¹ See in particular Gooding, David, Pinch, Trevor and Schaffer, Simon (eds), The Uses of Experiment. Studies in the Natural Science. (Cambridge University Press, 1989)Google Scholar, especially the Introduction which clearly sets out the aims of this book. My own first attempt on this subject was ‘The relationship between concept and instrument design in eighteenth-century experimental science’, Annals of Science, vol. 36 (1979), pp. 205–224,CrossRefGoogle Scholar primarily based on my work on eighteenth-century electrical instrumentation and experiments, and which caused one philosopher to comment that I had used the word ‘interaction’ in at least five different ways.

² Two recent books are Kemp, Martin, The Science of Ar. (London: Yale University Press, 1990)Google Scholar and Gage, John, Colour and Cultur. (London: Thames and Hudson, 1993).Google Scholar Kemp's title is misleading as he deals primarily with colour theory and perspective, as he points out in the sub-title, ‘Optical Themes in Western Art from Brunelleschi to Seurat’. Indeed perspective is the most common ‘scientific’ theme dealt with by art historians, see for instance the fine study by Edgerton, S. Y., The Heritage of Giotto's Geometry. Art and Science on the Eve of the Scientific Revolutio. (Cornell University Press, 1991).Google Scholar Other aspects that could be studied with profit are (a) how scientists produced idealized images of their experiments often with the help of artists, (b) how such images were used to communicate their results, (c) how artists used scientific metaphors in their paintings (of which Hans Holbein's ‘The Ambassadors’ (1533) is a well-known example), and (d) how artists used scientific aids (such as Dürer's perspective frame, and later the camera obscura and camera lucida) in their compositions. On (a) and (b) see the papers in Mazzolini, R. G., Non-Verbal Communication in Science Prior to 190. (Florence: Leo S. Olski, 1993)Google Scholar, in particular my paper ‘Natural philosophy textbook illustrations 1600- 1800’, pp. 169–196, and John J. Roche, ‘The semantics of graphics in mathematical natural philosophy’, pp. 197–233, which have useful bibliographies.

³ My study of these implications is mainly restricted to experimental nature philosophy, the precursor of physics, although some of my observations concerning analogy and model experiments can be applied more generally, see for instance, David Gooding et al., The Uses of Experimen., the ‘Standards and models’ section of the Introduction, pp. 3–4. Most studies by instrument historians have not been philosophical but artifact-orientated or in terms of cultural artifacts, see my discussion in ‘Instrumentation in the theory and practice of science: scientific instruments as evidence and as an aid to discovery’, Annali dell’Istituto e Museo di Storia della Scienza di Firenz., vol. 10, pt 2 (1985), pp. 87–115.Google Scholar

⁴ Here I am in agreement with Harré, Rom, ‘The dependence of “hitec” science on technology through the rôle of experimental technique’, chapter X in Hackmann, W. D. and Turner, A. J. (eds), Learning, Language and Inventio. (Aldershot: Variorum, 1994).Google Scholar

⁵ Not only began nature to be seen in terms of a giant machine, but machines were developed to elucidate and simulate the phenomena in the laboratory.

⁶ See Peter Galison and Alexi Assmus, ‘Artificial clouds, real particles’, in Gooding et al., The Uses of Experimen., pp. 225–274. They demonstrate that Wilson's cloud chamber only makes historical sense when studied in the context of the confluence of two traditions at the Cavendish Laboratory: the analytic and mimetic, the analytical research in the structure of basic matter (J. J. Thomson's ‘transcendental physics’) and the reproduction of natural phenomena (in Wilson's case, weather phenomena such as cloud formation).

⁷ I have used the 1667 edition of Hooke, R., Micrographia: or Some Physiological Descriptions of Minute Bodies Made by Magnifying Glasse. (London, 1667), Preface, and pp. 175–180, PI. XXIV.Google ScholarWilson, Catherine, ‘Visual surface and visual symbol: The microscope and the occult in early science’, Journal of the History of Ideas, vol. 69 (1988), pp. 85–108,CrossRefGoogle Scholar has rightly called this image a ‘landmark in scientific iconography’. See also Harwood, J. T., ‘Rhetoric and graphics in Micrographia’. in Hunter, M. and Schaffer, S. (eds), Robert Hooke New Studie. (Bury St Edmunds: Boydell Press, 1989), pp. 119–147Google Scholar, which illustrates the subtle changes between Hooke's manuscript drawings and the published engravings, and Alpers, Svetlana, The Art of Describing. Dutch Art in the Seventeenth Centur. (John Murray in association with Chicago University Press, 1983), pp. 72–118, esp. pp. 73–74, 83–84 (Hooke and microscopy).Google Scholar Alpers works has fascinated historians of science without gaining universal approval from art historians.

