Mivart's Dilemma Confronted

* This is Part 2 of a two-part installment of “Problematica.” It is about the problem of explaining novel structures in late nineteenth century evolutionary science. You can find the link to Part 1 here. (And Part 2.5 here!) Problematica is written by Max Dresow…

In the first part of this essay I made a perhaps-somewhat-controversial claim. After introducing three nineteenth century arguments against the idea that natural selection could explain the origin of characters, I said:

Mivart’s Dilemma [the inability of natural selection to account for the incipient stages of useful structures] played the most decisive role in structuring attitudes towards natural selection during the decades following Darwin’s death.

I take this to be controversial because Darwin’s struggles in the late nineteenth century are often put down to a faulty understanding of heredity: an interpretation that has led many to regard Fleeming Jenkins’s “swamping argument” as a biggest challenge. But Mivart’s dilemma was a more sophisticated bit of reasoning, and in the event, a more difficult one to answer. The implication is that the main locus of difficulty for Darwinism did not have to do with heredity per se. It had to do with the origin of novel structures.

Now I want to defend this claim, using as my weapon the chatty historical vignette. I have four of these to recommend to your consideration. Two concern general treatments of evolutionary theory. These are Herbert Conn’s Method of Evolution (1900) and Vernon Kellogg’s Darwinism To-day (1907). The other two describe research programs— those of William Bateson and Theodor Eimer, respectively. I will begin with the least well-known of these figures, the bacteriologist Conn, before proceeding to discuss Bateson, Kellogg, and Eimer in that order. In closing, I will offer a few remarks on the historiography of the post-Darwinian period, traditionally known as the “eclipse of Darwinism.”

Herbert Conn and the Problem of New Organs

Herbert William Conn was born in January 1859, the same year the Origin of Species exploded into public consciousness. As a teenager he read morphology at Johns Hopkins and became interested in evolution. Later he helped to found the Society for American Microbiologists, now the American Society for Microbiology (Noll et al. 2014). In addition, he penned one of the most thoughtful treatments of evolutionary theory to appear around the turn of the century, The Method of Evolution (1900). Broadly sympathetic to Darwinism, the work represented a shift from his earlier survey, Evolution of To-day (1886), in which he claimed that “Natural selection, or Darwinism, is… almost everywhere acknowledged as insufficient to meet the facts of nature, since many features of life are not explained by it” (Conn 1886, 243). Still, Conn had concerns.

Herbert Conn’s microscope (left) and some illustrations he made while studying under the important American zoologist W. Keith Brooks at Johns Hopkins (right). Conn’s dates are 1859–1917. (Photos from Noll et al. 2014)

Conn began The Method of Evolution with a frank appraisal of late nineteenth century evolutionary studies. “[While] scientists are agreed to-day… that the theory of descent represents the actual history of the organic world,” it is “probably not incorrect to say that upon the question of method [causes] there is to-day greater uncertainty of opinion and greater confusion than at any previous time” (Conn 1900, 10–11). The problem was not one of inadequate knowledge, however. To the contrary: “[it] is because the problem has not proved to be the simple one at first conceived, but to have many factors entering into it not at first expected,” that opinions varied in every conceivable direction (11). Notwithstanding this variation, Conn believed that researchers had every reason to be optimistic about the prospects of evolutionary science. Now that the major problems with competing evolutionary theories had been identified, scientists could get to work on solving them; and indeed, as Conn intended to show, this work was already underway.

Of the problems that had emerged since the early days of Darwinism, one in particular stood out: the problem of new characters. As Conn put it: “Selection cannot start the process of evolution… it is evident that minute variations on either side of a mean, leading to the selection of averages, does not account for the beginnings of organs” (1900, 139). Having absorbed the debates of the 1880 and ‘90s, Conn was aware of the pedigree of this argument, to say nothing of its relevance for contemporary evolutionary studies:

From the very outset of the discussion aroused by the theory of natural selection, it has appeared that one of the most serious difficulties in the way of accepting natural selection as a satisfactory solution of descent is the seeming impossibility of accounting for the beginning of organs. Selection explains elimination and the moulding of organs to new functions, but it doesn’t explain origin. This difficulty appealed to Darwin and he candidly discussed it, suggesting various facts looking toward a solution of the difficulty. But the same question has been raised by nearly everyone who has discussed selection, and it has been studied from various standpoints. (Conn 1900, 134)

What standpoints were these? As Conn knew, an important part of the difficulty was the inability of selection to seize upon the first rudiments of organs that would be useful in a more developed state. This was Mivart’s dilemma, and the typical response was to assert that organs do not originate de novo, but arise from preexisting structures via changes in function. So fins become limbs become flippers or wings— but at no point were tetrapod appendages non-functional, or even uni-functional. Consider that a proto-wing may continue to function as an arm, or may confer some other benefit (in reproduction, say) prior to its acquisition of a flight function. This was how Darwin dealt with the problem of new structures in the Origin (1859, 191), and the solution was consolidated and extended by Anton Dohrn in his Der Ursprung der Wirbelthiere, published 1875:

The transformation of an organ occurs through the succession of functions, the bearer of which remains one and the same organ. Each function is a resultant of many components, of which one forms the main, or primary function, while the others represent subsidiary, or secondary functions. The sinking of the main function and the rising of a subsidiary function changes the total function. The subsidiary function gradually becomes the main function, the total function becomes another one, and the result of the entire process is the transformation of the organ. (Dohrn 1875, 60, quoted in Bowler 1996, 163)

Anton Dohrn, photographed in 1883

Curiously, while this argument is now highly regarded, it did little to dispel Mivart’s Dilemma in the eyes of late nineteenth and early twentieth century biologists. Of course, everyone recognized that avian wings were derived from tetrapod forelimbs, which in their ancestral state were adapted to non-aerial functions. Yet a mystery remained: how did anything like a limb originate in the first place? Hypotheses abounded, attached to such prestigious names as Owen, Gegenbaur, Huxley, and Balfour. But none seemed quite adequate. After all, selection requires that an organ have some use; but “what could have produced even a rudimentary useful appendage on the sides of a body previously smooth?” (Conn 1900, 136). This led Conn to observe, twenty-five years after Dohrn’s Ursprung, that “in spite of all that has been said it can hardly be claimed by anyone that these difficulties have been removed by one or all the considerations that have been urged” (135). The difficulties, moreover,

[have] led some of our most thoughtful and observant naturalists to question seriously whether selection can be regarded as a vera causa, while it has convinced others that we can never find the explanation of descent by the study of natural selection, no matter how vigorously we pursue the subject, and that the only chance for further progress is in the study of variations themselves. (Conn 1900, 138, emphasis added)

In this statement— and particularly in the injunction to study variation directly— William Bateson’s influence is manifest. Bateson was a man obsessed with beginnings. The program of research described in his 1894 book, Materials for the Study of Variation, might even be regarded as an attempt to do evolutionary biology under the sway of Mivart’s dilemma, which by the 1880s had become a fixture of evolutionary discourse.* To better understand Conn’s recommendation, then, it will be useful to review the basics of Bateson’s research program, since it was a response to precisely those issues Conn identified as impediments to progress in evolutionary theory.

[* Conn was steeped in Bateson’s work, having taken on the colossal task of reviewing Materials for the Study of Variation just five years earlier.]

William Bateson’s Inductive Science of Beginnings

William Bateson seems to have had a difficult personality: “forgivably petty as a youth, [and] difficult to bear as a leading figure in the… field he named [genetics]” (Cock and Forsdyke 2008). He is frowning in nearly every picture I have seen of him, and contemporary accounts make it seem as if this is no accident. Yet it would be wrong to put his (often very grumpy) criticisms of Darwinism down to a personality disorder or a series of personal rivalries, as has sometimes been done. Like many of his contemporaries, Bateson’s low opinion of natural selection was supported by apparently sound arguments— arguments that many in the late nineteenth century regarded as lethal to the selection theory.

William Bateson (right) wearing his characteristic scowl, with Reginald Punnett of “Punnett Square” fame (left). Bateson’s dates are 1861–1926

The most pressing of these concerned the origin of novel structures, which Bateson treated as a focal point in the debate over the causes of evolution:

The preliminary question, then, of the degree of continuity with which the process of Evolution occurs, has never been decided. In the absence of such a decision there has nevertheless been a common assumption… that the process is a continuous one. The immense consequence of…this will appear from a consideration of the gratuitous difficulties which have been introduced by this assumption. Chief among these is the difficulty which has been raised in connexion with the building up of new organs in their initial and imperfect stages, the mode of transformation of organs, and, generally, the Selection and perpetuation of minute variations… We know that certain devices and mechanisms are useful to their possessors; but from our knowledge of Natural History we are led to think that their usefulness is consequent on the degree of perfection in which they exist, and if they were at all imperfect, they would not be useful. Now it is clear that in any continuous process of Evolution such stages of imperfection must occur, and the objection has been raised that Natural Selection cannot protect such imperfect mechanisms so as to lift them into perfection. (Bateson 1894, 15–16, emphasis added)

“Of these objections which have been brought against the Theory of Natural Selection,” he grumbled, “this is by far the most serious.” Darwinism was deficient because it had no answer to Mivart's dilemma.

It had begun so differently. Bateson began his career as an avid Darwinian, trained by the leading embryologist of his generation, Francis Balfour. While still a young and relatively unheralded morphologist, he undertook an expedition across the Asian Steppes with the aim of investigating the origin of the Aral Sea basin lakes. The project came to nothing; yet Bateson perceived that the lakes afforded an ideal opportunity to test the principles of adaptation in invertebrates (specifically, the effects of salinity on freshwater mollusks). He expected that the features of organisms would display a close correlation with secular conditions, testifying to the power of selection to forge a precise match between organism and environment (Cock and Forsdyke 2008, 28). Yet his investigations failed to uncover any evidence that the features of organisms graduate in sympathy with environmental conditions, at least to a degree that would establish this as a key aspect of their natural history. While the action of selection could explain some of the correlations he documented, it seemed unable to account for the majority, and further study turned up more instances in which natural selection seemed impotent to account for observed variations (e.g., Bateson and Bateson 1891).

