Lamarck therefore opted for an intermediary solution. He would not yet dismember Vermes, but he would establish two subdivisions within the class: vers externes (external worms) for earthworms and their allies, and vers intestins (internal worms) for tapeworms and their relatives. He stressed the simple anatomy of the parasitic subgroup, and defended their new name as a spur to further study, while arguing that knowledge remained insufficient to advocate a deeper separation:
It is very important to know them [the internal worms], and this name wall facilitate their study. But aside from this motive, I also believe that such a division is the most natural… because the internal worms are much more imperfect and simply organized than the other worms. Nevertheless, we know so little about their origin that we cannot yet make them a separate order.
At this point, the crucial incident occurred that sparked Lamarck to an irrevocable and cascading reassessment of his evolutionary views. He attended Cuvier’s lecture during the winter of 1801–2 (year 10 of the revolutionary calendar), and became convinced, by his colleague’s elegant data on the anatomy of external worms, that the extensive anatomical differences between his two subdivisions could not permit their continued residence in the same class. He would, after all, have to split the heart of Vermes. Therefore, in his next course, in the spring of 1802, Lamarck formally established the class Annelida for the external worms (retaining Vermes for the internal worms alone), and then separated the two classes widely by placing his new annelids above insects in linear complexity, while leaving the internal worms near the bottom of the ladder, well below insects.
Lamarck formally acknowledged Cuvier’s spur when he wrote a history of his successive changes in classifying invertebrates for the Philosophic zoologique of 1809:
Mr. Cuvier discovered the existence of arterial and venous vessels in distinct animals that had been confounded with other very differendy organized animals under the name of worms. I soon used this new fact to perfect my classification; therefore, I established the class of annelids in my course for year 10 (1802).
The handwritten note and drawing in Lamarck’s 1801 book, discussed and reproduced earlier in this essay, tells much the same story—but what a contrast, in both intellectual and emotional intrigue, between a sober memory written long after an inspiration, and the inky evidence of the moment of enlightenment itself!
But this tale should now be raising a puzzle in the minds of readers. Why am I making such a fuss about this particular taxonomie change—the final division of Vermes into a highly ranked group of annelids and a primitive class of internal worms? In what way does this alteration differ from any other previously discussed? In all cases, Lamarck subdivided Linnaeus’s class Vermes and established new phyla in his favored linear series—thus reinforcing his view of evolution as built by contrasting forces of linear progress and lateral adaptation. Wasn’t he just following the same procedure in extracting annelids and placing them on a new rung of his ladder? So it might seem—at first. But Lamarck was too smart, and too honorable, to ignore a logical problem direcdy and inevitably instigated by this particular division of worms—and the proper solution broke his system.
At first, Lamarck did treat the extraction of annelids as just another addition to his constandy improving linear series. But as the years passed, he became more and more bothered by an acute problem, evoked by an inherent conflict between this particular taxonomie decision and the precise logic of his overarching system. Lamarck had ranked the phylum Vermes, now restricted to the internal worms alone, just above a group that he named radiaires—actually (by modern understanding) a false amalgam of jellyfishes from the coelenterate phylum and sea urchins and their relatives from the echinoderm phylum. Worms had to rank above radiates because bilateral symmetry and directional motion trump radial symmetry and an attached (or not very mobile) lifestyle— at least in conventional views about ladders of progress (which, of course, use mobile and bilaterally symmetrical humans as an ultimate standard). But the parasitic internal worms also lack the two most important organ systems— nerve ganglia and cords, and circulatory vessels—that virtually define complexity on the traditional ladder. Yet echinoderms within the “lower” radiate phylum develop both nervous and circulatory systems. (These organisms circulate sea water rather than blood, but they do run their fluids through tubes.)
If the primary “force that tends incessandy to complicate organization” truly works in a universal and exceptionless manner, then how can such an inconsistent situation arise? If the force be general, then any given group must stand fully higher or lower than any other. A group cannot be higher for some features, but lower for others. Taxonomie experts cannot pick and choose. He who lives by the line must die by the line.
This problem did not arise so long as annelids remained in the class of worms. Lamarck, after all, had never argued that each genus of a higher group must rank above all members of a lower group in every bodily part. He only claimed that the “principal masses” of organic design must run in pure linear order. Individual genera may degenerate or adapt to less complex environments in various parts—but so long as some genera display the higher conformation in all features, then the entire group retains its status. In this case, so long as annelids remained in the group, then many worms possessed organ systems more complex than any comparable part in any lower group—and the entire class of worms could retain its unambiguous position above radiates and other primitive forms. But with the division of worms and the banishment of complex annelids, Lamarck now faced the logical dilemma of a coherent group (the internal parasitic worms alone) higher than radiates in some key features but lower in others. The pure march of nature’s progress—the keystone of Lamarck’s entire system—had been fractured.
