Structure Theory and the Problem of Independent Codiscovery

Even setting aside the problem of Kolbe's and Frankland's contributions to the development of structure theory, the sequence of events leading to Kekulé's 1857-1858 papers is complex. In both of these papers, Kekulé tried to make clear his intellectual debts to his predecessors, especially Gerhardt, Williamson, Odling, and Wurtz, and he disclaimed any great originality. There is no reason to read any exaggerated modesty into his protestations, for the theory of polyatomic radicals and the valence ideas being discussed by these and other chemists did indeed provide at least the implicit basis for most of the content of Kekulé's articles. What Kekulé had done that was new was to provide the first really systematic exposition of these ideas and to explore some novel implications—especially that of the combination of carbon atoms with one another, within the restrictions of tetravalence.

The latter proposition cannot be underestimated; it is the foundation of structural organic chemistry. And yet we can find intimations of even this suggestion in the work of many of the above-named predecessors of valence theory. Moreover, less than a month after Kekulé's second paper was published, Archibald Scott Couper independently proposed a theory that was identical to Kekulé's in all of its essentials, including carbon self-linking.^[55] Couper's claim to be an independent codiscoverer was valid, but his work both before and after 1858 was virtually without influence on the science of his day. Thus, it is not further discussed here, other than to remark that a case can be made that Wurtz' subatomic speculation may have influenced the genesis of his, as well as Kekulé's, theory.^[56] Frankland and Kolbe, of course, also had claims to some elements of Kekulé's theory, and their story is treated in the next chapter.

Williamson, a model of gentlemanly reserve, never urged his priority for polyatomic radicals and valence theory beyond what Gerhardt, Wurtz, and Kekulé granted him, even though many historians now see him as the real genius behind the entire reform movement culminating in the Karlsruhe Congress. As for William Odling, he had been at the

end of 1853 the second after Frankland to attempt a general statement on atomic valence and the first to do so using atomic weights—at the same time developing the concept of "mixed types." In 1855 he had also clearly laid out the marsh gas type. He tactfully pointed out these contributions in an article published shortly after Kekulé's second structure theory paper. He remarked, "Kekulé has recently published a paper in which these views are brought forward more definitely, and with an amount of illustration heretofore not possible."^[57] Any incipient priority fight between Kekulé and Odling essentially died at this time. Kekulé's own view, which as far as can be determined remained unexpressed directly to either Williamson or Odling, was that he was the first to apply the Englishmen's ideas systematically to atoms rather than radicals.^[58]

Wurtz reviewed Kekulé's structure theory paper in the first issue of his Répertoire de chimie pure , a sort of French version of Berzelius' Jahresbericht and an official organ of the new Société Chimique de Paris. In general quite complimentary, Wurtz approved of Kekulé's call to go "back to the elements" by constructing molecules according to atomicities of the atoms, analogous to the way the theory of polyatomic radicals had been developed. He pointedly noted, however, that others before Kekulé had already gone back to the elements: Williamson, Odling, and he himself, in 1855. Moreover, "I do not think I deceive myself," he wrote, "in believing that my recent work on the syntheses of polyatomic alcohols has provided this idea with the experimental confirmation which it lacked." He cited with approval Kekulé's comment that his (Wurtz') ideas had formed Kekulé's "point de départ."^[59] Just after Wurtz' review was published, Wurtz wrote to Kekulé congratulating him on his recent call. He also sent a copy of the first issue of the Répertoire and asked if he would consent to being the correspondent for Belgium. In a letter dated 15 February 1859 Kekulé gladly consented, and used the opportunity to make an argument for his own priority. (He thought enough of the matter to write an exact copy of the letter for his own files; it is only this copy that has survived.)

