The Conversion of Wurtz

Meanwhile, Wurtz had reached a decisive turning point in his theoretical commitments. On 18 July 1853, Gerhardt read to the Académie what Wurtz subsequently characterized as "one of his most beautiful papers," demonstrating that not only could alkyl radicals substitute more than once in ammonia, but acetyl and other acyl radicals could do the same, producing secondary amides. What was particularly

remarkable about these compounds 'was that the basic property of the ammonia was not simply suppressed, but rather was eliminated entirely; the new substances were acidic , even though derived from the "type" of ammonia.^[63] Whereas Dumas in the older type theory had declared arrangement (or "constitution") to be the principal determinant of chemical properties, Gerhardt was arguing a thesis of dualistic chemical theory, that the kinds of atoms in the molecule determine properties; or, more precisely, Gerhardt had taken the flexible position that both factors were decisive.

Two weeks later, Wurtz provided an elaboration of Gerhardt's idea.^[64] He found that his isocyanates react with acetic anhydride to yield a tertiary amide containing one ethyl and two acetyl radicals united by a nitrogen atom, but he expressed the results in terms of the water rather than the ammonia type. He then added, "I note here that the relations that exist between water and substances derived from the water type are expressed in a neater and simpler manner with the aid of the equivalents adopted by M. Gerhardt than by employing the notation ordinarily used."^[65] Moreover, he found the opportunity to state that he regarded ether as containing two (rather than one) ethyl radicals substituting the hydrogen of water (rather than adding to oxygen), a hallmark of Laurent's and Gerhardt's new ideas. Hitherto he had been faithful to Liebig's and Dumas' conviction that alcohol was nothing more than a hydrated form of ether. However, although here clearly signaling his new view that major elements of the Laurent-Gerhardt system were superior to the older one, he failed to use the new "equivalents" in this paper.

We possess an eyewitness report of the personal encounters between Wurtz and Gerhardt that summer. A student of Wurtz reminisced:

There [in Wurtz's laboratory] lively, passionate scientific discussions were held between the two young masters. Gerhardt, of an outspoken, brusque, and violent character, and Wurtz, fiery, but glib and subtle, argued over atoms and molecules. At times Gerhardt, whose conviction was profound, became impatient at the resistance of his contradictor, and I have seen him break in his fingers the piece of chalk with which he had just traced the formulas on the blackboard.^[66]

Gerhardt responded formally at the Académie to Wurtz' paper. He declared that it made little difference to him whether the new amides were portrayed as derived from the water or the ammonia type, for all of his formulas were only synoptic —general, flexible, and empirical and designed to do no more than summarize the experiments. Wurtz, however (he said), was still under the thrall of Dumas' types. Accord-

ing to that concept, the chemical properties of the derivatives had to be directly related to the properties of the progenitor, so it made no sense to classify acidic derivatives under a basic type. Gerhardt concluded that the issue came down to Wurtz' treatment of formulas as hypothesized before experiments and used as a guide in interpreting them, a procedure that once more revealed Dumas' influence.^[67]

Two weeks later (on 29 August 1853), Wurtz responded in turn. Yes, their viewpoints regarding the constitution of the amides were essentially the same since they were assuming the same groupings of radicals. Yes, his attitude toward formulas was different from Gerhardt's. His were constitutional in orientation, possessing a "true molecular signification," indicating, inter alia, "the arrangement of the simple or compound molecules," by which he meant, in modern vocabulary, the arrangement of atoms and radicals within the molecule. But no, he did not regard types as "purely mechanical and inert" with respect to properties; the type imprints "un cachet particulier" on all its derivatives, hence the inappropriateness of deriving acidic secondary amides from the basic ammonia type.^[68] In all of this Wurtz was thereby pleading guilty, and proudly so, to Gerhardt's charge that Wurtz was still loyal to Dumas' type theory. It was only in his first systematic theoretical treatise a decade later that Wurtz finally averred, fully in the spirit of Gerhardt, that for the sake of simplicity, compound amides should be regarded as ammonia derivatives even though they do not resemble amines chemically.^[69]