Lamarck struggled with this problem for several years. He stuck to the line of progress in 1802, and again—for the last time, and in a particularly uncompromising manner that must, in retrospect, represent a last hurrah before the fall—in the first volume of his seminal work, Philosophic zoologique, of 1809. But honesty eventually trumped hope. Just before publication, Lamarck appended a short chapter of “additions” to volume two of Philosophie zoologique. He now, if only tentatively, floated a new scheme that would resolve his problem with worms, but would also unravel his precious linear system.
Lamarck had long argued that life began with the spontaneous generation of “infusorians” (single-celled animals) in ponds. But suppose that spontaneous generation occurs twice, and in two distinct environments—in the external world for a lineage beginning with infusorians, and inside the bodies of other creatures for a second lineage beginning with internal worms? Lamarck therefore wrote that “worms seem to form one initial lineage in the scale of animals, just as, evidently, the infusorians form the other branch.”
Lamarck then faced the problem of allocating the higher groups. To which of the two great lines.does each belong? He presented his preliminary thoughts in a chart—perhaps the first evolutionary branching diagram ever published in the history of biology—that directly contradicted his previous image of a single ladder. (Compare this figure with the version presented earlier in this essay, taken from volume one of the same 1809 work.) Lamarck begins (at the top, contrary to current conventions) with two lines, labeled “infusoires” (single-celled animals) and “vers” (worms). He then inserts light dots to suggest possible allocations of the higher phyla to the two lines. The logical problem that broke his system has now been solved—for the radiaires (radiate animals), standing below worms in some features, but above in others, now rank in an entirely separate series, directly following an infusorian beginning.
When mental floodgates open, the tide of reform must sweep to other locales. Once he had admitted branching and separation at all, Lamarck could hardly avoid the temptation to apply this new scheme to other old problems. Therefore, he also suggested some substantial branching at the end of his array. He had always been bothered by the conventional summit of reptil
es to birds to mammals, for birds seemed just different from, rather than inferior to, mammals. Lamarck therefore proposed (and drew on his revolutionary chart) that reptiles branched at the end of the series, one line passing from turtles to birds (oiseaux) to monotrèmes (platypuses, which Lamarck now considers as separate from mammals), the other from crocodiles to marine mammals (labeled m. amphibies) to terrestrial mammals. Finally, and still in the new spirit, he even posited a threefold branching in the transition to terrestrial mammals, leading to separate lines for whales (m. cétacés), hoofed animals (m. ongulés), and mammals with nails (m. onguiculés), including carnivores, rodents, and primates (including humans).
Finally, Lamarck explicitly connected the two reforms: the admission of two sequences of spontaneous generation at the bottom, and a branching among higher vertebrates at the top: “The animal scale begins with at least two branches; in the course of its extent, several branches seem to end in different places.”
After Philosophic zoologique of 1809, Lamarck wrote one additional major book on evolution, the introductory volume (1815) to his Histoire naturelle des animaux sans vertèbres. Here, he abandoned all the tentativeness of his 1809 revision, and announced his conversion to branching as the fundamental pattern of evolution. In direct contradiction to the linear model that had shaped all his previous work, Lamarck stated simply, and without ambiguity:
Dans sa production des differents animaux, la nature n’a pas executé une série unique et simple.
[In its production of the different animals, nature has not fashioned a single and simple series.]
He then emphasized the branching form of his new model, and explained how the division of worms, inspired by Cuvier’s observations, had broken his former system and impelled his revision:
Lamarck’s first depiction of a branching model for the history of life. From the appendix to Philosophic zoologique (1809).
The order is far from being simple; it is branching [rameux] and even appears to be constructed of several distinct series…. The animals that belonged to the class of worms display a great disparity of organization…. The most imperfect of these animals arise by spontaneous generation, and the worms [now restricted to vers intestins, with annelids removed] truly form their own series, later in origin than the one that began with infusorians.
Lamarck’s third and last chart (reproduced here from his 1815 volume) shows how far he had progressed both in his own confidence, and in copious branching on his new tree of life. He tides the chart “presumed order of the formation of animals, showing two separate and subbranching series.” Note how the two major lines of separate spontaneous generation—one beginning with infusorians, the other with internal worms—are now clearly marked and separated. Note also how each of the series also divides within itself, thus establishing the process of branching as a key theme at all scales of the system. The infusorian line branches at the level of polyps (corals and jellyfish) into a line of radiates and a line terminating in mollusks. The second line of worms also branches in two, leading to annelids on one side and insects on the other. But the insect Une then splits again (a tertiary division) into a lineage of crustaceans and barnacles (labeled cirrhipèdes) and another of arachnids (spiders and scorpions).