Kekulé pointed out that in his thiacetic acid paper he had explicitly drawn attention to the monatomic character of hydrogen and chlorine and to the diatomic nature of oxygen and sulfur. This was more than a year before Wurtz' 1855 paper appeared. Kekulé continued

Your lovely investigations on the glycols are even more recent; so they can have served just as little as a point of departure for my views, and indeed I only remarked that the further development of my views only became possible through your experiments. As for the views of William-

son and Odling in the papers you cited, I gladly acknowledge, and did so in my publication, that the ingenious considerations of these chemists formed the point of departure for my views. But whether the idea of the basicity of the atoms as we presently conceive it is expressed "clearly" in these papers is less certain. In fact it doesn't seem so to me; I think that if one goes back that far, then no doubt Laurent and Gerhardt could be considered to be the originators of this idea.^[60]

Wurtz replied to Kekulé: "I assure you that had I known of those passages in your [1854] memoir I would not have written the little note to which you allude." He promised to rectify the error publicly. In his next published article, he added a footnote wherein he expressed regret at having unfairly omitted citation of Kekulé's 1854 paper.^[61]

About a month later, during Easter vacation of 1859, Kekulé had the opportunity to visit Wurtz in Paris. He described his trip to Erlenmeyer and raved about Wurtz: "Capital fellow!—it is not possible that two people could agree more closely with the general conception of a science than we two."^[62] In 1865 he counseled his student Ladenburg to continue his education in Paris with Berthelot, since "Wurtz is the same as I."^[63]

Shortly thereafter, the first fascicle of Kekulé's textbook appeared, and one of the first colleagues to whom he sent a complimentary copy was Wurtz. In the accompanying letter, Kekulé asked Wurtz, "if you should happen to have a free moment," for his opinion of the theory contained therein. Kekulé wrote

I think you will not consider it mere customary flattery when I tell you that I lay more value on your judgment than on that of most modern chemists. In principle I will enter no priority claim for theoretical views as long as this can possibly be avoided; I will also not, à la Couper, blow my own horn, trumpeting my views as a "new chemical theory"; I will leave it calmly to time and to the legal sense of others to demonstrate whether in the development of these views anything reverts to me, and how much. If I now request your opinion on this, it is because I believe that the viewpoint which the majority of chemists in the future will have can be little different irom the viewpoint that you have now.^[64]

In this fascicle Kekulé had dealt with Wurtz' research in a complimentary fashion. Wurtz had been the first, Kekulé affirmed, to have correctly interpreted Berthelot's work on the glycerides, and his "ingenious conceptions and brilliant discoveries" of the di- and trialcohols had subsequently served to develop the theory of polyatomic radicals.^[65] Wurtz replied to Kekulé's letter

I willingly recognize that you have gone further than your predecessors in seeking to consider the fundamental properties and basicity of the elements in the conceptions regarding the constitution of compounds and chemical formulas. I consider your idea of the polyatomic nature of carbon as both sound and fertile, susceptible of important development. I must also thank you for the role you ascribed to me in your chapter on polyatomic radicals. I would not wish to claim more that what you have attributed to me.^[66]

But in the same letter, Wurtz expressed frustration that his work, appreciated in Germany and England, was being neglected in his own country. This frustration explains why, despite his agreeable concessions to Kekulé, during the next few years he repeatedly (though carefully and with proper appreciation for the work of others) urged his own priority in certain respects.^[67] In brief, Wurtz thought he deserved credit in two senses. The subatomic speculations in his 1855 article, and the associated exposition of "multiply condensed hydrogen" as the schematic basis for multiple and polyatomic types, provided a general theoretical foundation for valence that had not hitherto existed. This contribution was historically significant and heuristically valuable even if the theoretical details were still speculative, indeed, even if they were false.

More importantly, Wurtz thought that his elucidation of the glyceryl radical and his discovery of glycols offered the first real experimental confirmation of the existence of polyatomic radicals and that it properly illustrated their significance and fruitfulness. He was not the first to introduce the idea of polyatomic radicals, he was happy to concede, but his work provided the completion and perfection of the theory explaining them. Moreover, Wurtz believed that his proof of the existence of polyatomic radicals constituted a solid argument for the existence of polyatomic elements.

But these were relatively small merits compared to the great synthesis and advancement made by Kekulé, and he always granted Kekulé full credit for the application of valence ideas to carbon itself (even urging the superiority of Kekulé's claim to that of his own student, Couper). There is no question that Wurtz recognized the fundamental breakthrough achieved by Kekulé, and he no doubt felt chagrined that he had not gotten there first. In the fall of 1858, presumably under the influence of Kekulé's articles, he finally decided to adopt the atomic weight notation of Gerhardt, Laurent, Williamson, Odling. and Kekulé.

Wurtz' first paper using the new formulas—the same paper in which he corrected his earlier omission of Kekulé's 1854 valence argument

for oxygen and sulfur—was a long review article on glycols, in which a great deal about structure theory appears. Wurtz stated that Williamson's work had had "a large part in the development of this theory," but that Kekulé's "extremely important theoretical article" of 1858 had put the pieces together, literally as well as figuratively. The principle behind the theory was to place the idea of bonds (liens ) between polyatomic elements and radicals in the foreground. The new theory not only unified radical and type theories, but it also provided the basis for demonstrating the unity of inorganic and organic chemistry. According to Wurtz, organic radicals "offer more points of attack" than inorganic radicals, which lends a greater "fragility" to "the molecular edifice," otherwise they are the same. "The ideas, the hypothesis if you will, which express the molecular structure of the first, are exactly applicable to the second."^[68]

Although since 1853 he had openly defended the reformed chemistry without adopting the new formulas, from this time onward Wurtz exhibited total commitment to the reforms and a missionary zeal for propagating the work of Laurent and Gerhardt, both of whom had since died before the age of forty-five. In 1860 he was a principal organizer, along with Kekulé and Weltzien, of the Karlsruhe Congress, whose subtext was reform. He also presented a formal lecture before the Société Chimique on his discovery of the glycols, using it as a pretext for evangelism. In 1862 he delivered an "Éloge" in memory of both Laurent and Gerhardt to the Société des Amis des Sciences,^[69] again taking the role of advocate as well as obituarist. In the same year he spoke before the Chemical Society of London "On Oxide of Ethylene" and its theoretical implications.

In 1863 he presented a book-length series of lectures to the Société Chimique covering all of chemical theory, which was subsequently printed in both French and English. In 1864 he lectured at the Collège de France with similar intent and results. In 1866 he began publishing his Leons élémentaire de chimie moderne , using the reformed weights. Finally, in 1868 Wurtz began to publish a detailed dictionary of chemistry similar to the Handwörterbuch der reinen und angewandten Chemie , the first fascicles of which consisted of a history of chemical theory. Both the history and the dictionary itself proved to be clever means of propagating the new doctrines. However, despite all these efforts, many in France continued to use the older equivalents long after they were rejected elsewhere. Wurtz tried one last time, writing a historical precis of the atomic theory in 1879. But it was only after his death in 1884 that the ultimate victory for atomic weights in France was won.^[70] ç

One more claimant to structure-theoretical ideas, Aleksandr

Mikhailovich Butlerov, requires discussion. A student of Nikolai Zinin and Karl Klaus, Butlerov had already been teaching for nine years at the University of Kazan when he took a nine-month trip to western Europe in 1857-1858. He met with Kekulé twice during this period and spent five months in Wurtz' laboratory, at just the time when Couper was formulating his ideas on structure theory. He was already then theoretically active, as demonstrated by a presentation he gave to the Société Chimique in February 1858 and a commentary on Couper's theory published in Liebig's Annalen later that year. At this time, Butlerov's thinking on the constitutions of organic compounds was similar not only to Couper's but also to Kekulé's, Frankland's, and Kolbe's. However, he took exception to Couper's too absolute formulations and to his complete rejection of type theory in both its older (Dumas') and newer (Gerhardt's) forms. He said he much preferred Kekulé's version of structure theory, which in any case had appeared slightly earlier than Couper's.^[71]