It would seem that it was only during 1853 that Wurtz became aware of Williamson's dramatic experimental work on the ethers and his theoretical development of the water type, for his first mention either of Williamson's name or of the water type occurred in his first paper of this series. Williamson had published his three seminal papers on the water type in 1850 and 1851, but only the first had appeared in a French translation and that was not in one of the principal French chemical journals. Wurtz may have first come across Williamson's work in connection with his duties as foreign correspondent for the Annales de chimie (after January 1852), through his visits to England and his friendship with a mainstay of the London chemists, Hofmann, or through his discussions that summer with Gerhardt. Once familiar with the work, however, he quickly assimilated it to his own concerns, as documented in the series of papers just discussed. It must have been in the summer or autumn of 1853 that he wrote Williamson to ask him to prepare a résumé in French of all three articles for publication in the Annales . The sixteen-page résumé appeared in the January 1854 issue accompanied by editorial notes by Wurtz explaining the background and circumstances.^[70]

Wurtz' high opinion of Williamson's work is revealed in a letter of 18 April 1854.

I would be delighted, my dear Williamson, to send you occasional summaries of my investigations; I will do this each time I present something to the Institute and even before I have fully completed my researches. I hope thereby to be in the position soon of proving to you my good will!

He then praised Williamson's most recent publication, on sulfonyl chloride, and offered to publish it in the Annales de chimie . He added, "I must tell you that your article on etherification has created a sensation" in Paris.^[71]

Clearly it had created a sensation with Wurtz, at least. Williamson had developed a strongly realist and constitutionalist program in connection with the water type; his formulas, he stated carefully, depict "an actual image of what we rationally suppose to be the arrangement of constituent atoms in a compound." In his third paper (1851), he had even suggested a mechanistic interpretation of Wurtz' 1849 isocyanate hydrolysis.^[72] Wurtz had always signaled a similar constitutionalist orientation, but until this point he had not found the ideas of Laurent and Gerhardt compelling. Wurtz' reorientation dated from his polemic with Gerhardt in August 1853 and his approximately simultaneous discovery of Williamson—not to mention his promotions. But clearly he much preferred Williamson's realist interpretation of the new chemistry to Gerhardt's positivist one.

Wurtz stated this reorientation unambiguously in subsequent papers. In later historical accounts, he repeatedly accorded Williamson principal authorship for the newer theory of types—even after he became an ardent champion of Gerhardt.^[73] In 1855 he published two remarkable papers, both of which show the strong (and explicitly acknowledged) influence of Williamson. Inspired by Williamson's development of the multiple water type, Wurtz showed how a triple water model could well account for the reactions of glycerin, with the glyceryl radical forming a bond ("lien") between the three water molecules.^[74] Later that year he revealed what history knows as the "Wurtz reaction," using it to make a theoretical argument —explicitly modeled after Williamson's of 1850—to provide what he regarded as "decisive" and "conclusive proof" that the interpretation of the new school regarding the isolated hydrocarbon "radicals" was superior to Kolbe's and Frankland's explanation. At the end of this paper, he provided a brief but reasonably complete summary of the principal points basic to the new chemistry, unequivocally declaring his adherence to it.^[75]

Wurtz' public conversion, we have seen, appears to date from the summer of 1853, the same point when he achieved financial and personal security. Was there a connection? Is it not possible that he was in fact converted much earlier, refraining from declaring himself for fear of alienating Dumas and others during a period when he was still trying to achieve the difficult task of establishing a Parisian career? Although this thesis is attractive, there are difficulties with it. If he were a secret partisan of Laurent and Gerhardt in 1851 and 1852, he would not have gone out of his way during those years to posit rational formulas directly contradicting the new chemistry, such as water = HO, alcohol = C4H⁵ O,HO, and "sulfobutylate de potasse"^[76] =

It would have been a simple matter to write formulas that were consistent both with Dumas' and with Gerhardt's and Laurent's ideas. Such empiricist formulas were in fact quite prevalent then and would have excited no comment at all. In such a way he could easily have preserved both his orthodoxy and his flexibility for the future revelations of his present convictions.

If Wurtz had been a closet reformer, he also would not have pursued an oral and written polemic with Gerhardt in the summer of 1853, or at least not with the form and content that we see. Finally, we would not see evidence of a gradual conversion that summer, such as his retention of elements of Dumas' theory. The evidence suggests that he was won to the new views not exclusively or even principally by patronage, peer or career pressures, but by evidence and argument—even though it is true that he may well have felt freer to be convinced by those arguments after the possible career penalties had been removed.

Another sign of the gradual character of his conversion is the curious fact that he continued to refrain from actually using the new "equivalents;" he adopted them in his papers only from the beginning of 1859. Old habits die hard; moreover, he may still have felt a certain loyalty to his patron Dumas that inhibited him, even after he had achieved financial and career security. Another factor may have been what I would characterize as a certain visceral conservatism in Wurtz' character, a trait that can be discerned at many points in the preceding narrative.