Painting showing the shore of the Aral Sea (1848). Once the world’s third largest lake, the Aral Sea is now largely dried up in an ongoing catastrophe driven by the diversion of waters for irrigation

An intriguing possibility thus presented itself. Perhaps a large number of characters came into being discontinuously, for reasons having to do with their intrinsic “stability” as opposed to their usefulness in the struggle for life. Worries about the nonfunctionality of incipient characters could then be bypassed, since there would be no incipient characters of the sort Mivart had in mind. Selection, then, would have nothing at all to do with the origin of novel structures; its only roll would be to preserve or destroy them once they happen to appear.

Here was his ticket. Reasoning from these considerations, Bateson concluded that the sudden birth of distinct characters is a key process in evolution— the key process, even. Living forms are not separated from their ancestors by a gliding series of intermediate forms, like pages in a flip-book. Rather, they are just the latest members of a graded series whose members are separated by sharp morphological discontinuities. It is the nature of the organism that determines the transformations it can undergo, just as the nature of a molecule determines the kinds of reactions it can participate in. He elaborated the theme in an 1899 lecture:

Think of what organisms are— complex systems of chemical and molecular actions. We are asked to suppose that they have no natural or inherent order— that without Selection or environmental control they would be Chaotic. But when [we] turn to the facts of variation [it is] not necessary to suppose any such thing. (Bateson 1899 [lecture], quoted in Cock and Forsdyke 2008, 140)

Again: natural selection is not required to explain the major features of evolution, including the novel structures that adorn the branches of the tree of life. Although selection can decide between discrete varieties, preserving the fit and destroying the unfit, it has nothing to do with the origin of varieties or the definiteness of their forms (see, e.g., Bateson 1894, 64-65, 266, 417ff). These find their explanation in the nature of the organism and its tendency to throw up certain kinds of variations and not others. The conclusion seemingly follows progress in the study of evolution requires the “direct” study of morphological variation (Bateson 1894, 17). So this is what Bateson did. Materials for the Study of Variation is an approximately 600 page inductive treatise on the nature of variation and its expression throughout the organic world, intended (among other things) to overcome the “gratuitous difficulties” introduced by Darwin’s assumption that evolution is continuous. This makes it an early attempt to do evolutionary biology under the sway of Mivart’s dilemma. It was, in other words, an attempt to build a research program around the inductive (observational) study of (discontinuous) evolutionary novelty.

A characteristic image from Materials for the Study of Variation, showing meristic variation in a hydrozoan. (Meristic variation is an obsession of Bateson’s, and is characterized by a modification in number or geometrical relation of body parts.) On the left is the normal form, with four radii; on the right is an abnormal form with six radii. Both animals are pictured from above.

In Materials, Bateson’s interest in the nature of the organism is on full display. It shows up most clearly in his interest in patterns (like “the almost universal presence of Symmetry and of Repetition of Parts among living things”) and the modes of their expression over onto- and phylogenetic time (Bateson 1894, 21). In order to understand how organisms “come to be what they are,” the investigator must confront the familiar question of “the origin and meaning of patterns… the outward and visible expression of that order and completeness which inseparably belongs to the phenomenon of Life”:

If anyone will take into his hand some complex piece of living structure, a Passion-flower, a Peacock’s feather, a Cockle-shell, or the like, and will ask how it has come to be so, the part of the answer that he will find it hardest to give, is that which relates to the perfection of its pattern. And it is not only these large and tangible structures that the question arises, for the same challenge is presented in the most minute and seemingly trifling details. In the skeleton of a Diatom or of a Radiolarian, the scale of a Butterfly, the sculpture on a pollen-grain or on an egg-shell, in the wreaths and stars of nuclear division, such patterns again and again recur, and again and again the question of their significance goes unanswered. (Bateson 1894, 21–22)

Darwinism, Bateson implied, could explain (certain) variations in patterned structures. Yet it was the patterns themselves, not their variable expression, that required attention, and the former was the more difficult thing to explain. He did not deny that “minute and seemingly trifling details” may have their uses (although he was bearish on our ability to discover them). His point was rather that after the utility of a structure had been demonstrated, it remained for the investigator to account for the structure itself: and this brought us face to face with the problem of beginnings. To understand evolution, one needed to understand the origin of new characters, and the coming-to-be of new patterns. That was why Bateson expended so much effort compiling information on variation. It was as if he hoped to crush the problem of beginnings under an avalanche of observations.

Another characteristic illustration from Materials, showing a “double hand” (I) along with some bones from the elbow of the same appendage (II)

To summarize, Bateson believed that “the building up of new organs in their initial and imperfect stages” constituted the most pressing difficulty on Darwin’s theory, and he motivated it by pointing to Mivart’s dilemma. It was because Bateson faced this problem head-on that Conn devoted such a large portion of The Method of Evolution to Bateson’s research on variation and pattern formation. Yet it was not for Bateson to convert the world to the cause of discontinuous evolution. That honor fell to Hugo de Vries, whose research on the evening primrose purported to show the origin of species per saltum, bypassing Mivart- and Jenkin-style arguments (Der Vries 1900). De Vries’s “mutation theory” would go on to dominate evolutionary discussion for almost a next decade. It remained a burning topic when Vernon Kellogg sat down to write his celebrated survey of the evolutionary literature, Darwinism To-day (1907).