Finally, we must recognize that these major changes do not only affect the overt geometry of animal organization. The conversion from linearity to branching also—perhaps even more importandy—marks a profound shift in Lamarck’s underlying theory of nature. He had based his original system, defended explicidy and vociferously until 1809, on a fundamental division of two independent forces—a primary cause that builds basic anatomies in an unbroken line of progress, and a subsidiary lateral force that draws single lineages off the line into byways of immediate adaptation to local environments. A set of philosophical consequences then spring from this model: the predictable and lawlike character of evolution lies patent in the primary force and its ladder of progress reaching to man, while accidents of history (leading to local adaptations) can then be dismissed as secondary and truly independent from the overarching order.
Lamarck’s fully developed tree of lifefrom 1815.
But the branching system destroys this neat and comforting scheme. First of all, the two forces become intermingled and conflated in the branching itself. We can no longer distinguish two independent and orthogonal powers working at right angles. Progress may occur along any branch to be sure, but the very act of division implies an environmental impetus to split the main line—and Lamarck had always advocated a complete and principled distinction between a single and inexorable main line and the numerous minor deviations that can draw off a long-necked giraffe or an eyeless mole, but can never disrupt or ramify the major designs of animal life. In the new model, however, environment intrudes at the first construction of basic order—as one group arises spontaneously in ponds, and another inside the bodies of other creatures! Moreover, each of the two resulting lines then branches further, and unpredictably, under environmental impetuses that were not supposed to derail the force of progress among major groups—as when insects split into a terrestrial line of arachnids and a marine line leading to crustaceans and barnacles.
Second, the forces of history and natural complexity have now triumphed over the scientific ideal of a predictable and lawlike system. The taxonomy of animals could no longer embody an overarching plan of progress, illustrating the fundamental order, harmony, and predictable good sense of the natural world (perhaps even the explicit care of a loving deity, whose plans we may hope to understand because he thinks as we do). Now, the confusing, particular, local, and unpredictable forces of complex environments hold sway, ready at any time to impose a deviation upon any group with enough hubris to suppose that Emerson’s forthcoming words could describe their inevitable progress:
And striving to be man, the worm
Mounts through all the spires of form.
VI. LAMARCK’S EPILOGUE AND MY OWN
Following his last and greatest treatise on the anatomy of invertebrate organisms, Lamarck published only one other major work—Analytic System of Positive Knowledge About Man (1820). This rare book has not been consulted by previous historians who traced the development of Lamarck’s changing views on the classification of animals. Thus, traditional accounts stop at Lamarck’s 1815 revision, with its fundamental distinction between two separate lineages of spontaneous generation. The impression therefore persists that Lamarck never fully embraced the branching model, later exemplified by Darwin as the “tree of life”—with a common trunk of origin for all creatures and no main line of growth thereafter. Lamarck had compromised his original ladder of progress by advocating two separate origins, but he could continue to stress linearity in each of the resulting series.
But his 1820 book, although primarily a treatise on psychology, does include a chapter on the classification of animals—and I discovered, in reading these pages, that Lamarck did pursue his revisionary path further, and did finally arrive at a truly branching model for a tree of life. Moreover, in a remarkable passage, Lamarck also recognizes the philosophical implications of his full switch by acknowledging a reversal in his ranking of natural forces in one of the most interesting (and honorable) intellectual conversions that I have ever read.
Lamarck still talks about forces of progress and forces of branching, and he does argue that progress will proceed along each branch. But branching has triumphed as a primary and controlling theme, and Lamarck now frames his entire discussion of animal taxonomy by emphasizing successive points of division. For example, consider this epitome of vertebrate evolution:
Reptiles come necessarily after fishes. They build a branching sequence, with one branch leading from turdes to platypuses to the diverse group of birds, while the other seems to direct itself, via lizards, toward the mammals. The birds then … build a richly varied branching series, with one branch ending in birds of prey.
(In previous models, Lamarck had viewed birds of
prey as the top rung of a single avian ladder.)
But much more radically, his 1815 model based on two lines of spontaneous generation has now disappeared. In its place, Lamarck advocates the same tree of life that would later become conventional through the influence of Darwin and other early evolutionists. Lamarck now proposes a single common ancestor for all animals, called a monad. From this beginning, infusorians evolve, followed by polyps, arising “direcdy and almost without a gap.” But polyps then branch to build the rest of life’s tree: “instead of continuing as a single series, the polyps appear to divide themselves into three branches”—the radiates, which end without evolving any further; the worms, which continue to branch into all phyla of segmented animals, including annelids, insects, arachnids, crustaceans, and barnacles, each by a separate event of division; and the tunicates (now regarded as marine organisms closely related to vertebrates), which later split to form several Unes of mollusks and vertebrates.
Lamarck then acknowledges the profound philosophical revision implied by a branching model for nature’s fundamental order. He had always viewed the linear force of progress as primary. As late as 1815, even after he had changed his model to permit extensive branching and two environmentally induced sequences of spontaneous generation, Lamarck continued to emphasize the primary power of the linear force, compared with disturbing and anomalous exceptions produced by lateral environmental causes, called l’infiuence des cir-constances. To restate the key passage quoted earlier in this essay: