Exploring Atomic Ecologies:
Erlenmeyer and Kolbe

Ultimately, what made the theory of structure so powerful was an aspect that brought it closer to the physicists' scientific culture, namely, its ability to be axiomatized. In its simplest form, it could be reduced to a series of simple algorithmic or combinatorial rules, whereby possible molecular structures could be schematically derived by following straightforward valence assignments. The theory in this simple, powerful form is roughly the same as what Kekulé and Wurtz referred to in the late 1850s as the theory of "atomicity of the elements." It had roots in and similarities to both radical and type theories, but it was cleaner and more mechanical and was manipulable by simple algorithms rather than complex chemical intuition.

For example, most radical theorists regarded radicals as irreducible molecular moieties that were to be considered as wholes. A simple valence bond between two atoms was foreign, even antithetical, to the theory. Similarly, the type theory suggested more than mere genetic relationships among series of compounds: it asserted familial chemical similarities among members of the same type. This was one of the central claims of the theory that Kolbe (and others) began to contest in the 1840s and early 1850s. It was also what Wurtz was referring to when he wrote in 1861, "In a word, the idea of types is an artifice, a pure convention. In my opinion it must be subordinated to a more fundamental notion, that of the atomicity of the elements ."^[28]

Although Kekulé and Couper laid the foundation for the theory of atomicity of elements in 1857-1858, even advocates of this theory did not immediately eliminate old type/radical thought patterns from their

minds and begin thinking in the simple structural combinatorics that would soon become second nature to organic chemists. There was a certain period of transition for the field, during which the language and thoughts of leading theorists were somewhere in between the two stages. By the time Hofmann was composing his Introduction to Modern Chemistry in 1865, this transition period was about over, and Hofmann could correctly refer there to the "revolution" that had taken place in the science.^[29]

By the late 1860s, then, many if not most leading theoreticians had become full structuralists and thereafter thought largely in terms of those simple algorithmic patterns. However, this does not mean that organic chemistry had become a simple arithmetic problem, as Kolbe often charged. The schematic simplicity of the theory had to be compared and constantly checked and modified against a sometimes bewildering array of chemical peculiarities on the actual atomic—molecular level. The empirical contingency of atoms and their idiosyncratic chemical interrelationships contrasted starkly with the schematic simplicity of the theory. In effect, chemists had to emulate the naturalists once more, in exploring and determining what amounted to seemingly arbitrary (or at least then unmotivated) "ecological" relationships among the atoms in a molecule. In such a situation, it was vital that chemists maintain open attitudes. The theory was an indispensable guide in a general sense, but empirical flexibility was the most pragmatic course for questions of detail.

Nothing could have predicted, for instance, that "double bonds" would become the best way to view the relationship between adjacent carbon atoms in olefins, as opposed to maintaining free carbon affinities or two affinities satisfying each other, or even reducing the actual valence. Nothing could have predicted the eventually established circumstance that the four valences of carbon are all chemically equivalent. It was not at all clear why certain simple structures perfectly consistent with the theory, such as OCH₂ and C(OH)₄ , resisted all attempts at realization. Kolbe asserted as early as 1850 that the carbonyl carbon of acetic acid was bound both to oxygen and to alcohol functions. In 1859, Kekulé made the same statement in structural terms and argued that this atomic arrangement defined organic acid behavior in the general case. But, again, exactly why this was true remained mysterious.

Chapter 9 examined the work of some of the leading organic chemists of the 1860s in exploring some of these contingent "chemical ecologies" and pondered the effects such activity had in reshaping the character of the science. This section and the next continue this line of investigation, concentrating on the thoughts and writings of two

of the most avid converts to structural ideas: Emil Erlenmeyer and Aleksandr Mikhailovich Butlerov. Both men had interesting relationships with Hermann Kolbe.