⁸ Thus instrumentalists had to devise a kind of grammar to make communication possible on how their instruments had to interact with nature. For a key paper, see Dennis, M. A., ‘Graphic understanding: instruments and interpretation in Robert Hooke's Micrographia’, Science in Contex., vol. 3 (1989), pp. 309–364,Google Scholar which has not yet received the serious response which it deserves.

⁹ Indeed Gaston Bachelard has argued that the introduction of the microscope was an impediment: ‘In truth it was only a case of spinning out the old dreams with the new images which the microscope delivered. That people sustained such excitement over these images for so long and in such literary form is the best proof that they dreamed with them’, La Formation de I'esprit scientifiqu., 4th ed. (Paris, 1965), p. 160.Google Scholar

¹⁰ Hooke discusses his instrumental approach in ‘A General Scheme, Or Idea of the Present State of Natural Philosophy, and How its Defects may be Remedied by a Methodical Proceeding in the Making Experiments and collecting Observations Whereby to Compile a Natural History, as the Solid Basis for the Superstructure of True Philosophy’, published by Waller, R. (ed.), The Posthumous Works of Robert Hook. (London, 1705), pp. 1–75, especially pp. 35–37.Google Scholar For a brief discussion and further references, see my paper ‘Attitudes to natural philosophy instruments at the time of Halley and Newton’, Polhem, vol. 6 (1988), pp. 143–158, esp. pp. 143, 146–149.Google Scholar Hooke was well aware of the power of his pictures.

¹¹ Burgess, J., Marten, M. and Taylor, R., Under the Microscope. A Hidden World Reveale. (Cambridge University Press, 1990), pp. 55 and 61Google Scholar, figs 3.22 and 3.33. Figs 1.1—1.4 are a good example of the microscope as a technical/observational frontier: unexpected features are revealed as magnification increases, in this case the magnified head of a pin reveals the rod-shaped bacteria. On the visual impact of these technology-driven images, see Darius, John, Beyond Visio. (Oxford University Press, 1984)Google Scholar, but which does not discuss philosophical implications.

¹² Success in these terms is judged (a) by the discovery of a new phenomenon, in Leeuwenhoek's case that of bacteria, and (b) persuading fellow practitioners that this phenomenon really exists, primarily by making replication possible (see note 8). In the case of Faraday, the only way he could get fellow practitioners to replicate a specific experiment was by sending them a miniature version of the apparatus.

¹³ See Bachelard, G., Les intuitions atomistique. (Paris: Presses Universitaires de France, 1933)Google Scholar, La formation de I'ésprit scientifique (Paris: Presses Universitaires de France, 1938)Google Scholar, and L'activité rationaliste de physique contemporaine (Paris: Presses Universitaires de France, 1951)Google Scholar, and for a discussion of his work, see Gaukroger, S., ‘Bachelard and the problem of epistomological analysis’, Studies in the History and Philosophy of Science, vol. 7 (1976), pp. 189–244CrossRefGoogle Scholar, and Schaffer, S., ‘Natural philosophy’, in Rousseau, G. S. and Porter, R. (eds), The Ferment of Knowledge. Studies in the Historiography of Eighteenth-Century Scienc. (Cambridge University Press, 1980), pp. 77–91.Google Scholar

¹⁴ A classic example is Kuhn's, T. S. account of the discovery of the Leyden jar in The Structure of Scientific Revolution. (University of Chicago Press, 1962), pp. 61–62,Google Scholar which Kuhn agrues as ‘theory-induced’. I would describe such a development much more in terms of organic development and natural selection. See also note 47.

