"Kekulé Always Rides a Fiery Steed"
Although by 1862 Kolbe was encountering difficulties in substantiating his early aromatic conjectures and had partly repudiated them by 1865, he continued to work (and to assign his students projects) on aromatic compounds, and his laboratory produced many novel substances in the early 1860s. In the spring of 1863, two of his students in Marburg simultaneously and independently discovered the first case of triple isomerism in the aromatic series. Konstantin Zaitsev (in German transliteration, C. Saytseff), one of the first of numerous Russian students, never made a name for himself in chemistry, but he was the brother of the much more famous Aleksandr Mikhailovich Zaitsev, who arrived at Kolbe's lab a semester later. Early in 1863, Konstantin prepared from anisic acid (modern p -methoxybenzoic acid) a new hydroxybenzoic acid that was demonstrably distinct both from salicylic acid and also from Gerland's "Oxybenzoësäure." He named it "Paraoxybenzoësäure," the prefix simply indicating that it was an isomer.[23]
About the same time at a neighboring lab bench, Georg Fischer (who is even more obscure than K. Zaitsev) obtained the same compound starting from toluene through nitrodracylic (modern p -nitrobenzoic) and a novel para-aminobenzoic acid. The fact that there were now three hydroxybenzoic acids suggested to Fischer that, in addition to benzoic and salylic acids, there must be a yet undiscovered third isomer, "Parabenzoësäure."[24] Zaitsev's and Fischer's articles were published back to back in the Annalen . In that same issue, Beilstein argued that the aminobenzoic acids were now three in number. Shortly thereafter, H. Hlasiwetz and L. Barth revealed the existence of a third dihydroxybenzene, christened "resorcinol," and N. Sokolov concluded that three nitrobenzoic acids could be sharply distinguished. Finally, in a communication dated August 1864, Beilstein announced the discovery of a third distinct chlorobenzoic acid and concluded that there must also be three bromobenzoic and three iodobenzoic acids.[25]
What all this means is that within the space of about three years, the isomer problem in the aromatic series had suddenly and radically
changed. In 1861, it seemed highly probable that there were two of each species, in other words, two isomers of benzene itself, two isomers of each monoderivative such as benzoic acid, and two of each diderivative such as salicylic acid. By the end of 1864, it was becoming clear that there was only one benzene and only one of each mono-derivative of benzene, but that there were in general three of each diderivative such as hydroxy benzoic or nitrobenzoic acid. Kolbe's conjectures were no longer viable, and Fischer noted plaintively that the "very important question" of explaining aromatic isomerism was not presently answerable. In extenuation of his mistaken identification of salylic acid, Kolbe likewise commented in a note written in March 1865 that he had been led astray by his urgent desire to explain the genuine generic isomer problem for aromatic compounds.[26] But it would appear that Kolbe had given up on the problem as hopeless.
It was none other than Kolbe's nemesis Kekulé who first succeeded in finding a satisfactory solution to this problem, and this was precisely as Kolbe was writing the remark just cited. Kekulé hinted at double bonds in benzene in 1858, and he later implied that he had already formulated his benzene theory at that time. This claim, however, was probably disingenuous because, as we have seen, he accepted the existence of salylic acid and of two isomeric benzenes in 1861. But the new discoveries of 1861-1864 dramatically altered the requirements for a successful theory. If we can believe his famous dream anecdote (and circumstances suggest that we can), he had the idea for his benzene theory around early 1862, but this was still too early in the chronology to found a successful empirically based theory. By the end of 1864, it was time.