The Problem of Beginnings in Vernon Kellogg’s Darwinism To-day

To many in the first decade of the twentieth century, de Vries’s demonstration of spontaneous speciation in Oenothera lamarckiana constituted “the most promising new development to appear in biology since Darwinism itself” (Bowler 1983, 40). “[No] work since the publication of Darwin’s Origin of Species has produced such a profound sensation in the biological world,” F.C. Baker proclaimed in 1906. The eminent zoologist Charles Davenport agreed:

De Vries’s great work “Die Mutationstheorie” marks an epoch in biology as truly as did Darwin’s “Origin of Species”… Ever since Darwin’s time most biologists have been content to discuss and argue on the modus operandi of evolution. The data collected by Darwin have been quoted like scriptural texts to prove the truth of the most opposed doctrines. We have seen biologists divided into opposing camps in defense of various isms, but of the collection of new data, and above all, of experimentation we have had little. The great service of de Vries’ work is that, being founded on experimentation, it challenges to experimentation as the only judge of its merits. It will attain its highest usefulness only if it creates a widespread stimulus to the experimental investigation of evolution. (Davenport 1905, 369)

The evening primrose, Oenothera lamarckiana, color plate in Die Mutationstheorie (de Vries 1900)

By the year 1900, the most important desideratum for a theory of evolution was not explanatory power, but empirical rigor (Magnus 2000). This was the main selling point of the mutation theory. “One of the greatest values of de Vries’s work,” C.S. Gager proclaimed, “was in the fact that he was present when the transition took place, and gives, not a theory at all but the record of a fact observed again and again” (Gager 1906, 86). The virtue counterbalanced the theory’s uncertain scope— a shortcoming that blood many to hold it at arm’s length. De Vries, after all, had been privy to the mutable period of just one species, and this a rara avis, chromosomally speaking. In addition, he had not so much exercised control over the mutation as he had domesticated it, utilizing the quasi-experimental method of “pedigree-culture” (Jordan 1906). While de Vries’s observations were valuable— not least because they seemed to show, “experimentally,” how genuine novelties could arise in the course of evolution— more observations were needed before the mutation theory could be accepted at anything resembling face value.

These criticisms were not lost on Vernon Kellogg (1867–1937), the Stanford entomologist who, in 1907, authored two books sharply critical of the mutation theory (Kellogg 1907; Jordan and Kellogg 1907). Born in Kansas in 1867, Kellogg was among the first group of students “who never [had] to face the question of whether or not to accept evolution as a natural phenomenon” (Largent 2009, 8). Because of this, he was able to focus his attention entirely on the process of evolution, as opposed to the evidence for descent or the complexities of genealogical relationships. This emphasis was evident in Darwinism To-day, Kellogg’s celebrated exploration of the debates in evolutionary theory around the turn of the century. He began, like Conn, with an appraisal of the field:

The fair truth is that the Darwinian selection theories, considered with regard to their claimed capacity to be an independently sufficient mechanical explanation of descent, stand to-day seriously discredited in the biological world. On the other hand, it is also fair truth to say that no replacing hypothesis or theory of species-forming has been offered by the opponents of selection which has met with any general or even considerable acceptance by naturalists. Mutations seem to be too few and far between; for orthogenesis [straight-line evolution] we can discover no satisfactory mechanism; and the same is true for the Lamarckian theories of modification by the cumulation, through inheritance, of acquired or ontogenic characters. Kurz und gut, we are immensely unsettled. (Kellogg 1907, 5)

Unsettled, but not hopeless. Like Conn, Kellogg was optimistic that the situation and evolutionary theory could be improved, and would be improved in the coming decades. His task was to prepare the ground for such an improvement: not so much to plant the field, but to break the stubble land with his plow.

Although Darwinism To-day ranges over many topics, perhaps none was closer to Kellogg's heart than “determinate variation,” or variation directed along non-random, or “determinate,” lines. To hear Kellogg tell it, “The problem of the existence or non-existence of determinate variation…is one of the most important matters in connection with the whole great problem of descent, that is, of evolution” (Kellogg 1907, 34–35). “It is the basic problem of evolution,” he stated, “for it is the problem of beginnings” (35, emphasis added). Kellogg’s discussion of this problem was scattered throughout the book, beginning in a series of chapters titled “Darwinism Attacked.” Here the reader is treated to a catalog of problems that have been urged against the theories of natural and sexual selection, including “the objection... long ago strongly stated by Mivart... that numerous useful characteristics or adaptations of organisms are useful only in a highly perfected state, often involving a complex and considerable structural development of old (then much modified) or quite new parts, and hence could not have arisen by gradual modification by the selection of minute variations” (Kellogg 1907, 49). This objection, Kellogg disclosed, “has long appealed strongly to me” in connection with his own field of study: “the utility of colour and pattern among insects.” Just how did these patterns arise in the first place, especially when they are only useful in their developed state?

An American monarch butterfly (left) and its mimic, the viceroy (right)

To illustrate the worry, Kellogg mentioned “the mimicry of our common American monarch butterfly, Anosia plexippus, by the viceroy butterfly, Basilarchia archippus.” “Now, Anosia is distasteful to birds,” but the viceroy is not, and “would be a welcome bonne bouche to any bird that could distinguish it” (Kellogg 1907, 50). There is accordingly “huge usefulness [in the viceroy mimicking the monarch], and selection can well be the steadfast maintainer of the viceroy’s dissimulation.”

