Restaging Liebig: A study in the replication of experiments

Annals of Science

Volumn 62, Issue 1, 2005, Pages 1-55

  Usselman, Melvyn C ^a   Reinhart, Christina ^a   Foulser, Kelly ^a   Rocke, Alan J ^b  

^a UNIVERSITY OF WESTERN ONTARIO   (Canada)

^b CASE WESTERN RESERVE UNIVERSITY   (United States)

Author keywords

[No Author keywords available]

Indexed keywords

CARBON; CARBOXYLIC ACIDS; HYDROGEN; NITROGEN; RESEARCH LABORATORIES;

CINCHONINE; PHILOSOPHY; RACEMIC ACID;

ORGANIC COMPOUNDS;

EID: 27844448611     PISSN: 00033790     EISSN: None     Source Type: Journal    
DOI: 10.1080/00033790410001711922     Document Type: Review

References (181)

1. J. Liebig, 'Ueber einen neuen Apparat zur Analyse organischer Körper, und über die Zusammensetzung einiger organischen Substanzen', Annalen der Physik, 21 (1831), 1-43; he had already indicated his interest in improving organic analysis in 'Ueber die Analyse organischer Substanzen',
2. Annalen der Physik, ibid., 18 (1830), 357-68. On Liebig and the analytical method based on the new device,
3. see J. B. Morrell, 'The Chemist Breeders: The Research Schools of Liebig and Thomas Thomson', Ambix, 19 (1972), 1-46;
4. Joseph S. Fruton, 'The Liebig Research Group - a Reappraisal', Proceedings of the American Philosophical Society, 132 (1988), 1-66;
5. F. L. Holmes, 'The Complementarity of Teaching and Research in Liebig's Laboratory', Osiris, 5 (1989), 121-64;
6. William Brock, Justus von Liebig: The Chemical Gatekeeper (Cambridge, 1997), 48-51;
7. A. J. Rocke, 'Organic Analysis in Comparative Perspective: Liebig, Dumas, and Berzelius, 1811-1837', in Holmes and T. H. Levere, eds., Instruments and Experimentation in the History of Chemistry ( Cambridge, 2000), 273-310;
8. idem, 'Origins and Spread of the 'Giessen Model' in University Science', Ambix, 50 (2003), 90-115;
9. and Melvyn Usselman, 'Liebig's Alkaloid Analyses: The Uncertain Route from Elemental Content to Molecular Formulas', Ambix, 50 (2003), 71-89.
10. A. J. Rocke, Chemical Atomism in the Nineteenth Century (Columbus, 1984);
11. U. Klein, Experiments, Models, Paper Tools: Cultures of Organic Chemistry in the Nineteenth Century (Palo Alto, 2003).
12. In nineteenth-century parlance, and in this paper, 'carbonic acid' is synonymous with gaseous carbon dioxide.
13. Wöhler to Liebig, 28 November 1830, Bayerische Staatsbibliothek (BSB), Munich, Sammlung Liebigiana IIB (Wö) 13. All subsequent references to the BSB will be understood to refer to the Liebigiana collection; 'Wö' stands for the respective Wöhler file, followed by letter number. The enormous Liebig-Wöhler correspondence was edited in a massively expurgated fashion by
14. A. W. Hofmann, Aus Justus Liebig's und Friedrich Wöhler's Briefwechsel, 2 vols. (Braunschweig, 1888). Christoph Meinel has completed a full and reliable transcription of the entire correspondence (Justus Liebig - Friedrich Wöhler: Briefwechsel, in preparation), and he has kindly permitted us to cite verbatim extracts from this work (English translations here are by Rocke). We use Professor Meinel's transcriptions for all citations from the Liebig-Wöhler letters; he has silently modernized some of the spelling. See below, note 31.
15. Rocke (note 1, 2000), 285-89. Adoption of the kaliapparat was slower in Great Britain, partly because very few British chemists were attempting organic analyses in the early 1830s.