¹⁵ Newton turned an optical toy that could be bought at any fair into a scientific instruments, see Schaffer, Simon, ‘Glass works: Newton's prisms and the uses of experiment’, in Gooding, et al., The Uses of Experimen., pp. 67–104, esp. p. 78.Google Scholar

¹⁶ I have made a similar study of this topic in ‘Instruments and experiments. The case of atmospheric electricity in eighteenth-century Holland’ Tijdschrift voor de Geschiedenis der Geneeskunde, Natuurwetenschappen, Wiskunde en Technie., vol. 10 (1987), pp. 190 (60)–207 (77).Google Scholar See also my essay review, ‘Lightning rods and model experiments: Franklin's science comes of age’, Studies in History and Philosophy of Scienc., vol. 22 (1991), pp. 679–684.Google Scholar My distinction of passive and active instruments has not won universal acceptance, see for instance Bennett, J. A., ‘A viol of water or a wedge of glass’, in Gooding, et al., The Uses of Experimen., pp. 105–114,Google Scholar in which he demonstrates that an instrument can be transformed from being active to passive depending on its use with which I would agree. Bennett appreciates the problems of definition within the context of the various sciences (e.g. optical versus natural philosophy), but suggests that historians should preserve the distinctions (or classifications) established by the practitioners themselves. Thus, the categories ‘mathematical, optical and philosophical instruments’ grew out of the intellectual and craft distinctions of the late seventeenth century. However, as demonstrated by his paper ‘The mechanics’ philosophy and the mechanical philosophy’, History of Scienc., vol. 24 (1986), pp. 1–28,Google Scholar Bennett in his criticism is too biased towards the seventeenth century, and even here he sticks too rigidly to what he considers to be the category of ‘practical mathematics’. In fact, categories have always been quite fluid, as can be discovered when analysing the shift of the categories of instruments over time in contemporary encyclopediae, such as the various editions of Harris's, JohnLexicon technicum; or an Universal Dictionary of Arts and Sciences, explaining not only the terms of Art, but the Arts themselve. (1704).Google Scholar Such a study has not yet been undertaken. At any time, the act of categorizing must be to some extent ahistorical. What Bennett has not appreciated is that my distinction of ‘passive’ and ‘active’ was simply to facilitate a discussion about the role of instruments in experimental philosophy—for which our language is still inadequate. Historians like Bennett have concentrated on the history of astronomy and not on the development of laboratory practices.

¹⁷ For my most detailed discussion on this topic see my ‘Scientific instruments: models of brass and aids to discovery’, in Gooding et al., The Uses of Experimen., pp. 31–65.

¹⁸ Cavendish, H., ‘An account of some attempts to imitate the effects of the torpedo by electricity’, Phil. Trans. R. Soc., vol. 66 (1776), pp. 196–225.CrossRefGoogle Scholar Cavendish reached this conclusion by rearranging well-known electrical laboratory apparatus such as charged Leyden jars, brass chains, and electroscopes, see Hackmann. ‘The relationship between concept and instrument design’, pp. 221–222 and ‘Instruments and experiments’, pp. 55–56. Newly discovered electrical phenomena were only considered electrical if they passed what had become the traditional tests, such as charging a Leyden jar and deflecting a gold-leaf electroscope. For another case study see the identification of electricity produced by the friction of steam from steam engines by Lord Armstrong and Faraday in my paper ‘Electricity from steam: Armstrong's hydroelectric machine in the 1840s’, in Anderson, R. G. W., Bennett, J. A. and Ryan, W. F., (eds), Making Instruments Count. Essays on Historical Scientific Instruments Presented to Gerard L'Estrange Turne. (Aldeshot:Variorum, 1993), pp. 147–173.Google Scholar What is interesting in this example is that once it had been established that the prime cause of this unusual phenomenon was the friction of the water/steam particles, this was readily accepted as it did not challenge the existing ‘mind set’ concerning the behaviour of electricity.