But of what avail for this purpose of deceit was the first tiny tinge or fleck of red-brown on the starting black and white wings of the ancestral viceroy? How can one possibly conceive of an attainment of this identity of pattern between mimicker and mimicked by selection on a basis of life-or-death-determining advantage of slight chance appearances of brown or reddish flecks or tinges in successive viceroys? Not until practically full development of the mimicry pattern existed can this pattern have worked its advantage. (Kellogg 1907, 50)

The passage could have been written by Mivart himself, and indeed very nearly was. (For Mivart's remarks on variation see the Genesis of Species, 33–38.) Like Mivart, Kellogg saw no way around the difficulty. Elaborate mimesis of this sort could not have arisen by the selection of incremental changes, since the first of these changes would have conferred no advantage upon the would-be mimic. Natural selection simply had no way to get started on the creation of a good mimic, unless it had more to work with than small fluctuating variations around an ancestral morphological condition.

As noted, Darwinism To-day explored many objections to natural (and sexual) selection, among them that selection cannot make use of “the familiar and always occurring fluctuating variations called Darwinian” (the swamping argument); that it cannot explain “coadaptive and highly complex adaptations” (the problem of coadaptation); and that it cannot explain “qualitative differences in species, and many-branched descent” (Kellogg 1907, 137). Yet Mivart’s dilemma enjoyed a special place in the text. Writing in the chapter “Darwinism Defended,” Kellogg observed that

[on] the whole… I think I speak perfectly fairly in saying that the believers and defenders of the natural selection theory to-day admit in large measure the validity of those criticisms which are directed at the incapacity of Darwinism, in its long familiar form, to account for the development of variations and modifications up to the advantageous or disadvantage stage. (Kellogg 1907, 135–136)

That is, Kellogg supposed that even staunch Darwinians would admit the incapacity of natural selection to explain the development of characters before they become advantageous. But if selection was “radically weak at its base, being incapable of explaining the beginnings of useful variations,” then an explanation of beginnings must be sought elsewhere (Kellogg 1907, 134). Kellogg considered the possibility that evolution proceeds by jumps as Bateson and de Vries argued, but his sympathies went elsewhere— toward the “Lamarcko-Eimerian” school of thought. According to this school, “ontogenetic variations produced directly in response to environment” can originate evolutionary change before selection is afforded a handle for action. Indeed, selection may never be afforded such a handle: witness the many non-adaptive differences between closely related species. This kills two birds with one stone: (1) Mivart’s dilemma, and (2) the “existence...of species differences which are of no especial utility” (Kellogg 1907, 136).

Kellogg’s own views on this subject are interesting (Kellogg 1906). Yet instead of unpacking them, I want to explore the theoretical tradition in which Kellogg located himself (albeit tentatively and partially): that of orthogenesis or “straight-line” evolution. As Kellogg observed, orthogenesis arose in response to “[one] of the principal criticism of the selection theory,” namely, “the impossibility of explaining the beginnings of advantageous modification and the beginnings of new organs, by the selection of fluctuating individual variation” (Kellogg 1907, 274). So, for example, Theodor Eimer (of the “Lamarcko-Eimerian” school) wrote that “The Darwinian principle of utility, the selection of the useful in the struggle for existence, does not explain the first origin of new characters” (Eimer 1890, 2). Likewise, E.D. Cope claimed in The Origin of the Fittest (1887) that “[g]reat obscurity has arisen from the supposition that natural selection can originate anything,” in part because “[selection] requires that the structures preserved [by its action] be especially useful to their possessors” (15, emphasis). As a final case study in the influence of Mivart’s dilemma, then, let’s look at the Lamarcko-Eimerians, focusing especially on Eimer himself.

Theodor Eimer and the Origin of New Characters

There is no easy way of summarizing of the theory of orthogenesis, because there was no theory of orthogenesis. The term was coined by Wilhelm Haacke in 1893, and co-opted by Theodor Eimer shortly thereafter; but in most subsequent discussions, it was taken to refer to a loosely related group of ideas that took evolution to be directed by factors internal to the organism, or else by laws or principles that caused evolution to move in “straight lines” (Jepsen 1949). Haacke was a minor figure who quickly became a historical footnote. But Eimer enjoyed more success, chiefly on the strength of his studies of the wall lizard Lacerta muralis and swallowtail butterflies in the genus Papilio. An early convert to Lamarckism, Eimer’s enthusiasm for “definitely directed evolution” (orthogenesis) did not crest until the final decades of his life, at which point he set himself to the utter destruction of August Weismann’s Neo-Darwinism (Bowler 1983, 153–154). However, Eimer’s very personal hatred of Weismann was not the only motivation for his orthogenetic theorizing. Eimer’s most theoretical work, called Organic Evolution as the Result of the Inheritance of Acquired Characters According to the Laws of Organic Growth, began with the following words:

It seemed to me long ago of the greatest importance to undertake an investigation of the question whether the modification (variation) of the species of animals is not governed by definite laws. It had previously been assumed that variation occurred quite irregularly, in the most diverse directions, that it was abandoned completely to chance; in fact, the origin of species according to the Darwinian explanation is left entirely to chance… If, as I acknowledge, the principles of Darwinism are true because they can be shown to follow from natural laws, then it was to be expected that obedience to laws would also be discovered in that province which Darwin had surrendered to chance. But if variation were shown to follow certain laws, the same demonstration would apply to the origin of species. (Eimer 1890, 1)

Gustav Heinrich Theodor Eimer (1843–1898)

But why think that variation follows rigid laws that impart a definite direction to the course of evolution? Simply put, Eimer found it inconceivable that natural selection could explain the overall course of evolution, since selection “can only work with existing material, and it cannot even use that until it has attained a certain perfection, until it is already useful” (Eimer 1898, 21). But evolution does move in certain directions with apparent linearity, he supposed. It must then be the case that evolution moves along pre-set channels, at least some of the time, and that these channels are responsible for the shape of novel characters. A tenuous inference, perhaps: but notice the payoff. If Eimer was right, then the greatest of all evolutionary puzzles would be solved. The origin of characters would show itself to be a function of the laws governing variation in particular groups, which channel growth in certain directions, and ultimately, direct evolution down preferred pathways of morphological change. Not only would this deal a killing blow to Weismann's Neo-Darwinism, but it would place a distinctively Eimerian stamp on all future studies of the evolutionary process. Eimer wanted this. So he shot his shot.

Eimer is not always remembered as someone interested in the problem of novel structures. Yet for him, novelty was of the essence of evolution. It was also useful. As Mark Ulett has observed, “[Eimer leveraged] the general explanatory strengths of orthogenesis and the weaknesses of natural selection” by focusing on Darwin’s greatest bugbear: “the origin of novel characteristics” (Ulett 2014, 126). By doing so, “[he] was not quibbling over small details or academic minutiae” but instead focusing “on the ‘big picture’ of evolution”— those aspects of organic descent that most demanded an explanation. “Eimer knew that evolutionary theories emphasizing natural selection could not easily account for incipient features and the evolution of underdeveloped characters, a problem that Darwin… and many others had [commented] on.” He thus proposed to study the problem of beginnings directly, and chose as material for his study butterflies in the genus Papilio.

Eimer worked by arranging specimens into intergrading series, each beginning with the “fundamental” form, Papilio podalirius. In effect, his goal was to read the laws of evolution directly from the wings of his butterflies. Eimer asserted that P. podalirius was the ancestral form (on the grounds that it manifests the basic scheme of wing markings for the group with the fewest “aberrations”), and that it had retained its morphology even as derived lineages underwent coordinated changes. The rest of the species were then interpreted as modifications of this fundamental form, beholden to shared laws of variation. Like Bateson, Eimer considered contemporary forms only— he was not interested in the project of reconstructing the forms of hypothetical ancestors. His project was rather to identify the actual stages of phyletic development through which derived forms had passed in contemporary species, ruling out the possibility of extinct intermediates. This forced him to employ some tenuous hypotheses concerning, for instance, the “standstill of development [evolution] at definite stages” (termed epistasis), and the parallel development of similar (identical) morphologies (Eimer 1898, 30). Still, Eimer was apparently convinced that his project had succeeded, and that similar projects could be undertaken to reveal the laws of variation in other groups.

A characteristic figure from Eimer (1898), showing two wing morphologies in swallowtails. Eimer argues, for example, that wing spots (like those in P. machaon, here labeled with an “a”) are generated by the abridgement and fusion of lateral stripes; but this is only an example of a general tendency. Every detail of the wing he supposes to be controlled by law. “There is no chance in the transformation of forms”: just “organic growth” in its law-bound regularity. Even mimicry, in most cases, has “nothing to do with adaptation”; mimics “must have originated suddenly and in approximate perfection in order to have given selection any hold for its operations” (Eimer 1898, 54)

It will come as no surprise to learn that this project attracted critics. Most notable for our purposes was the American embryologist Charles Minot, whose exchange with Eimer’s assistant (Maria von Linden) following Eimer’s death is revealing. Von Linden was totally convinced of the validity of Eimer’s methods. Most important was the practice of arranging contemporaneous specimens into series of intergrading forms:

This short exposition of the direction of evolution in the genus Papilio seems to me to show that it was not arbitrariness on Eimer’s part to select Papilio Podalirius as the ancestral form of his group of butterflies. He has shown, by his study of the markings of one series of forms, that those of all its members can be reduced to one and the same scheme, and that aberrations from the forms which are nearest to this original scheme of markings vary so as to form transitions to nearly allied species, which again are connected with more distant species, and the conclusions drawn from the study of these phenomena are confirmed by the results of geographical distribution. This being so, I cannot understand how Minot can doubt that Eimer’s assertions are correct. (von Linden 1897, 311)

Minot’s reply was terse, and cut to the quick of Eimer’s approach:

All of Eimer’s evidence is essentially that he asserts that of a group of living species a certain form or certain forms are ancestral types. If one denies that assertion Eimer cannot prove that it is correct, but unless he proves it his deductions remain hypotheses. The reader is asked to consider whether Countess von Linden offers proof that a certain species in any given case is the ancestral race. (Minot 1897, 313, emphasis added)