16. See references in note 1. Among recent historians, the most doubts concerning the impact of the new method of analysis have been expressed by Ernst Homburg, 'Two Factions, One Profession: The Chemical Profession in German Society, 1780-1870', in D. Knight and H. Kragh, eds., The Making of the Chemist: The Social History of Chemistry in Europe, 1789-1914 (Cambridge, 1998), 54-55, 68-69.
17. Liebig, 'Analyse, organische', Handwörterbuch der reinen und angewandten Chemie, I (Braunschweig, 1836-42), 357-400 (1837); appeared simultaneously from the same publisher as a separate monograph,
18. Anleitung zur Analyse organischer Körper; trans. by W. Gregory as Instructions for the Chemical Analyses of Organic Bodies (Glasgow, 1839).
19. T. Settle, 'An Experiment in the History of Science', Science, 133 (1961), 19-23;
20. R. Stuewer, 'A Critical Analysis of Newton's Work on Diffraction', Isis, 61 (1970), 188-205;
21. and J. MacLachlan, 'Galileo's Experiments with Pendulums: Real and Imaginary', Annals of Science, 33 (1976), 173-85.
22. MacLachlan provides an overview in 'Experiments in the History of Science', Isis, 89 (1998), 90-92.
23. R. Root-Bernstein, 'Mendel and Methodology', History of Science, 21 (1983), 275-95.
24. P. Dean and M. Usselman, 'The "Synthetic Palladium" of Richard Chenevix: A Verdict on the Chemist and the Chemistry', Ambix, 26 (1979), 100-15;
25. B. Kronberg, L. Coatsworth, and M. Usselman, 'Mass Spectrometry as a Historical Probe: Quantitative Answers to Historical Questions in Metallurgy', in J. B. Lambert, ed., Archaeological Chemistry - III (Washington, D.C., 1984), 295-310.
26. The context is described and the modern (still to be published) experiments are mentioned in Usselman, 'Multiple Combining Proportions: The Experimental Evidence', in Holmes and Levere (note 1), 243-71 (esp. 268-70, notes 18, 43, 48).
27. L. Principe, 'Apparatus and Reproducibility in Alchemy', in Holmes and Levere (note 1), 55-74;
28. L. Principe, see also idem, '"Chemical Translation" and the Role of Impurities in Alchemy: Examples from Basil Valentine's Triumph-Wagen', Ambix, 34 (1987), 21-30.
29. David Gooding, 'History in the Laboratory: Can We Tell What Really Went On?' in F. James, ed., The Development of the Laboratory: Essays on the Place of Experiment in Industrial Civilization (New York, 1989), 63-82;
30. C. Blondel and M. Dörries, eds., Restaging Coulomb: Usages, controverses et réplications autour de la balance de torsion (Florence, 1994);
31. H. O. Sibum, 'Reworking the Mechanical Value of Heat: Instruments of Precision and Gestures of Accuracy in Early Victorian England', Studies in the History and Philosophy of Science, 26 (1995), 73-106;
32. essays by Sibum and by Falk Riess in M. Heidelberger and F. Steinle, eds., Experimental Essays - Versuche zum Experiment (Baden-Baden, 1999), 141-74;
33. and C. Meinel, ed., Instrument-Experiment: Historische Studien (Berlin, 2000).
34. For which, see, e.g., Allan Franklin, The Neglect of Experiment (Cambridge, 1986); James (note 13);
35. D. Gooding, T. Pinch, and S. Schaffer, eds., The Uses of Experiment (Cambridge, 1989);
36. David Gooding, Experiment and the Making of Meaning (Dordrecht, 1990);
37. Andrew Pickering, ed., Science as Practice and Culture (Chicago, 1992);
38. Andrew Pickering, ed., idem, The Mangle of Practice: Time, Agency, and Science (Chicago, 1995);
39. Peter Galison, Image and Logic: A Material Culture of Microphysics (Chicago, 1997);
40. H.-J. Rheinberger, Toward a History of Epistemic Things (Palo Alto, 1997); Heidelberger and Steinle (note 13); and Meinel (note 13).