A former apothecary and a student of Liebig and Will, Erlenmeyer first entered our story in chapter 7.^[30] Closely associated with Kekulé as a fellow Privatdozent in Heidelberg in 1856-1858, Erlenmeyer was one of the most inventive of the early structural chemists. In 1859, Kekulé trusted Erlenmeyer so implicitly that he asked him to perform all copyediting and proofreading for the first sheets of his textbook, then being printed, and he did not even want the final version to be sent back to him before printing. Predictably, this procedure created friction between the two men.^[31]

In 1858, Kekulé in collaboration with three other Heidelberg colleagues decided to found a new critical journal devoted principally to book reviews of the scientific literature and entitled Kritische Zeitschrift für Chemie, Physik und Mathematik . Kekulé's intent, as he wrote to Liebig, was to create a "dam" against the flood of "Schmierliteratur" then afflicting science.^[32] However, neither Liebig nor any other first-rank chemist lent his name publicly to the enterprise.

Kekulé had departed for Ghent long before even the first volume of this journal was complete, and Erlenmeyer took over the chief editor's role. From this time on, the journal restricted its scope to chemistry alone and began to publish reprinted (and occasionally original) articles. Beginning with volume three in 1860, Erlenmeyer took sole command of editorial production. The dramatic increase in the pace of chemical research and in the number of chemists, Erlenmeyer noted in a preface to that volume, meant that researchers were now in constant danger of unintentionally trespassing on each other's work. All the more vital was it that an organ be established that provided for rapid publication of preliminary results and small projects. Consequently, the Zeitschrift was henceforth to appear twice monthly in numbers of two sheets (thirty-two pages) each.^[33] What Erlenmeyer did not say publicly, but in private letters to colleagues in an attempt to boost circulation, was that he and others were also growing increasingly impatient with the slow publication schedule of Liebig's Annalen . His new model for the Zeitschrift was the Comptes rendus in Paris.^[34]

Kolbe began to send Erlenmeyer offprints of his Annalen articles for reprinting. In 1862 they began a friendly scientific correspondence, and by the end of the next year this had developed into a real friendship, including regular visits. Kolbe perceived enough commonality in their views that he began to think of Erlenmeyer as a compatriot.^[35] "I do not doubt," he wrote, "that my and your views will in time come into

their own and win general approval. Without this certainty the isolation in which I now find myself with my views would have become unendurable to me."^[36]

According to Erlenmeyer's later opinion, both he and Kekulé remained advocates of type theory (that is, they were not full structuralists) when Kekulé was in Heidelberg, even after he had enthusiastically adopted the point of view advocated in Kekulé's 1857-1858 papers. His conversion from this "one-sided" viewpoint to true structuralism came only late in 1861. The problem with the earlier viewpoint, Erlenmeyer said, was that its advocates always strove to identify a single de-terminative type, or atom defining the type, that could be used unambiguously to classify and characterize the molecule. He thought the correct viewpoint was to regard all pieces of a molecule, all of its radicals and atoms, to be ontologically equal, so that a molecule might be considered to be schematically dissectable in any possible fashion.^[37]

Examination of volumes four and five (1861 and 1862) of Erlenmeyer's Zeitschrift supports this autobiographical contention. In a paper published in April 1861, Erlenmeyer stated that he considered Kolbe as much a typist as Williamson, Gerhardt, and Wurtz, and he argued, in a Gerhardtian vein, that a significant accomplishment of the type theory was to have shown that one cannot prove absolute molecular constitutions. However, the goal of the paper was to argue a point that Kolbe would have considered constitutional, namely, to demonstrate that all alcohols and acids are monobasic.^[38] Erlenmeyer's thesis was pure Kolbean theory, and it would seem that at this time he would have described himself as a follower simultaneously of Gerhardt, Kekulé, and Kolbe, creating a useful syncretic viewpoint from these various influences.

However, early the following year Erlenmeyer began to publish a series of remarkable theoretical ideas that shows significant evolution from the previous year, ideas that are quite in the spirit of full structuralism. In a fervent defense of chemical theory in general and the newer theories in particular, Erlenmeyer provided one of the best early statements of the basic assumptions of the theory of atomicity of elements. The article contains the first suggestion (if not positive assertion) of the existence of carbon-carbon double bonds in olefins. He also proposed that the various affinity units (valences) of an atom have differing strengths, for on this assumption such otherwise anomalous compounds as carbonic oxide and nitrous gas could be assimilated into the theory of atomicity, since in such compounds not all affinities of some atoms are engaged with those of other atoms.^[39] Soon thereafter, Erlenmeyer developed the idea, apparently derived from Wurtz and Kekulé, that n-valent atoms are accretions of n-monovalent subatoms,

the subatoms differing in their absolute and relative affinities toward subatoms of various other atoms. This was a potentially powerful theory, but it proved to be empirically inaccessible and so was abandoned before the end of the decade.^[40]

In April 1862 Erlenmeyer published another critical article, this time on the lactic acid controversy. He rhetorically asked "die Herren Typiker" (among whom he obviously no longer counted himself) why they believed their type formulas so literally when doing so generated many misleading predictions. The proper approach, he said, was to consider a chemical formula as a "positional diagram" (Situationsplan ) for its constituent atoms, which specifies the "topographic positions" (topographische Lage ) of all of its components. So as not to be misunderstood, he added,

No one could imagine that with these formulas I intend to indicate the actual positions of the components. But I will remark that I very much intend thereby to express a topographic analogy of the derivative with the mother substance and to show that the mere reaction formulas of the typists no longer suffice, but rather we need to use relative constitutional formulas.^[41]

Erlenmeyer's "relative constitutional formulas" look to modern eyes, and no doubt to many of his contemporaries as well, like structural formulas. Indeed, he became the earliest German advocate of the graphic formulas pioneered by Couper and Crum Brown, which proliferated in his articles especially after 1865.

At the end of 1863, Erlenmeyer published a defense of Kolbe's views. Kolbe was being misunderstood by nearly everyone, Erlenmeyer argued; he was a typist in the full sense and had lost credit for some of his discoveries only because of the unusual form of his theories, while the content was quite conventional. Indeed, Kolbe's types were in some respects superior to other versions, Erlenmeyer asserted; they were more consistent, more genetically and synthetically derived, and more exclusively focused on carbon. Kolbe's biggest real differences with the other typists were his insistence on the goal of establishing absolute constitutional formulas and his use of equivalents for carbon and oxygen. There was a certain irony in each of these points of conflict, Erlenmeyer continued. For one, the typists' "reaction" formulas, despite the conventional rhetoric of their users, do ordinarily specify constitutions in Kolbe's sense, and more often than not they correspond essentially with Kolbe's formulations. Moreover, Kolbe may well be right in viewing carbon as C₂ (C = 6) rather than

C (C = 12). (Here Erlenmeyer was making reference to his own subatomic speculations.)^[42]

Kolbe was very grateful. After thanking Erlenmeyer, he added,

By undertaking to prepare the way for a better understanding of my views on the composition of organic compounds, against which there is widespread prejudice, you are taking on a somewhat dangerous task, since a large number of chemists want to know absolutely nothing about them, and will resent you for demanding that they think seriously about chemical questions. . . . Even though our views differ in points of detail, I am pleased that they are nearly the same regarding carbon.^[43]

Erlenmeyer was fast earning a reputation, and not altogether a favorable one, as a prolific but increasingly acerbic and incautious critic; even his formerly close relationship with Kekulé went on the rocks. It was partially for this reason that his journal was not flourishing. By the fall of 1864, there were only 150 subscribers (and more than half of them were Russians). He decided to give up his editorship and gave the journal over to three Göttinger Dozenten (Beilstein, Fittig, and Hübner).^[44] In its place, he did what many of his theoretically inclined colleagues were doing and began writing a textbook. In 1868 Erlenmeyer's future was finally secured (he was then a forty-three year old Privatdozent) by a call to the Munich Technische Hochschule.

Erlenmeyer maintained a close relationship with Kolbe until the latter's increasingly strong attacks on structure theory finally drew his fire. In June 1871, he wrote Kolbe saying that he was composing a public lecture to be given in honor of King Ludwig's twenty-fifth birthday. Without giving away details, he wanted, among other things, to persuade Kolbe "that your formulas are nothing more than my formulas, that at most a difference between them arises when we have a different view of the constitution itself."^[45]

The lecture was printed in the course of the summer, equipped with a ponderous 3500-word footnote dealing with Kolbe,^[46] and Erlenmeyer sent a copy of it to his friend. Kolbe was one of the few who were still contesting structure theory, Erlenmeyer noted, but what Kolbe did not realize is that operationally, his formulas were nearly identical to those of the structuralists. He was a type theorist in the strict sense, the sort that he (Erlenmeyer) and Kekulé had been before both began to operate as full structuralists about 1861. According to Erlenmeyer, the only substantial difference between Kolbe and the structural chemists was that Kolbe denied the existence of not only direct connections between carbon atoms (chain theory) but also the very concept of valence bonds.

Erlenmeyer argued that Kolbe's position was not logical or consistent. Kolbe averred that his theory was based upon atomicity, that carbon atoms were (almost always) tetratomic, and that methyl, for example, was monatomic because one of the carbon atomicity units was unsatisfied. Kolbe had angrily rejected the conclusion that it was precisely that fourth valence of carbon that formed the point of attack or attachment for the methyl radical. It was the methyl radical as a whole , he said, that substituted for the hydrogen of water to form methyl alcohol, of methane to form ethane, of formic acid to form acetic acid, and so on. However, Erlenmeyer protested, it was Kolbe himself who had begun as early as 1850 to speak of the "point of attack" of chemical action, in this case referring to the carbonyl carbon atom of acetic acid.

Erlenmeyer also chided Kolbe for the one-sidedness that characterized all genuine type theories. Kolbe's search for the dominating "Grundradikal" or "Hauptkohlenstoffatom" resulted in an over-emphasis on one radical or atom to the detriment of every other one in the molecule. It was downright silly to search for a single carbon atom in an alcohol that confers "family character" (Gattungscharakter) on the compound; it is rather functional group analysis (emphasis on hydroxyl groups, in this case) that has chemical significance. Kolbe had used various typographical techniques, including bolding, boxes, and size differentials, to distinguish among the carbon atoms in organic compounds. Were the alcohol of the eighteen-carbon stearic acid to be prepared, Erlenmeyer concluded, and were Kolbe to use sizing to distinguish all of the carbons, one would need a microscope to see the eighteenth.^[47] Erlenmeyer summarized his position in the following way:

Kolbe has remained a typist while most other chemists have abandoned the type viewpoint. He considers chemical compounds not according to their entire composition and constitution and all properties determined from these factors, but only as alcohols, as aldehydes, as carbonic acids, as sulfonic acids, as amines, etc. etc.^[48]

That this conservatism had operational hazards Erlenmeyer was also able to show. Kolbe had predicted as a consequence of his point of view that the two chlorine atoms of the unknown compound 1,3-dichloropropane would be found to be chemically distinct, whereas chain theory (structure theory) would predict them to be chemically identical. Erlenmeyer succeeded in testing this prediction, at least for the bromine analog: he had the compound synthesized and reported

that both bromine atoms hydrolyzed simultaneously to form the dialcohol.^[49]

Erlenmeyer could hardly have devised a more stinging rebuke than to say that Kolbe was a one-sided type theorist and needed to think more seriously and more chemically about compounds and reactions, for this was precisely Kolbe's longstanding complaint against the type theorists and the structuralists. Kolbe's fury can only be imagined. In September he composed yet another general attack on structure theory for his new Journal für praktische Chemie . To Vieweg he commented, "In this piece I become personally unpleasant only once, namely against Erlenmeyer, a man who has not a trace of logic in his head. He challenged me and forced me to this. Liebig may not like it, but I cannot help that."^[50] In that event, however, Kolbe decided to suppress his comment.^[51] But he did not hesitate to disparage Erlenmeyer ever after in his private correspondence. "Once a pharmacist, always a pharmacist," he grumbled to Volhard.^[52]