This methodological quarrel is not my main concern, however. Minot had von Linden dead to rights— nevermind that. What I find interesting is what von Linden’s remarks reveal about Eimer’s motivation for undertaking his studies. Responding to the allegation that Eimer rejected Darwinism because it did not explain the origin of variation, von Linden wrote: “On the basis of arguments which have hitherto been considered customary and convincing in biology, I believe I have shown that Eimer [is] far from rejecting Darwin’s theory…because ‘it does not explain the origin of variations’” (von Linden 1897, 312). She continued:

He knows as well as Minot that Darwin does not even attempt an explanation of the origin of new characters [here von Linden evidently conflated the origin of variations with the origin of new characters –MD]… As, however, the theory of the origin of species demands an explanation of the origin of new characters, Darwin has not, as Eimer shows, explained that which he wished to explain. Eimer, on the contrary, shows in [Die Artbildung und Verwandtschaft bei den Schmetterlingen] how new qualities develop; he explains the causes of their formation and traces the laws of their development. (von Linden 1897, 312, emphasis added)

Thus, to von Linden, Eimer’s thinking deserved the benefit of the doubt because it at least attempted an explanation of that most important of evolutionary riddles: the origin of new characters. We may be unconvinced by Eimer’s methods, but we should pay attention to its motivation and conceptual framing. Here, as in studies of discontinuous evolution, we find lurking the specter of Mivart’s Dilemma.

Eclipse of Darwinism?

I’m under no illusion that I’ve established my claim that Mivart’s dilemma played a decisive role in structuring attitudes towards natural selections in the decades following Darwin’s death. Still, I’ve taxed your patience long enough (and I’m out of vignettes), so I rest my case.

With the remainder of this essay I want to address a related matter. Repeatedly in this essay I have found myself in need of a term for the interval immediately following Darwin’s death. Usually, I have just said “the decades following Darwin’s death” or the “immediate post-Darwinian period”; but there is another expression that is sometimes used to refer to the same interval. Although I’ve shied away from it, I want to suggest in closing that it is a perfectly suitable expression, and that recent historical misgivings about it are overstated.

Since the 1980s, historians have referred to the period following Darwin’s death as the “eclipse of Darwinism” (Bowler 1983). The expression is meant to convey an ebbing in the popularity of Darwinism (natural and sexual selection), along with a growth of interest in alternative evolutionary mechanisms (like saltationism and orthogenesis). But what made this particular ebbing unusual is that, beginning in the 1930s, natural selection experienced a spectacular revival. (With sexual selection, the revival took longer, but it too eventually came.) The metaphor of “eclipse” therefore suggests itself, because eclipses are transient happenings. We don’t speak of the “eclipse of phlogiston,” because the concept of oxygen is presumably fixed in our scientific ontology. But with “Darwinism” the astronomical metaphor has more purchase— or perhaps not, if recent commenters are to be believed.

Julian Huxley (here pictured in front of grandpa Thomas Henry) coined the expression “eclipse of Darwinism” in Evolution: The Modern Synthesis. It was later popularized by Peter Bowler in his book, The Eclipse of Darwinism

Historians have expressed two main concerns about the expression, “eclipse of Darwinism.” The first concerns its descriptive adequacy. An eclipse can only be observed when the sun (or moon) is above the horizon. But Darwinism’s “sun” arguably never rose high enough to be eclipsed (at least when that sun is understood to be the theories of natural and sexual selection). So the expression “eclipse of Darwinism” is inaccurate (Meulendijks 2021).* Of course, inaccurate expressions can sometimes play a useful role in communication. But the way this expression functions is as propaganda for a Whiggish history of evolutionary thought in which all roads lead to the modern synthesis— the second criticism. Just as the expression “Dark Ages” functions to slander an entire not-so-benighted period in human history, so the “eclipse of Darwinism” tends to obscure a heady and energetic time in the history of evolutionary science (Largent 2009). As such, it has arguably outlived its usefulness as a historiographical construct.

[* Alternatively, if one takes a broad view of Darwinism, it is doubtful that Darwinism’s sun was ever eclipsed: not because it failed to rise, but because it continued to shine throughout the period. Meulendijks (2021), for example, argues that the only development resembling an eclipse was the German debate over Haeckel’s version of Darwinism. But this is hardly the same thing as a universal decline in support for Darwinism (broadly construed). So, again, the expression “eclipse of Darwinism” is inaccurate.]

But if not the eclipse, what should we call this period? Replacements have been proposed, but these seem no better than the term they aim to replace. Mark Largent, for example, has proposed that we call the traditional “eclipse” period interphase, in reference to the portion of the cell cycle in which the cell prepares for division (Largent 2009). He thinks this is preferable since the new term is less “teleological” than the old: it does not analyze “early twentieth century biology… merely in the context of what follows it.” Yet I fail to see how “interphase” is any less teleological than “eclipse.” To my eyes, it is more teleological. Eclipses are just accidents. They are astronomical coincidences in a fairly literal sense. Interphase, by contrast, happens in order to facilitate cell division. That is its telos, or purpose. But this has all the wrong connotations. Certainly we do not want to imply that the events of the 1880s to 1920s served only to prepare the way for later developments, including the much-maligned “[modern] evolutionary synthesis.” Whatever we do, then, let’s not call the period following Darwin’s death “interphase.”

