58} For example, NC drilling machines are capable of making boards, for rocket assemblies, with 800 very precisely drilled holes, with a minimum of scrap and inspection time. The machining of these and similar pieces was unthinkable earlier. (D. Schon, Technology and Change’, pages 226-7.)

^{note (1) page 59} Most Swedish NC users have changed from one to two or three shifts.

^{note (1) page 60} The shuttle runs across the loom; the fabric is made up of a horizontal and a vertical yarn—the weft and the warp; the shuttle carries the weft.

^{note (1) page 63} In Sweden the Maxbo loom was the first, but the firm which had introduced it closed down in 1962 (the year in which the Draper loom first appeared there) and its use, still confined to Sweden, seems likely in future to be limited even there to new cotton-type synthetic fabrics.

^{note (2) page 63} i.e. the technique of weaving the edges of the fabric.

^{note (1) page 65} In the United Kingdom sheet (window) glass accounts in square footage for more than half of all glass sold and float/ plate for about 15 per cent (the rest being cast or rolled glass), but in terms of value sheet accounts for about 40 per cent and float/plate for rather more. The proportions may, how ever, vary from country to country.

^{note (2) page 65} Possibly because of their expensive and relatively new existing equipment or the view that one float line (built by Boussois who signed a licence contract earlier) would provide the market primarily suited to the float quality.

^{note (3) page 65} Pittsburgh Plate Glass Co., Libbey-Owens-Ford Glass Co., and the Ford Motor Co.

^{note (1) page 66} Asahi Glass Co. and Nippon Sheet Glass Co.

^{note (2) page 66} St Gobain announced in July 1968 the decision by St Roch to construct a float line that will come into operation in 1970, when it will probably exceed in capacity any existing plant in continental Europe.

^{note (3) page 66} The inquiry covered 100 per cent of the Austrian, British, and Swedish flat glass industries; in France, Germany, and Italy the coverage was partial. Estimates made by NIESR, based on other national and trade sources, complete the information for these countries too. The coverage of the industry can therefore be considered as complete.

^{note (1) page 67} The information on other processes in use is not complete. In the United Kingdom the Pittsburgh system of sheet manu facture is the only flat glass manufacturing process still in use, apart from float (and apart from rolled glass for figured and wired manufacture). In Sweden the Fourcault and Pittsburgh methods are applied but by 1970 the last Fourcault tank is expected to disappear. The French industry mainly uses the Libbey-Owens and Pittsburgh methods, whereas in Germany the Fourcault system still appears to be widely used (as well as Libbey-Owens and Pittsburgh).

^{note (2) page 67} This is probably why the new Swedish/Danish glass factory now being built has decided against the float and in favour of an older process : the total market for high quality thick glass is probably not large enough for the production of one float line.

^{note (3) page 67} The inquiry did not cover other clay industries, such as china and pottery, where the tunnel kiln was accepted con siderably earlier and installed in larger numbers.

^{note (4) page 67} In Hoffman kilns they often cause troubles such as acid soot emission.

^{note (1) page 70} The National Coal Board is the United Kingdom's second largest brick producer; many of its brickworks produce brick from the coal slurry and the clay mined together and washed away. See also footnote (a) to table 30.

^{note (2) page 70} The Swedish experience seems to contradict the British, though on a much smaller scale. In Sweden too, the share of hollow bricks is not large and the share of facing bricks has been increasing in the last ten years—but these factors have not hindered the adoption of the tunnel kiln.

^{note (1) page 71} In 1951-54 coal accounted for 90 per cent of all primary energy supplies in the United Kingdom; in 1965 coal's share was still 67 per cent. Coal's percentage share in these two periods in the other countries was : 63 and 32 in Austria, 68 and 47 in France, 91 and 59 in Germany, 28 and 14 in Italy, and 25 and 11 in Sweden.

^{note (2) page 71} For the other four countries the equation y = 45.461 + 0.157x has values for R² of 0.919 and for s of ±0.033 (where x is the percentage share of solid fuels in total energy supplies and 1900 + y is the year of introduction).

^{note (1) page 72} Austria has no comparable shipbuilding industry.

^{note (1) page 73} It is customary to measure shipbuilding output by ton nage. This results in considerable distortion because of the different patterns of output. Japan's production mainly, though not exclusively, consists of relatively high-tonnage, low-value tankers whereas the European yards make, on the whole, more sophisticated ships of higher value but lower tonnage.

^{note (2) page 73} The equation y = 66.78 — 1.55x gives values of 0.863 for R² and ± 0.356 for s, where x is the average percentage share in world shipbuilding for every second year from 1958 to 1966 and 1900 + y is the year of introduction of the new type of machine.

^{note (1) page 76} See chart 8 (page 58).

^{note (2) page 76} As late as 1958 VW (the largest producer in Europe) had no ATL.

^{note (1) page 78} In 1926 a Japanese scientist, Kurosawa, showed that an extract from the fungus Gibberella Fujikuroi, caused abnormal growth in rice and other plants. The substance Gibberellin was however first crystallised only in 1938 and was shown to be made up of several components; one of them, Gibberellin A₃, came to be called GA. Pure GA was extracted and identified chemically in early 1950 by Dr P. W. Brian and a team of ICI biologists. In Europe the first report on it was from a Swedish investigator, Dr E. Sandegren, in 1958 and it was reported to the European Brewing Convention in 1959. Since then there have been a number of experiments both at pilot and at commercial level and GA has been widely adopted in malting.

^{note (1) page 79} The structure of the brewing industry is different in the United Kingdom from that in other countries; most British breweries maintain a large number of ‘tied houses’—i.e. retail outlets.

^{note (2) page 79} The maximum is 0.5 grammes per ton of barley, compared with up to 2.0 grammes, depending on conditions, in the United Kingdom and United States.

^{note (3) page 79} An application for authorisation to use GA was filed years ago with the Ministry of Health; no decision has yet been reached.

^{note (4) page 79} Large brewers, over one million barrels output a year; small brewers, under 100 thousand barrels. (One barrel = 36 gallons = 1.64 hectolitres.)

^{note (5) page 79} Two firms reported using GA in 1955 for the conditioning of the barley; this was considered as experimental.

^{note (6) page 79} ICI is the main producer of GA, though there are others in Belgium and Poland. In the United States production (under licence from ICI) is rather limited and most of it goes into agriculture; supply limitations may restrict the use of GA in malting, at least for the time being. Apart from Germany, there are a few other countries—such as Switzerland—where it is illegal.

^{note (1) page 81} Throughout this section the processes are abbreviated as follows : OXY = basic oxygen processes in steelmaking; CC = continuous casting of steel; SP = special paper presses; NC = numerically controlled machine tools; SL = shuttleless looms in cotton weaving; FG = float glass; TK = tunnel kilns in brickmaking; SCM = new steel cutting methods in shipbuilding; ATL = automatic transfer lines for car engines; GA = Gibberellic Acid in brewing/malting.