¹⁹ For a more detailed discussion on analogies and model experiments, see Hackmann ‘Instruments and experiments’, pp. 45–48, and the references cited therein. Galison and Assmus (in Gooding et al., The Uses of Experimen., p. 227) refer to the Victorian tradition of ‘mimetic experimentation’, and (p. 231) that a number of late nineteenth-century morphological scientists began to use the laboratory to reproduce natural occurrences such as cyclones or glaciers, etc. In 1892 the geologist E. Reyer wrote that researchers had given up either because quantitative experiments seemed impossible or because experiments had been unable to imitate (nachbilden. natural conditions. Now, by reproducing these phenomena, at least partially, much could be learned, thus heralding the beginning of new experimental physical geology. As I have shown, model-making was one of the main techniques of seventeenth- and eighteenth- century experimental philosophy, and in my study of the Dutch natural philosopher Martinus van Marum (1750–1837), I called it ‘imitative experiments’, based on van Marum's Dutch term nabootse., see Hackmann, ‘The relationship between concept and instrument design’, pp. 220–223, based on my unpublished Belfast MA thesis (1970), ‘The electrical researches of Martinus van Marum (1750–1837)’, pp. 225–226.

²⁰ Gunther, S., ‘Das Polarlicht im Altertum’, Beiträdge z. Geophysisik, vol. 6 (1904), pp. 98—107.Google Scholar For a good synopsis of these sources, see Harvey, E. NewtonA History of Luminescence From the Earliest Times Until 199. (Philadelphia:The American Philosophical Society, 1957).Google Scholar

²¹ Icelandic settlers in Greenland are said to have used the same term noroljo. for this phenomenon. For a typical late nineteenth-century textbook account, see Guillemin, A., Electricity and Magnetis., revised and edited by Thompson, Silvanus P. (London: Macmillan, 1891), pp. 90–134.Google Scholar

²² For the most comprehensive history on the aurora in relation to the history of electroluminescence, see Harvey, , A History of Luminescence (Philadelphia: pp. 37–40, 44–45, 83–84, 255–263, 399Google Scholar.

²³ As the study of gasses was called in the eighteenth century because they were collected over pneumatic troughs.

²⁴ René Descartes had already formulated a similar theory that the aurora was the result of the light of the sun below the horizon. Before the experiments proposed in the seventeenth and eighteenth centuries, the dominant view was that this phenomenon was caused by vapours from the earth set on fire.

²⁵ An interesting aspect here is the importance of contemporary technology as the work was first undertaken by the compass maker Robert Norman in his The Newe Attractiv. (1581) in his attempt to develop a new navigational instrument.

²⁶ For fuller description and references, see Hackmann, ‘Attitudes to natural philosophy instruments’, pp. 152–153, Pls 3 and 4.

²⁷ Ferguson, J., An Introduction to Electricity (London, 1770), pp. 19–20Google Scholar, and Experiment XVII, pp. 63–65, PI. I, fig. 4. Nollet, Abb Jean-Antoine in Leons de physique expérimentale (Paris, 1753)Google Scholar had already described a similar demonstration device in vol. 1 lettre IV, pp. 80–82, PI. I, figs 3–5. William Henley in 1774 called the aurora tube a ‘glass exhausted prime conductor’, made by the instrument maker Edward Nairne, but already described by Watson, William in ‘an account of the phenomena of electricity in vacuo with some observations thereupon’, Phil. Trans. R. Soc., vol. 47 (1752), pp. 362–376CrossRefGoogle Scholar, who remarked that the phenomenon ‘resembled very much the lively coruscations of the aurora borealis’. These immensely popular demonstration devices were described in all the contemporary textbooks.

²⁸ Franklin, B., Experiments and Observations on Electricity, made at Philadelphia in America. To which are added Letters and Papers on Philosophical Subjects (London, 1769), pp. 30–52Google Scholar, letter V, ‘Observations and suppositions toward forming a new hypothesis for explaining the several phenomena of thunder-Gusts’ (dated 29 April 1749); Harvey, A History of Luminescenc., pp. 259–260.

²⁹ The ‘voltameter’ was a technique developed to measure quantity of charge by the decomposition of water in its constituent gases of hydrogen and oxygen. Initially it was also a test to demonstrate that current electricity was a kind of static electricity as it decomposed water in the same way. For the early history of this device (and references), see Hackmann, Willem ‘Leopoldo Nobili and the beginnings of galvanometry’, in Tarozzi, G. (ed.), Leopoldi Nobili e la cultura scientifica del suo tempo (Instituto per i beni artistici culturali naturali della Regione Emilia-Romagna, 1985)Google Scholar. Gaston Plant constructed the first practical lead-acid storage battery or accumulator in 1854.