On a second trip to western Europe in 1861, Butlerov gave what proved to be an influential paper at the Naturforscherversammlung in Speyer. He now disavowed his earlier rejection of Couper, affirming with the Scotsman that both type theories must be completely discarded. He criticized Kekulé and other structuralists for retaining vestiges of types that had done nothing but introduce confusion and inconsistencies. What was necessary at this point in the science was to apply the "theory of chemical structure" in a thoroughly consistent fashion. "Chemical structure," in Butlerov's definition, can be considered to be equivalent to an application of Kekulé's "atomicity of the elements," with an added epistemological caution that actual physical constitutions (spatial arrangements) were as yet inaccessible to chemists, who could only determine apparent structures from purely chemical phenomena. Kekulé, Kolbe, Wurtz, and others had been making a similar caution, but it was Butlerov's vocabulary that caught on.^[72]

Soviet historians of chemistry during the Stalin and Khrushchev periods attempt to show that Butlerov's concept was unique, fruitful, and modern, hence that he was the founder of the "classical theory of chemical structure." This thesis is vitiated by the fact that Butlerov himself never made such a claim, merely asserting (vocally and repeatedly) that from the beginning he had called for a clear and consistent application of the theory, that he had been one of the most successful of the first generation of structural chemists, and that his contributions had often been neglected. All of these claims were accurate. In chapter 10 we will return to Butlerov and discuss his relationship with Kolbe.

All of the nuanced and often overly subtle priority skirmishes con-

cerning the origins of structure theory highlight the complexity of this period of rapid theoretical evolution. The speed of development and the near coincidence of multiple contributions is partially explained by the circumstances outlined at the beginning of this chapter. Valence and structure were certainly in the air breathed by chemists of the 1850s, a point noted by Kekulé in his autobiographical sketch. One critical thesis, however, remained unelaborated until Kekulé's structure theory papers appeared, a thesis upon which all of structural chemistry depended: the self-linking of carbon atoms. Why was this one crucial idea so relatively slow to emerge when it seems to be a simple application of valence?

The answer apparently is that electrochemical dualism was not completely dead, even among the "typists" of the 1850s, and for good reason. Among the followers of Laurent and Gerhardt, it may have become accepted that atoms of hydrogen, oxygen, nitrogen, and chlorine form elemental molecules by combining with themselves, but this phenomenon was troublesome since it remained theoretically inexplicable. A further expansion of the concept to carbon in nonelemental molecules must have stuck in the craw of even the most modern reformers. After all, it is an understandable and even necessary tendency for scientists to attempt to explain the microworld by referring to the macroworld, hence to invoke macroscopic forces such as gravity, or more commonly, coulombic attraction and repulsion, to explain chemical forces.

Although some typists (Brodie and Couper, for example) attempted to develop electrochemical theories that might account for at least some cases of self-combinations of atoms, these attempts never achieved much success.^[73] Indeed, one reason for the general retreat into positivism of European chemists around 1840 had been the perceived general failure of electrochemical theory and the absence of any fundamental analog to replace it. Little had changed during the following two decades. The key to the discovery of carbon self-linking—and hence, the key to structure theory—had to be a certain willingness to suspend belief in any fundamental explanatory analogy to chemical affinity and simply to trust the direction that the leading systematists of chemical reactions were heading. In other words, it required a certain empiricism, or even positivism, of outlook, and Kekulé clearly possessed such a vision.

Ironically, though, it was precisely because of Kekulé's realist tendencies, in the form of a structuralist conviction of the ontological reality of chemical atoms and the physical linking function that they can perform (a conviction derived from Williamson and Wurtz), that he was able to conceive of a theory of chemical structure . Kekulé himself

thought that this realist tendency could be traced back to his architectural training, which had engendered in him an "irresistible necessity for visualizability (Anschaulichkeit )" and had led him to become the architect of molecular rather than of wood, brick, or stone edifices.^[74]