41. See Usselman (note 1) for an extended discussion of this issue.
42. I. Hacking, Representing and Intervening (Cambridge, 1983), 150.
43. In O. Wallach, ed., Briefwechsel zwischen J. Berzelius und F. Wöhler, 2 vols. (Leipzig, 1901), I, 327.
44. Liebig to Berzelius, 8 January 1831, in J. Carrière, ed., Berzelius und Liebig, Ihre Briefe von 1831-1845, 2 ed. (Munich, 1898), 4-5. Liebig wrote his parents on New Years Day 1831, 'Ich bin seit 3 Monaten mit mehreren grösseren Arbeiten beschäftigt gewesen, welche jetzt beendigt sind und die bald in Druck erscheinen werden'. Universitätsarchiv Giessen (UAG), Liebig-Brief Nr. 1126.
45. Liebig (note 1). The actual date of appearance (March) is mentioned in Liebig's letter to Wöhler of 3 April 1831, in BSB, IIA 1 (Wö) 24.
46. Regarding the Liebig-Wöhler correspondence, see note 4 above. Twelve of Liebig's laboratory notebooks are also held by the BSB, the earliest of which (Liebigiana IC1, consisting of 89 [+2] leaves) bears the date '1830' on its cover. There are three difficulties in using this notebook. First, no entries are dated (however, it is clear that the content does indeed derive from 1830, probably from the spring of that year until the last days of December). Second, it appears that the entries are generally but not always in chronological order; Liebig apparently sometimes used blank pages out of order, and at other times added annotations to previously written pages. And finally, when one compares the notebook to his correspondence and publications, it is clear that this is by no means a complete record of his laboratory work that year.
47. Liebig to Wöhler, 24 June 1830, BSB IIA1 (Wö) 9.
48. It was 'noch nicht ganz untadelhaft, deshalb teile ich Ihnen seine Konstruktion noch nicht mit'; Liebig to Wöhler, 4 August 1830, BSB IIA1 (Wö) 10.
49. 'Was halten Sie von den organischen Basen?' he asked Wöhler on 8 March 1830, looking toward a subject for collaboration: BSB IIA1 (Wö) 8. Wöhler did not respond to this question. For background, see John E. Lesch, 'Conceptual Change in an Empirical Science: The Discovery of the First Alkaloids', Historical Studies in the Physical Sciences, 11 (1981), 305-328.
50. All this is described in Liebig to Wöhler, 12 October, BSB IIA 1 (Wö) 15; Liebig's letter to Magnus has apparently not survived.
51. E.g., ibid.; Liebig to Trommsdorff, 18 November, UAG Nr. 2751 (request for 'Emetin' and 'Brucin'); Wöhler to Liebig, 28 November, loc. cit. (sending 'Anemonin' from H. Rose, and 'Columbin' from Wittstock); H. Rose to Liebig, 28 November, BSB IIB (Rose); Liebig to Wöhler, 7 December, BSB IIA 1 (Wö) 20 (request for 8 specific alkaloids from Wittstock - with the quoted passage); Wöhler to Liebig, 18 December, BSB IIB (Wö) 16. The last letter has Wöhler strongly advising Liebig to ask Wittstock directly, and to give him something in return - for 'nichts für ungut - Du velangst verdammt viel'.
52. Another way to distinguish unequivocally between the older and newer methods is sample size. The older method could not handle samples larger than about a tenth of a gram, whereas the newer method normally used half-gram samples, and sometimes a gram or more. Both of these circumstances make it simple to tell when Liebig is using which method.
53. Liebig to Wöhler, 12 October 1830, loc. cit.: that 'ich ... mich zu jeder Arbeit disponiert und disponible finde' and that he was already 'mit einer langen und weitläuftigen Arbeit über die Analyse der vegetabilischen Basen beschäftigt'.
54. Liebig to Wöhler, 18 November, BSB IIA1 (Wö) 18. In their correspondence both with Liebig and with each other, Wöhler and Berzelius repeatedly remarked on Liebig's extraordinary productivity.
55. See Magnus's letter to Liebig of 19 November, BSB IIB (Magnus) 1. The seven analyses of fols. 54v-59v (strychnine, cinchonine, salicin, morphine, narcotine, roccellic acid, and caincin) are all in the same hurried hand and colour of ink. Up to this point the only newer analyses represented in the notebook were morphine, narcotine, strychnine, and cinchonine (possibly all obtained from Merck). On fol. 59v is a list of all compounds hitherto analysed by the new method. They are sixteen in number, but this also includes names of compounds added later, whose analyses occur on fols. 60v to 78r (quinic acid, pseudoerythrine, columbin, urea, cyanic acid, brucine, and quinine). The list also includes racemic acid and piperine; the first of these analyses is described only in the published paper, while the second is described neither in the published paper nor in the notebook.
56. Notebook, fol. 55av. This page from the notebook was published as an illustration in an exhibit catalog published by the Justus-Liebig- Universität Giessen in 2003 (Justus Liebig (1803-1873): Der Streitbare Gelehrte, 10), which contains several outstanding essays. However, the illustration was accompanied only by a short caption and no historical analysis. The large words on the page are 'Kupfer', 'Sauerstoff' (twice), 'Kohlenstoff', 'Wasserstoff', and 'Stickstoff'. In this notebook. Liebig used a curious mixture of Latin script and old German handwriting, even sometimes French; these words are obviously in Latin letters. Compare to the German script above: '0,500 gr [Salicin] lieferten 1,000 Kohlensäure und ... 0,287 Wasser'.
57. Wöhler to Liebig, 28 November, BSB IIB (Wö) 13. 'Dein vorletzter Brief an Poggendorff hat uns im höchsten Grade durch Deine Analysier-Manie amüsiert. ... Auf Deinen Apparat bin ich sehr gespannt, denn ich hatte es versprochen, nicht eher eine solche Analyse zu machen, als bis man eine andere Methode entwickelt haben würde'.
58. Dumas and Pelletier, 'Recherches sur la composition élé mentaire et sur quelques propriétés caractéristiques des bases salifiables organiques', Annales de chimie et de physique, 24 (1823), 163-191.
59. Liebig to Trommsdorff, 18 November 1830, UAG, Nr. 2751; Liebig to Wöhler, 7 December 1830, BSB IIA1 (Wö) 20; Liebig (note 1), 14.
60. Names of compounds that appear in the notebook but not in the published paper were piperine, veratrine, caffeine, salicin, and 'Serullas's oil'. Actual analyses of all but piperine appear in the notebook.
61. Liebig to his wife, 25 October 1828: 'Denke Dir, ich habe mir Unterricht im Glasblasen genommen, wenn man nichts kann, wird man die Gelegenheit benutzen etwas zu lernen'; and 2 November 1828: 'Denke Dir, ich habe Unterricht im Glasblasen genommen, und bin nun weit genug gekommen, um mir alle Glasapparate selbst machen zu können'. UAG, Nr. 2209 and 2212. In his letter of 17 November 1830, Wöhler asked Liebig to have a Glasbläsertisch like Liebig's made for him, and to send it to him: BSB IIB (Wö) 12. Liebig had purchased his in Paris, as he told his wife in his letter, written from Arras, of 7 November 1828: UAG, Nr. 2211.
62. J. Volhard, Justus von Liebig, 2 vols. (Leipzig, 1909), I, 222; Volhard was only born in 1834, and could not have known this firsthand. Ettling graduated as a pharmacy student at Giessen in 1831, did some analytical work with Liebig and became his paid assistant in 1835. He made a kaliapparat in the presence of a Glasgow glassblower at the BAAS meeting in 1834 at Manchester;
63. see W. H. Brock and Susanne Stark, 'Liebig, Gregory and the British Association, 1837-1842', Ambix, 37 (1990), 134-147, on 140.
64. We were not the first to do this. Some years ago, Willi Conrad reconstructed a Kaliapparat for museum display and to demonstrate its efficacy. He published no analytical results, nor did he distinguish among the independent components of the apparatus for the measurement of carbon, hydrogen and nitrogen, but he did comment that his replicated Kaliapparat was able to produce 'astonishingly accurate results'. Conrad, Justus von Liebig und sein Einfluss auf die Entwicklung des Chemiestudiums und des Chemieunterrichts an Hochschulen und Schulen (doctoral dissertation, Technische Hochschule Darmstadt, 1985).
65. Usselman (note 1), 73. This is true because of the various assumptions the analyst must make, and the necessary requirements to favour some types of experimental data over others, in order to convert analytical data into a molecular formula.
66. Ibid., 71-89.
67. Ibid., 88.
68. Liebig (note 1), 13. He was particularly critical of Dumas's and Pelletier's silence on this point.
69. Liebig, 'Analyse' (note 7), 360;
70. Instructions, 3. Unless otherwise noted, we use Gregory's translations of 'Analyse', which are reliable, if sometimes overly literal.
71. Sibum (note 13); idem, 'Working Experiments: A History of Gestural Knowledge', The Cambridge Review May (1995), 25-37.
72. Liebig, 'Analyse' (note 7), 365;
73. Instructions, 12-13.
74. Our reconstructed combustion tubes differed in one way from Liebig's, in that they were fashioned from different glass. Liebig used 'Bohemian potash glass', for its toughness and crack-resistance. Our tubes (and all other glassware fashioned in the project) were made from borosilicate glass, which is easier to work than potash glass and more stable chemically and thermally. There is no reason to suspect that the substitution of modern glass for Bohemian potash glass would have any effect on the reconstruction results.
75. Liebig, 'Analyse' (note 7), 362;
76. Instructions, 7.
77. Liebig (note 1), 10. Although the statement was of little help to anyone new to organic analysis, it was probably all that seasoned analysts needed to know, for it meant (accurately or not) that Liebig was not preparing his compounds for analysis in any novel way.
78. Liebig, 'Analyse' (note 7), 366;
79. Instructions, 13.
80. Liebig, 'Analyse', 375;
81. Instructions, 25.
82. Liebig, 'Analyse', 374;
83. Instructions, 23.
84. Liebig, 'Analyse', 366-67;
85. Instructions, 13-14.
86. F. Szabadváry, History of Analytical Chemistry (Reading, 1992), 288.
87. Liebig, 'Analyse', 366-67;
88. Instructions, 13-14.
89. The forced expulsion of the drying tube can become an explosion if the tube is very tightly connected via its cork seal with the combustion tube. Combustion analyses are not without risks to the operator. In our reconstructions, the operator always wore protective laboratory clothing and safety glasses, and performed the analyses behind safety glass in a modern fume hood. Liebig and his students did not have these advantages.
90. Liebig, 'Analyse', 368;
91. Instructions, 16.
92. Liebig, 'Analyse', 364-65;
93. Instructions, 10-11.
94. Liebig (note 1), 4-5.
95. Liebig, 'Analyse' (note 7), 380;
96. Instructions, 32.
97. Liebig, 'Analyse', 367;
98. Instructions, 15.
99. The glass-blower in the Chemistry Department at the University of Western Ontario, Bruce Harwood, provided much expert assistance in all aspects of the project that involved glass. He made the kaliapparat that we used in all our analyses, as well as the apparatus for analyzing nitrogen discussed below.
100. Liebig, 'Analyse', 365;
101. Instructions, 11-12.
102. Liebig, 'Analyse', 362-63;
103. Instructions, 7-8.
104. The use of this tube, and the smaller overall size of the kaliapparat, were the only details in which the 1837 analyses differed from the 1831 paper.