³⁰ According to Halley, the rotation of the internal kernal was the cause of the diurnal and annual variations of magnetic declination. A layered earth is, of course a very old idea, described for instance by Dante, and in the nineteenth century in Jules Verne's A Journey to the Centre of the Eart. (English trans. 1874) in which the intrepid travellers discover the inhabitants the existence of which had already been suggested by Halley.

³¹ Quarrell, W. H. and Mare, M. (eds), London in 1710 from the Travels of Zacharias Conrad von Uffenbach (London: Faber and Faber, 1934), p. 99Google Scholar; Bennett, J. A., The Mathematical Science of Christopher Wren (Cambridge University Press, 1982), pp. 44–54Google Scholar.

³² Royal Society Journal Book, vol. 12, pp. 203–204 (15 February 1721). I am grateful to Patricia Fara for this information in her unpublished Cambridge Ph.D. thesis, ‘Magnetic England in the eighteenth century’ (September 1993), pp. 51 and 63.

³³ For an example made between 1791 and 1797, Edge, J. R., ‘a terrestrial globe by William Bardin, with some unusual magnetic features’, Bul. of the Scientific Instrument Society, no. 39 (1993), pp. 16–18Google Scholar.

³⁴ Canton, J., ‘An attempt to account for the regular diurnal variation of the horizontal magnetic needle; and also for its irregular variation at the time of an aurora boralis’, Phil. Trans. R. Soc., vol. 51 (1759), pp. 400–402CrossRefGoogle Scholar.

³⁵ He made this discovery at a time when there was no common agreement on the physical relationship between electricity and magnetism. Oersted's conviction concerning the unity of natural forces which made him do this experiment was based not on contemporary experimental philosophy, but on the ideas of Johann Wilhelm Ritter and Kant, and the German naturphilosophe.. See my paper ‘Leopoldo Nobili’, pp. 203–233, esp. pp. 210–211.

³⁶ Once iron filings were used to make visible the unseen magnetic fields around lodestones and artificial magnets, the representations became increasingly abstract, eventually being simply curved lines in space, see Hackmann, ‘Scientific instruments’, p. 192, but especially David Gooding, ‘Magnetic curves and magnetic field: experimentation and representation in the history of a theory’, in Gooding et al., The Uses of Experimen., pp. 183–223, esp. pp. 186–190. For a more detailed discussion see Gooding, David, Experiments and the Making of Meaning (Dordrecht, Boston, London; Kluwer, 1990), pp. 95–113CrossRefGoogle Scholar.

³⁷ Barlow made his electromagnetic terrella at the instigation of George Birkbeck of the London Institution to illustrate experimentally that Ampere's hypothesis that all magnetism was due to electric current could explain the terrestrial magnetic field. Sturgeon, W., ‘Account of an improved electromagnetic apparatus’, Annals of Philosophy, vol. 12 (1826), pp. 357–361Google Scholar. Sturgeon's electromagnetic globe is reproduced in Gooding, Experiments and the Making of Meanin., pp. 206–211, figs 7.12 and 7.13. Nobili's globe is described in his Memorie ed osservazioni edite ed inedite del Cavaliere Leopoldo Nobili colla descrizione ed analisi de'suoi apparati ed instrumenti, 2 vols (Florence, 1834), vol. 2, pp. 22–24Google Scholar, PI. IV, fig. 3, described by Hackmann, and Brenni, P., ‘Gli strumenti scientific’, in the exhibition catalogue L'eredità scientifica di Leopoldo Nobili (Comune di Reggio Emeila: Biblioteco Municipale ‘A Panizzi’ and Instituto e Museo di Storia della Scienza, 1984), item 26, pp. 73–74Google Scholar, and in my Catalogue of the Pneumatic, Magnetic, Electrostatic, and Electromagnetic Instruments in the Museo di Storia della Scienza (Florence: Giunti, 1995)Google Scholar, item 249.

³⁸ Faraday, Michael, Experimental Researches in Electricity, vol. 1 (London, 1839)Google Scholar, article 192. He visualized the terrestrial globe, surrounded by its magnetic lines of force, revolving on its own axis, and as it cut through its own lines of force generating electricity, the currents moving from the equator through the earth to the poles, and there leaping into the air to return to the equator through space, possibly in the form of the aurora borealis and australis. This is an interesting example of the transference of ideas, in this case from magnetic induction in the laboratory to the earth as a huge electromagnetic generator similar in action to the electromagnetic machine being developed from Faraday's work. Faraday was interested in the aurora phenomenon for quite some time judging by his annotations dated from 1832 to 1849. See William, L. Pearce, Michael Faraday (London: Chapman and Hall, 1965), pp. 207–208, 224Google Scholar, note 11. See also Dibner, Bern, Faraday Discloses Electro-magnetic Induction (Burndy Library, 1949)Google Scholar.

³⁹ On its early history, see Hackmann, W. D., ‘The induction coil in medicine and physics 1835–1877’, in Blondel, C., Parot, F., Turner, A. J. and Williams, M. (eds), Studies in the History of Scientific Instruments (London: Roger Turner Books for the Centre de Recherche en Histoire des Sciences et des Techniques de la Cité des Sciences et de l’Industrie, Paris, 1989), pp. 235–250Google Scholar.

⁴⁰ de la Rive, A-A., ‘Extrait d'une lettre à M. Regnault (sur les Aurores boréales) [regarding effect of magnet on electric discharge in gas and the aurora borealis]’, Comptes rendus Ac. Sci., vol. 29 (1849), pp. 412–415Google Scholar; also in Phil. Mag., vol. 35, pp. 446–449. See also the following other papers by de la Rive, : ‘On the diurnal variation of the magnetic needle, and on the aurorae boreales’ Phil. Mag., vol. 34 (1849), pp. 286–294Google Scholar, in which he demonstrates that electric sparks are effected by the poles of a strong electromagnet from which he moved on to his aurora experiment: ‘I cannot conclude this abstract without drawing attention to the circumstance, that M. Arago had already pointed out in 1820, shortly after Œrsted's discovery, the possibility of acting upon the voltaic arc by this magnet, and the analogy which might result between this phenomenon and that of the aurora borealis’ (p. 294); ‘On the rotation of the electric light round the Pole of an electro-magnet’, Phil. Mag., vol. 15 (1858), pp. 463–466CrossRefGoogle Scholar, which refers to his description of this apparatus in his textbook translated by Walker, C. V., A Treatise on Electricity in Theory and Practice, 3 vols (London, 1853–1858), vol. 2 (1856), p. 308Google Scholar; ‘On the action of magnetism upon the electric discharge in highly rarified gaseous media’, Phil. Mag., vol. 33 (1867), pp. 512–530CrossRefGoogle Scholar: ‘I demonstrated the existence of this action as early as 1849, by showing that a magnetic pole causes jets of electricity which escape from it radically to rotate’. See also Harvey, A History of Luminescenc., p. 29.

⁴¹ Guillemin, Electricity and Magnetis., pp. 119–122, fig. 49. I know of three versions of this rare apparatus, one in Museo Scientifico dell’Instituto di Fisica G. Marconi dell’Università in Rome, signed ‘N. 496 Societé genevoise/ 113 Plainpalais/ Genève’, another in the Conservatoire des Art et Métier in Paris, while the prototype is in the Musée d’Histoire des Sciences in Geneva, made in the workshop of Professor Thury in about 1859 under the direction of Eugène Schward, a skilled German instrument maker. It is described by de la Rive, in ‘Nouvelles recherches sur les aurores boréales et australes, et description d'un apparail qui les reproduit…’, Mémoires de la Société de Physique et d’Histoire naturelle de Geneva, vol. 16 (1862), pp. 313–342Google Scholar and in ‘Description d'un appareil qui reproduit les aurores boréales et australes avec les phénomènes qui les accompagnent’, Comptes rendus Ac. Sci., vol. 54 (1862), pp. 1171—1175Google Scholar, and in English ‘Further researches on the aurorae boreales and the phenomena which attend them’, Phil. Mag., vol. 23 (1862), pp. 546–553CrossRefGoogle Scholar.

⁴² It is now thought that the primary cause of the aurora is charged particles which approach the earth at high speed and are deflected by the earth's magnetic field into a ring. These aurora-producing particles may well be part of the outer Van Allen belt. Spectroscopic studies have indicated that the auroral light is radiation from atoms and molecules of oxygen and nitrogen, with some radiation by hydrogen atoms. Auroral displays are most frequent during heightened sunspot activity in particular when there is a solar flare burst out. Magnetic storms occur at the same time. The simultaneous activity of sunspots and magnetic storms was already observed in the eighteenth century, but the significance not understood until the twentieth.

⁴³ Rom Harré, ‘The dependence of “hi-tec” science’.

⁴⁴ On the Ham diviners and the lodestone spoon, see Needham, J., Science and Civilisation in China, vol. IV: 1 (Cambridge University Press, 1962), pp. 261–269Google Scholar, and for details of the cultural transmission of the magnetic compass, W. D. Hackmann, ‘Jan van der Straet (Stradanus) and the origins of the mariner's compass’, in Hackmann and Turner, Learning, Language and Inventio., chapter IX.

⁴⁵ Both Hacking and Gooding have argued for the possibility of the independence of experiment from theory. Certainly exploratory work can create phenomena for which no contemporary theoretical explanation exists—discoveries in eighteenth-century electricity is a good case in point, but even here there were broad guiding principles, such as whether nature was corpuscular and behaving according to the laws of mechanics. Although I would agree with Gooding that the theoretical hypotheses guiding an experiment may fail to specify the relevant parameters and observational conditions necessary to obtain a result, and that these had to be learned in the process of doing the experiment. Gooding refers in this context to Trevor Pinch who has identified such observations as having a low externalit.—they are not predicted by, or dependent upon a particular theory, referring again to my own work on eighteenth-century electricity, I cannot think of any experiment that was totally conceived outside any contemporary theoretical framework. I am not sure whether the same would be true of the images produced by the microscope, such as Hooke's compound eye of the fly or Leeuwenhoek's bacteria. However, here I would claim that these images are still within a contemporary cultural framework, based on assumptions to which Mari Hesse refers as coherence condition.. See Gooding, ‘Magnetic curves and magnetic field’ pp. 191—192, and his ‘How do scientists reach agreement about novel observations?’, Studies in the History and Philosophy of Science, vol. 17 (1986), pp. 205–230CrossRefGoogle Scholar; Hesse, M. B., Revolutions and Reconstructions in the Philosophy of Science (Hassocks: Harvester Press 1980), pp. 131–134Google Scholar; Pinch, T., ‘Towards an analysis of scientific observations: the externality and evidential significance of observation reports in physics’, Social Studies of Science, vol. 15 (1985), pp. 3–36CrossRefGoogle Scholar.

⁴⁶ A common strategy adopted in the eighteenth century by an opponent of a particular model was to argue that the modeller was involved in a circular argument. According to this, the same concepts used to make the model where then ‘proofed’ by the model if it behaved as predicted or recreated the phenomenon. It is difficult to see how analogies can fully avoid circular arguments.

⁴⁷ The same is true of the analogous arguments: Galileo's comparison of the surface of the moon with the mountains of Tuscany is more directly related to the observed phenomena than Hooke's extension to this analogy in which he suggests that a more powerful microscope would perhaps show that the moon is populated by sheep grazing on grass like that covering the hills of Salisbury Plain—see Waller, The Posthumous Works of Robert Hook., p. 243. In the case of model experiments, their success depends on how close the parameters of the real phenomenon can be modelled or paralleled in the laboratory. Cavendish's model of the electric fish was far more successful than Priestley's electric earthquake model in which he passed an electric discharge between two wet planks (the earth) the force of which toppled blocks of wood (the buildings) (Hackmann, ‘Instruments and experiments’, p. 52). This paper has only dealt with the empiricism of modelling, not with such important practicalities as scaling.

⁴⁸ To reiterate note 14 above, instead of Kuhn's paradigm shift, it seems to me that a great deal of this development can be seen in organic or evolutionary terms, with natural selection taking place between competing models.