I’m not a historian; I just play one on the internet. Still, it seems to me that the phrase “eclipse of Darwinism” is a perfectly suitable one in light of: (1) the more or less complete dismissal of sexual selection between 1882 and 1930, and (2) the widespread concerns about the efficacy of natural selection, especially in relation to novel structures. Yes, the word “eclipse” has some misleading connotations (the “bright sun” of Darwinism and all that). Yes, there were— and continue to be— serious questions about “who [was able to] claim the mantle of Darwin’s name as endorsement for their [ideas]” (Hale 2015, 16). But the expression “eclipse of Darwinism” strikes me as a useful way of summarizing an invaluable observation: that for a time “the Darwinian selection theories” were regarded as obviously and irreparably deficient by a large group of sophisticated scientists. Mivart’s dilemma, along with other anti-Darwinian arguments, had done their work. Once they had been gotten rid of, the “eclipse” was effectively over.

For a little after-dinner mint on Mivart’s dilemma, read Part 2.5 of this essay…

References

Baker, F. C. 1906. Application of De Vries’s Mutation Theory to the Mollusca. The American Naturalist 40:327–333.

Bateson, W. & Bateson, A. 1891. On Variations in the Floral Symmetry of certain Plants having Irregular Corollas. Journal of the Linnaean Society of London 28:386–424.

Bateson, W. 1894. Materials for the Study of Variation, Treated with Respect to Discontinuity in the Origin of Species. New York: Macmillan.

Bowler, P. J. 1983. The Eclipse of Darwinism. Baltimore: Johns Hopkins University Press.

Bowler, P. J. 1996. Life’s Splendid Drama. Chicago: University of Chicago Press.

Cock, A. G. & Forsdyke, D. R. 2008. Treasure your Exceptions. The Science and Life of William Bateson. New York: Springer.

Conn, H. W. 1887. Evolution of To-day. New York: G.P. Putnam’s Sons The Knickerbocker Press.

Conn, H. W. 1900. The Method of Evolution. New York: G.P. Putnam’s Sons The Knickerbocker Press.

Cope, E. D. 1887. The Origin of the Fittest: Essays on Evolution. New York: D. Appleton & Co.

Davenport, C. B. 1905. Species and Varieties, Their Origin by Mutation, by Hugo de Vries [Review]. Science 22:369–372.

De Vries, H. 1900. Die Mutationstheorie (v1). Leipzig: Veit & Comp.

Dohrn, A. 1875. Der Ursprung der Wirbelthiere und das Princip des Functionswechsels. Genealogishce Skizzen. (English translation Ghiselin, M. 1995. The origin of vertebrates and the principle of succession of functions. History of the Philosophy of Life Science 16:5–98.

Eimer, T. 1890. Organic Evolution as the Result of the Inheritance of Acquired Characters According to the Laws of Organic Growth. London: Macmillan & Co.

Eimer, T. 1898. On Orthogenesis, and the Impotence of Natural Selection in Species Formation. Chicago: The Open Court Publishing Company.

Gager, C. S. 1906. De Vries and His Critics. Science 24: 81–89.

Hale, P. 2015. Rejecting the myth of the non-Darwinian revolution. Victorian Review 41:13–18.

Hoquet, T. 2024. Beyond blending inheritance and the Jenkin myth. Journal of the History of Biology 57:17–49.

Huxley, J.S. 1942. Evolution: the modern synthesis. London: George Allen and Unwin, Ltd.

Jepsen, G. 1949. Selection, “Orthogenesis,” and the Fossil Record. Proceedings of the American Philosophical Society 93:479–500.

Jordan, D. S. 1906. Discontinuous Variation and Pedigree Culture. Science 24:399–400.

Jordan, D. S. & Kellogg, V. L. 1907. Evolution and Animal Life. New York: D. Appleton & Company.

Kellogg. V. L. 1906. Is There Determinate Variation? Science 24:621–628.

Kellogg, V. L. 1907. Darwinism To-day. New York: Henry Holt & Company.

Largent, M. 2009. The So-called Eclipse of Darwinism. In J. Cain & M. Ruse (Eds.) Descended from Darwin, 3–21. Philadelphia: American Philosophical Society.

Magnus, D. 2000. Down the Primrose Path: Competing epistemologies in early twentieth-century biology. In R. Creath & J. Maienschein (eds.) Biology and Epistemology, 91–121.

Meulendijks, M. 2021. Eclipsing the eclipse? A neo-Darwinian historiography revisited. Journal of the History of Biology 54:403–443.

Minot, C. S. 1897. Reply to M. von Linden. Science 6:313.

Mivart, S. J. 1871. On the Genesis of Species. London: Macmillan & Co.

Noll, K. M., Marco, M. L. and Cochrane, B. J. 2014. Microbiology milestone: Herbert W. Conn, lifetime achievements. Microbiome 9:401–405.

Ulett, M. A. 2014. Making the case for orthogenesis: The popularization of definitely directed evolution (1890–1926). Studies in History and Philosophy of Biological and Biomedical Science 45:124–132.

von Linden, M. 1897. Eimer’s evolution of butterflies. Science 6:308–313.