105. Liebig, 'Analyse', 369;
106. Instructions, 17-18. Carbon dioxide and water, the products of complete combustion, are tasteless. Pyrolyzed organic matter, on the other hand, is exceedingly flavourful (as every cook knows).
107. Liebig, 'Analyse', 368-70;
108. Instructions, 16-18.
109. Berzelius, 'Ueber die Zusammensetzung der Weinsäure und Traubensäure ... nebst allgemeinen Bemerkungen ueber solche Körper, die gleiche Zusammensetzung, aber ungleiche Eigenschaften haben', Annalen der Physik, 19 (1830), 305-35 (321).
110. Liebig (note 1), 7. The name of the substance appears in Liebig's notebook only once, on fol. 59v (ca. 18 November 1830), in a list of all 16 compounds to date that had been analysed by the new method. If we take Liebig at his literal word, Hess (an otherwise unknown student of Liebig's) must have done this analysis in early October.
111. Liebig, 'Analyse', 360;
112. Instructions, 4.
113. Liebig's balance was capable of weighing substances with a precision of half a milligram, so weights given to three decimal places are appropriate. Modern electronic balances measure weights accurately to four decimal places. It is the norm in modern science to report all experimental numbers, and calculations based upon them, to appropriate numbers of significant figures. We have used consistent significant figures for reported numbers based on modern stoichiometry and calculations. As was customary in his day, Liebig reported figures in which the number of significant figures varied depending on the nature of the apparatus used, and the extent to which arithmetical operations were carried to extra decimal places. In this paper we always report Liebig's numbers as published, without adjusting significant figures according to modern standards.
114. Usselman (note 1), 71-89.
115. Liebig stated his atomic weight standard explicitly in 'Ueber einen neuen Apparat' (note 1), 13, but it is also evident by deconstructing his arithmetic. These weights were promulgated by Berzelius in his revised system of 1826.
116. See Usselman (note 1), 71-89, for further discussion of other calculational irregularities in Liebig's results.
117. Liebig (note 1), 7.
118. Use of Liebig's values for the carbon content of carbon dioxide and the hydrogen content of water yield his published percent compositions in most, but not all, instances. Inexplicable, and usually quite small, differences are found in some of Liebig's numbers, which suggests that small computational errors, or variance in numbers of decimal places employed in calculations, are present in Liebig's published percent compositions. In the absence of the original calculations it is not possible to identify the source of the numerical differences. The largest differences are encountered in the racemic acid results, where Liebig's published percent values for the three composite elements differ from those calculated using the numbers which work correctly elsewhere.
119. As explained in Usselman (note 1), 71-89, Liebig's values give percent compositions for carbon that are 1.32% high, and for hydrogen that are 0.98% low, compared to the percentages that would result by use of modern atomic weights (assuming theoretical perfect accuracy of the combustion analysis).
120. Liebig (note 1), 1-2.
121. Lesch (note 23), 354.
122. Liebig (note 1), 23-24.
123. Ibid., 17.
124. Liebig's alkaloid paper contains examples of both. See Usselman (note 1), 71-89.
125. Lesch (note 23), 311.
126. Liebig (note 1), 29.
127. Ibid.
128. Liebig, 'Analyse' (note 7), 384;
129. Instructions, 35.
130. Liebig (note 1), 3-4.
131. Ibid., 9-10.
132. Liebig, 'Analyse' (note 7), 385;
133. Instructions, 37.
134. Liebig, 'Analyse', 382;
135. Instructions, 34.
136. Using copper oxide that had already been employed for one or two previous combustion analyses might seem to be a risky decision, since it is possible that reused copper oxide might contain, other than the desired reduced copper, organic contaminants from the preceding analyses. We compared analytical results for carbon and hydrogen on sample compounds from runs differing only in the use of pure versus reused copper oxide. There was no difference in the results and we believe Liebig must have done similar comparative tests. In fact, the oxidative destruction of organic compounds by excess copper oxide is so complete that we were able to purify the troublesome batch of copper oxide containing small amounts of carbonate by combusting it with organic compounds in the normal way.