Early in 1865 Kekulé published a paper (in French, since he was then working in Francophone Belgium) that argued that a cyclohexatriene formula for benzene could probably account for all the known isomeric relationships in the aromatic series. The "Kekulé formula" would allow only one benzene and one benzoic acid, but would appear to predict three isomers (today referred to as ortho, meta, and para isomers) of every disubstituted benzene. Kekulé expressed himself cautiously in this article and spent most of his time discussing how the theory could explain isomer relationships of side-chains rather than positional (ortho/meta/para) isomers. But he was more definite and more confident on the crucial question of positional isomers in a long German-language article published in Liebig's Annalen in February 1866. A definitive treatment was published in the sixth fascicle of his textbook, which appeared that summer.[27]
Kolbe read Kekulé's Annalen paper with interest (he probably had not seen the earlier French papers, since he rarely read French jour-
nals). "What do you think about Kekulé's new philosophy in the February issue of the Annalen ?" he asked Erlenmeyer. "He always rides a fiery steed."[28] We do not possess Erlenmeyer's reply to this question, but there can be little doubt that it was favorable, for an article appeared in the very next issue by Erlenmeyer praising Kekulé's theory as the "most probable" representation of aromatic compounds (although Erlenmeyer also made a number of suggestions modifying and extending the theory). Kolbe must have been astounded at the overwhelmingly favorable response to Kekulé's theory. Not only did students and friends of Kekulé, such as Erlenmeyer, Williamson, Müller, Beilstein, Baeyer, Ladenburg, L. Meyer, Dewar, Hübner, and Körner, become advocates of the theory but also many who had no direct personal relationship to him, such as Wilbrand, Watts, Naquet, Fittig, and V. Meyer. In fact, three former members of Kolbe's own research group became strong advocates of the new theory, namely, Graebe, Crum Brown, and Claus, and Kolbe's good friend and theoretical comrade Frankland also joined the crowd.[29]
Some historians of chemistry have been misled by the lengthy but often minor disagreements over certain structural details into thinking that Kekulé's theory was poorly accepted until close to the end of the century. In fact, the proposers of principal "rivals" to Kekulé's structure thought of themselves essentially as advocates of the theory who were offering only reinterpretations of it. Ladenburg, the author of the "prism" formula, always stressed the similarities of his and Kekulé's formulas. In 1874 he conceded that Kekulé's formula was in most respects superior to his own; in 1876 he stated that virtually all chemists had accepted Kekulé's theory, at least to a certain degree, and that the cyclohexatriene formula was "at least as appropriate, if not more so, than the prism." Even Kolbe conceded in 1874 that Kekulé's theory was "accepted by the great majority of chemists."[30]
Kolbe's concession was by no means a surrender. Despite his repudiation of salylic acid in 1865, he did not immediately abandon his belief that there are isomeric "phenyl" and "benzyl" radicals corresponding to the formula C12 H5 , reaffirming this idea two years later.[31] Appalled by the growing popularity of Kekulé's "chaining" theory of structure and deeply concerned that Kekulé's cyclohexatriene interpretation of benzene was already then being viewed by many as the successful capstone of that theory, Kolbe was desperate to produce a coup in the aromatic realm. His interest in aromatics was of even longer standing than Kekulé's, and he shared Kekulé's identification of the central desideratum, namely, a general theoretical explanation for isomerism. He found it particularly galling that Kekulé, who had
been a thorn in his side on many occasions and whom he considered an upstart and a "superficial chatterer," had now become his chief rival. Even many of his own advanced students and junior associates during the years 1866-1870, such as Ernst Carstanjen, Carl Graebe, Hermann Ost, and Henry Armstrong, disagreed with his views on aromatics, and some of them preferred Kekulé's.
It was in response to the arguments and urgings of Markovnikov in 1866-1867 and possibly also Graebe in the summer of 1868 that Kolbe finally accepted the larger atomic weights for carbon, oxygen, and sulfur, which by that time had been adopted by virtually all German chemists. The first occasion when Kolbe employed the new weights was in his dedicatory lecture for his new institute on the Waisenhaus-strasse, on 16 November 1868. The topic of this printed academic dissertation was his conception of the constitution of hydrocarbons; its centerpiece was a novel theory of aromatic compounds. Both were offered explicitly as alternatives to Kekulé's ideas. Indeed, Kolbe's second sentence averred that his ideas were "essentially different from the views of the chemists of so-called modern chemistry, and especially from those based on Kekulé's chain theory." He continued,
I do not intend here to subject the latter to a thorough criticism; I merely state clearly and succinctly, in order at once to remove any doubt concerning my estimation of its significance, that I consider the chain theory, just like the earlier type theory, to be an ephemera among recent chemical hypotheses, and its principles to be untenable and erroneous, without denying that it is ingeniously devised, and, as experience confirms, is well suited to blind younger chemists.[32]
The next few pages of Kolbe's lecture recounted his views on aliphatic hydrocarbons, in their current state of evolution. Kolbe's readers (as well as the readers of this book) had seen this material before.