Finally, it should be noted that the atomic weight reform of the 1850s and 1860s differed in character from previous seemingly similar shifts. A number of different conventional systems had been proposed and used during the previous decades: Dalton's, Davy's, and Thom-

son's early weights; Wollaston's "equivalents"; the French system initiated by Gay-Lussac and Dumas; Berzelius' atomic weights, represented in two principal modifications (before and after 1826); the proposals of Gerhardt and Laurent, from 1842; and finally, the reform championed by Gmelin and Liebig from about 1845, to which most European chemists pledged fealty. The older chemists had lived through most of these battles and must have been growing tired of the constant changes. Every shift made major portions of familiar pedagogy obsolete, and each time the formulas for most substances had to be recast and committed to memory anew. The latest (Gmelin-Liebig) reform was only a few years old, within memory of even the youngest members of the community, when men such as Williamson and Kekulé began to use the Gerhardt-Laurent two-volume formulas. As noted earlier, even Gerhardt himself had used conventional equivalents throughout most of his Traité , for he was afraid that otherwise the book would suffer poor sales. Indeed, the guiding thought behind the Gmelin-Liebig reform was that the new equivalents were regarded as the most empirical of all possibilities, possessing explicitly conventional status. Since they were fully independent of any theory, they could be used indefinitely: the final and ultimate atomic weight system. In short, resistance by the community to yet another shift was natural and understandable.^[77]

Furthermore, advocates of the newest reform were making an important new claim, namely, an ontological one. Despite the positivistic form of Gerhardt's theories, he had no hesitation in declaring the old formulas pure and simply "false" and his the only correct ones. Laurent and Williamson took a much more strongly and explicitly realist stance toward the new weights and formulas, with Williamson providing compelling chemical arguments for their truth. Kekulé made a similar explicitly ontological claim in 1854: "It is not merely a difference in notation, but rather an actual fact" that the formula for water is H₂ O.^[78] So one may sympathize with Wurtz in pondering his decision whether to sign on to the new reform: such a move would signal not merely an opinion that the new weights and formulas were more convenient, but that they were true . One; would want to be very certain before making such a claim. In 1853 he would only go so far as to say that the new formulas were "simpler and neater," not that they were the only correct ones.

Wurtz appears to have become fully convinced of the actual truth of the Gerhardt-Laurent weights and formulas as a result of the work done during an immensely productive period of his life, 1856 to 1858. During these three years, he published eighteen papers on glycol and its many derivatives and on organic acids, diacids, and hydroxyacids.

He was also influenced by Kekulé's and Couper's 1857-1858 publications on structure theory. Kekulé regarded himself in some measure as a disciple of Wurtz, giving Wurtz significant credit for leading ideas in his major paper of 1858. Couper was working in Wurtz' laboratory when he published a substantially similar paper nearly coincidentally with Kekulé. Wurtz was also a major influence on the third principal founder of structure theory, A. M. Butlerov. In the autumn of 1858, Wurtz finally decided to discard equivalents; in all his papers from 1859, he adopted the new atomic weights and formulas using a notational convention devised by Williamson.^[79]

It was just before this final shift by Wurtz that he expressed feelings of isolation and insecurity to his old teacher Liebig. At the last election to membership in the Académie, Fremy had been chosen, and Wurtz conceded that he was not undeserving; but even Berthelot had received more votes that he. Balard had voted for Deville, and Dumas had not supported him. This was an unfortunate omen, indicating that "for my future nothing is secure." He felt more appreciated in Germany than in his homeland.^[80] It is curious that this pessimistic letter was written in February 1858, just before Wurtz decided to shift to the new atomic weights and to carry out an extended public campaign for the new chemistry; it also suggests that Wurtz did not yet feel professionally secure. Nothing could have been better calculated to increase his feelings of insecurity and isolation than to begin such a campaign for views that he knew would be unpopular with important people. That he went right ahead with it suggests that he was convinced that the new chemistry was true and would ultimately prevail. He must have been heartened by the obvious successes of the Gerhardt-Laurent reforms in Germany. Until the victory in France was won, however, he was bound to create a professionally uncomfortable life for himself. He could not have known in 1858 that at his death twenty-six years later the reforms would still not have fully succeeded in his native France.