137. Liebig, 'Analyse', 381;
138. Instructions, 33.
139. Liebig (note 1), 8.
140. Ibid., 8-9. Liebig took for granted that one volume of nitrogen contained twice as many nitrogen atoms as one volume of carbon dioxide did of carbon atoms. Thus, the number of atomic weights of nitrogen was taken to be twice the number of carbon atomic weights.
141. Rocke (note 1, 2000), 290.
142. Sibum (note 13); idem (note 43); Gooding (note 13); Simon Schaffer, 'Experimenters' Techniques, Dyers' Hands, and the Electric Planetarium', Isis, 88 (1997), 456-83. There is also a rich earlier literature on tacit and craft knowledge in scientific experimentation, going back to the work of such scholars as M. Polanyi and J. R. Ravetz.
143. Gooding (note 14, 1990); Simon Schaffer, idem, 'Putting Agency Back into Experiment', in Pickering (note 14, 1992), 65-112.
144. D. Pestre, 'La pratique de reconstruction des experiences historiques, une toute première reflection', in Blondel and Dörries (note 13), 17-30, esp. 19, 29.
145. S. Shapin and S. Schaffer, Leviathan and the Air Pump (Princeton, 1985).
146. Sibum (note 13), 77-82.
147. P. Heering, 'The Replication of the Torsion Balance Experiment', in Blondel and Dörries (note 13), 47-66. Heering's conclusions were disputed by John Heilbron ('On Coulomb's Electrostatic Balance', ibid., 151-61), who suggested that Coulomb could have accidentally or deliberately discovered means to eliminate such residual static charges. Besides, Heilbron noted, Coulomb really did know the answer he was after, and that makes a difference: he was 'illustrating a law of nature', rather than 'specifying a contingent fact' like Faraday or Joule (or Liebig). Heilbron's speculation appears more probable in the light of Christine Blondel's finding that a device that is functionally equivalent to the electrically insulating 'Faraday cage' was known at least from 1781 ('La "mécanisation" de l'electricité: Idéal de mesures exactes et savoir-faire qualitatifs', ibid., 99-119).
148. Holton, The Scientific Imagination: Case Studies (Cambridge, 1978), 25-83.
149. The term is Peter Galison's (note 14, 46 and passim). It refers to the collaborative efforts between scientists on the one hand, and technologists and artisans on the other.
150. George Johnson, 'Here They Are: Science's Ten Most Beautiful Experiments', New York Times, 24 September 2002;
151. Michael Freemantle, 'Chemistry at its Most Beautiful: Pasteur's Separation of Enantiomers Tops List of the Most Memorable Discoveries in Chemistry', Chemical and Engineering News, 25 August 2003, 27-30.
152. Liebig, 'Sur un nouvel appareil pour l'analyse des substances organiques; et sur la composition de quelques-unes de ces substances', Annales de chimie, [2] 47 (1831), 147-97;
153. Dumas, 'Lettre de M. Dumas à M. Gay-Lussac, sur les procédés de l'analyse organique', Annales de chimie, [ ibid., 198-213 (on 198).
154. Wöhler to Berzelius, 24 November 1831, in Wallach (note 17), 381; Berzelius to Wöhler, 20 March 1835, ibid., 609.
155. Liebig to Pelouze, 27 November 1831, Dossier Pelouze, Archives de l'Académie des Sciences, Paris. 'M. Oppermann [the bearer of this letter] a travaillé 2 ans dans mon laboratoire et vous pouvez de lui apprendre la méthode dont je me sers pour les analyses organiques plus parfaitement, que par la description que j'en ai fait. Je serais flatté si on reconnaîtrait son utilité à Paris'.
156. 'Depuis quelque temps ... avec un plein succès ...': Dumas, 'Sur l'esprit pyro-acétique', Annales de chimie, [2] 49 (1832), 208-10.
157. These details are all treated in Rocke, Organische Chemie im Wandel: Der Einzug des Liebig'schen Apparates in französische Laboratorien', Mitteilungen der Fachgruppe Geschichte der Chemie, Gesellschaft Deutscher Chemiker 15 (2000), 98-112.
158. See M. and W. Schneider, 'Das "Handwörterbuch" in Liebigs Biographie', in W. Dressendörfer and W.-D. Müller-Jahnke, eds., Orbis pictus: Kultur- und Pharmaziehistorische Studien (Frankfurt, 1985), 247-54;
159. the Liebig citation is from his letter to Eduard Vieweg of 11 January 1837 (251-52).
160. M. Dörries, 'Balances, Spectroscopes, and the Reflexive Nature of Experiment', Studies in History and Philosophy of Science, 25 (1994), 1-36, on 5.
161. H. M. Collins, Changing Order: Replication and Induction in Scientific Practice (London, 1985), 19, 73-76, 83-84, 129-130 (quote on 130).
162. Cf. Donna Haraway's striking metaphor: 'Nothing insoluble precipitates out of the solution of science as practical culture and cultural practice. Remaining in 'solution' is the permanent condition of knowledge-making' (Social Text, 15 [1997], 126).
163. In Collins's sentence, the first occurrence of 'can' means 'can sometimes'. The second occurrence of the word, as a negative compound, cannot mean 'can sometimes not', but can only mean 'can never'.
164. Ibid., 169-70.
165. Ibid., 170 (sic).
166. Collins, 'A Strong Confirmation of the Experimenters' Regress', Studies in the History and Philosophy of Science, 25 (1994), 493-503, on 503;
167. replying to Allan Franklin, 'How to Avoid the Experimenters' Regress', Studies in the History and Philosophy of Science, ibid., 463-91.
168. Steve Woolgar, ed., Knowledge and Reflexivity (London, 1988);
169. Ronald Giere, Science Without Laws (Chicago, 1999), 20-21.
170. Y. Gingras and B. Godin, 'Expérimentation, instrumentation, et argumentation', Didaskalia, 11 (1997), 149-60;
171. Godin and Gingras, 'The Experimenters' Regress: From Skepticism to Argumentation', Studies in History and Philosophy of Science, 33 (2002), 133-48;
172. H. Collins, 'The Experimenters' Regress as Philosophical Sociology', Studies in History and Philosophy of Science, ibid., 149-56.
173. Hacking (note 16), 231.
174. Mary Hesse, 'Changing Concepts and Stable Order', Social Studies of Science, 16 (1986), 714-26.
175. Hacking (note 16), xiv, 262.
176. Ibid., 251-53. Hacking's subsequent argument for the mutual interdependence of theory, apparatus, and experiment, an interdependence that (he says) makes each of these elements 'essentially irrefutable' and, as he puts it, self-vindicating, appears at first glance to return to Duhem and Quine, but really represents more a development than a contradiction of this position: 'The Self-Vindication of the Laboratory Sciences', in Pickering (note 14), 29-64, on 30.
177. For a clear summary statement of his position with respect to this and related controversial matters, see his The Social Construction of What? (Cambridge, 1999).
178. B. Barnes, D. Bloor, and J. Henry, Scientific Knowledge: A Sociological Analysis (Chicago, 1996). They characterize Collins's methodological idealism as not only impolitic for advocates of the sociology of knowledge, but also simply 'wrong' (13, 32-33, 73-80).
179. Dörries (note 108); Pickering, 'Living in the Material World: On Realism and Experimental Practice', in Gooding et al. (note 14), 275-97; Barnes et al. (note 120); Rheinberger (note 14); Giere (note 114);
180. Alex Pang, 'Visual Representation and Post-Constructivist History of Science', Historical Studies in the Physical and Biological Sciences, 28 (1997), 139-71;
181. Larry Laudan, Beyond Positivism and Relativism: Theory, Method, and Evidence (Boulder, CO, 1996), 4.