5-9B. At the apex of this phylogeny, De Vries shows the 7 mutant forms derived from Oenothera Lamarckiana. The rest of the diagram illustrates De Vries' general view of evolution; with most lineages stable nearly all the time, but entering short mutational episodes when several new species may arise virtually at the same time. From Volume 2 of De Vries' Mutation Theory.
[Page 430]
The theory of mutations is a starting point for direct investigation while the general belief in slow changes has held back science from such investigations during half a century” (1905, p. 30).
As an experimental reductionist, committed to finding the mechanism for large-scale evolution of phenotypes in the smallest cellular parts, de Vries sought the causal basis for his mutation theory in the character of variation and its putative causes. As a foundation for all his theorizing, de Vries proposed a strict separation between two distinct types of variation: fluctuating and mutational. (This division, of course, establishes the same false dichotomy that prompted the famous “biometrician” vs. “Mendelian” debate — a struggle that de Vries' context did much to promote). De Vries stated that, in the early 1870's, he had read Quetelet's work on normal curves and Galtonian regression to the mean — and had determined thereby that the omnipresent, small scale or, in his favored term, “fluctuating” variation could not be parlayed into directional evolutionary change, as Darwin's theory required (Fig. 5-10). Evolution must therefore require a conceptually and causally distinct mode of sudden, larger-scale, true breeding and non-regressing variation — a necessary source eventually found in the “mutations” that yielded distinct new phenotypes in Oenothera.
De Vries acknowledged that selection of fluctuating variation could produce new agricultural races and stocks of domesticated animals. But this Darwinian alteration can only yield a minor change from the mean of a parental stock: “It is responsible only for the smallest lateral branches of the pedigree, but has nothing in common with the evolution of the main stems. It is of very subordinate importance” (1905, p. 801). These new races, if not constantly superintended, will rapidly revert towards parental characters by regression to the mean (1909a, volume 1, pp. 88-89). De Vries, who understood the logic of Darwinism so keenly (see pp. 446–451), knew that the most promising
5-10. De Vries' illustration of continuous “fluctuating” variation, which he regarded as ineffective as a source for evolutionary change. From volume 1 of De Vries' Mutation Theory.
[Page 431]
Darwinian escape from this paradox lay in a claim that fluctuating variation would be reconstituted symmetrically about a newly established mean — and that continual, directional selection to great cumulative effect could therefore be achieved. De Vries, following the classic argument about a “rigid sphere” of variation, as famously formulated by Fleeming Jenkin (1867), simply denied that such reconstitution could occur. Again, using Wallace as Darwin's surrogate, de Vries stated (1909a, volume 1, p. 42): “I admit that with this assumption [of limitless and reconstitutable variation] it would be very easy and simple to account for the phenomenon of adaptation ... If [Wallace's] assumption is once granted everything follows. But it is, as a matter of fact, fallacious.”
De Vries summarized his views on the inefficacy of fluctuating variation (which, he suspected, did not have a particulate Mendelian basis and arose by influences that we would now call ecophenotypic). How could substantial and permanent evolutionary change originate from a style of variation that (1) always regressed toward the mean; (2) arose in strictly limited extent, with preponderance near the mean, and only rarely at a useful phenotypic difference (the normal curve); (3) enjoined a strictly linear set of effects, only producing more or less of a feature, while “creative” evolution required the development of true novelty: “Individual variability, when tested by sowing, reverts to its original mean, the forms of its variants are connected together, are coherent and not discontinuous. It is centripetal in as much as the variations are grouped most densely around a mean. Finally — and this is very important — it is linear; because the deviations occur in only two directions — less or more.”
By default therefore, but not at all as a negative argument, evolutionary novelty must arise by a phenomenologically and causally distinct style of variation that de Vries called “mutational” — i.e., sudden, fortuitous (and therefore nonadaptive), true breeding and nonreverting saltations. De Vries called these saltational variants “species,” but we must understand (as he did) that such units cannot be equated with traditional Linnaean taxa of the same name. With his mutation theory, de Vries entered (and largely shaped, though he did not originate) a major debate in systematics.
Obviously, a de Vriesian saltation does not, in se, make a new species in our usual sense of the term — for the single mutant plant is only an individual with a discontinuous phenotype, however true breeding in self-fertilization. In what sense, then, could de Vries insist that he had discovered the mechanism for the origin of new species?
In large part, de Vries based his claim upon an attempted redefinition. He argued that the traditional Linnaean species encompasses an imprecise, compound aggregation including varying numbers of phenotypically distinct, true-breeding entities (and a fair amount of continuous variability as well, based on the fluctuating style). The true-breeding subtypes represent nature's genuine units and should be so designated. They arise by discontinuous saltation, without intermediates, and should be called “elementary species.” As a [Page 432] practical point, de Vries did not propose that all traditional taxonomy be restructured. He would allow the Linnaean names to persist as “species,” with his smaller, “real” units termed “elementary species.” (De Vries did not follow his own recommendation consistently, for he gave new species names, in traditional Linnaean form, to his mutational variants. For example, the large tetraploid became Oenothera gigas derived from O. Lamarckiana). In this argument, de Vries supported a movement, then current in systematics, to designate the traditional Linnaean units as linneons and the true-breeding subtypes as jordanons (to honor the botanist Alexis Jordan) — recognizing both as species of different sorts and by different criteria (with the jordanon as “more” biological and the linneon as tolerated by practical necessity). De Vries wrote:
We may conclude that systematic species, as they are accepted nowadays, are as a rule compound groups. Sometimes they consist of two or three, or a few, elementary types, but in other cases they comprise 20, or 50, or even hundreds of constant and well-differentiated forms (1905, p. 38).
The systematic species are the practical units of the systematists and florists, and all friends of wild nature should do their utmost to preserve them as Linnaeus has proposed them. These units, however, do not really exist entities; they have as little claim to be regarded as such as the genera and families have. The real units are the elementary species ... Pedigree culture is the method required and any form which remains constant and distinct from its allies in the garden is to be considered as an elementary species (1905, p. 12).
De Vries' historical argument for changing emphasis from the linneon (ordinary species) to the jordanon (de Vriesian elementary species) provides an interesting insight into his worldview and rhetorical style. Before Linnaeus, he claims, genera stood as the “natural” units of common discourse: “The old vulgar names of plants, such as roses and clover, poplars and oaks, nearly all refer to genera” (1905, p. 33). Linnaeus, also searching for the natural unit, failed to extend his argument far enough. He began with genera and then moved “down” to species. He knew that he might proceed to still smaller units, but chose to go no further:
Afterwards Linnaeus changed his opinion on this important point, and adopted species as the units of the system. He declared them to be the created forms, and by this decree at once reduced the genera to the rank of artificial groups. Linnaeus was well aware that this conception was wholly arbitrary, and that even the species are not real
indivisible entities. But he simply forbade the study of lesser subdivisions. At this time, he was quite justified in doing so, because the first task of the systematic botanists was the clearing up of the chaos of forms and the bringing of them in connection with their real allies (1905, p. 34). [Page 433]
Just as the establishment of Linnaean species had made genera artificial, so too does the recognition of de Vriesian elementary species relegate the conventional Linnaean species to a congeries with no natural status. De Vries argues that this theoretical progress from larger to successively smaller units of natural “reality” illustrates the general advance of science as a reductionistic enterprise. A “stepping down” from the Linnaean species to the de Vriesian elementary species can claim both the sanction of history and the virtue of utility: “What is to guide us in the choice of the material? The answer may only be expected from a consideration of elementary species. For it is obvious that they only can be observed to originate, and that the systematic species, because they are only artificial groups of lower unities, can never become the subject of successful experimental inquiry” (1905, p. 517).
This redefinition of species as discontinuous saltations inevitably raised the issue of whether de Vries' new units (“elementary species,” or “jordanons” of other terminologies) always originated from single monophyletic sources, or represented discrete phenotypes that could arise more than once — thus divorcing this supposedly “most real” taxon from the usual genealogical criterion of monophyly for a basic unit in a phylogenetic system. De Vries, following both the logic of his argument and his observation that elementary species of Oenothera arose again and again, accepted the implication (so strange to modern “population thinking” and genealogically based taxonomy) that the same species could, and usually did, arise many times. In fact, such a propensity for multiple origins established a major criterion for potential success. De Vries noted as central to his concept (1909a, p. 208) “the assumption that the new form or species does not arise merely once from the parent species but ... a great many times and with some degree of regularity.”
De Vries devised an interesting set of subtypes for his saltations — thus revealing another aspect of the philosophical complexity of his ideas (not always expressed with consistency, but often replete with interesting psychological and sociological influences). If “stepping down” from the linneon to the jordanon revealed a reductionist bias usually interpreted as “modernist” in his time, then de Vries's classification of mutations reveals an allegiance to notions of progress and regress that might be deemed archaic in its implied fascination for the scala naturae.
De Vries recognized some of his mutations as starkly different (in a qualitative sense) from the parental form. But others could be linked in a series, with the parent as prototype, either by loss of ancestral characters or by simple quantitative alteration. All categories included equally “good” species in the causal or genetic sense — that is, equally discontinuous entities, formed suddenly without intermediates, and true breeding under self-fertilization. But only the first category established genuine novelty in evolution; thus, only these truly different species contributed to the progress of life's history. All other categories comprised variations on parental forms (usually based on loss or diminution of characters), and could only constitute a series around the parental prototype. Therefore, in the oldest taxonomic ploy of evolutionary [Page 434] thought (dating to Lamarck's distinction of progressive increments from lateral branches — see Chapter 3), de Vries subdivided, by their presumed phylogenetic effect, these taxa of similar genetic status. He labelled mutations yielding phyletic novelties as “elementary species,” while phenotypic departures still linked to parental morphologies became “varieties.” De Vries then made a further subdivision among varieties, distinguishing taxa formed by loss of a character (“retrograde varieties”) from those that may seem more advanced than the parent but really display nothing new (atavistic reappearance of characters present in closely related species, for example, or simple enhancement of a character already present). De Vries wrote (1905, pp. 246-247):
There is a real difference between elementary species and varieties. The first are of equal rank, and together constitute the collective or systematic species. The latter are usually derived from real and still existing types. Elementary species are in a sense independent of each other; while varieties are of a derivative nature . . . We have assumed that the first came into existence by the production of something new, by the acquirement of a character hitherto unnoticed in the line of their ancestors. On the contrary, varieties, in most cases, evidently owe their origin to the loss of an already existing character, or in other less frequent cases, to the reassumption of a quality formerly lost. Some may originate in a negative way, others in a positive manner, but in both cases nothing really new is acquired.
In his most forthright statement about the differing phyletic roles of progressive and regressive mutations, de Vries then stated (1905, p. 15):
Many instances could be given to prove that progression and retrogression are the two main principles of evolution at large. Hence the conclusion that our analysis must dissect the complicated phenomena of evolution so far as to show the separate functions of these two contrasting principles. Hundreds of steps were needed to evolve the family of the orchids, but the experimenter must take the single steps for the object of his inquiry. He finds that some are progressive and others retrogressive, and so his investigation falls under two heads, the origin of progressive characters, and the subsequent loss of the same. Progressive steps are the marks of elementary species, while retrograde varieties are distinguished by apparent losses.
The logic of de Vries' system may be sound, but he faced — as he well understood — a major empirical dilemma. He had found consistent mutations in only one lineage, the genus Oenothera, and with high frequency only in the species O. Lamarckiana. (He noted an isolated example or two in other lineages, particularly in a plant with the intriguing name of “peloric toad-flax,” but found no consistently mutable form besides Oenothera. He also tested other species of Oenothera, but found most immutable or, in the case of O. biennis for example, much less subject to alteration than O. Lamarckiana. In [Page 435] an interesting later paper (de Vries, 1915), he tried to calculate the “coefficient of mutation in Oenothera biennis” vs. O. Lamarckiana, concluding that the latter species showed a 6 to 10 fold increase in mutability. De Vries attributed this augmentation to a transition of one or more pangenes from stable to labile positions — a pure speculation, but again consistent with his system and logic.)
Why only one? Did such rarity mean that de Vries had only discovered an oddball, with no general message for evolution? De Vries recognized that such an inference would destroy his system, and he therefore argued that all (or at least many) species maintain potential for entering a “mutable period,” but that very few actually exist in such a state at any moment. (We know, after all, that most species are stable in both current and paleontological perspective. If all lineages were as mutable as O. Lamarckiana, we would never be able to designate Linnaean taxa, for nature would then present a constantly changing and unbreakable continuum, rather than a set of discrete and recognizable populations.) De Vries considered himself fortunate that he had located even one species in such a state — for if “mutable periods” constitute an almost incalculably tiny fraction of a species' lifespan, the probability of finding any given species in such a state at any moment becomes effectively zero. In trying to turn the tables on his adversaries, de Vries argued that the discovery of even one case presupposes a generality for extremely rare “mutable periods” — for if such mutability could be dismissed as simply freakish and unique, he could not have expected to encounter even a single example!
The view that it might be an isolated case, lying outside of the usual procedure of nature, can hardly be sustained. On such a supposition it would be far t
oo rare to be disclosed by the investigation of a small number of plants from a limited area. Its appearance within the limited field of inquiry of a single man would have been almost a miracle... The mutable condition ... must be a universal phenomenon, although affecting a small proportion of the inhabitants of any region at one time: perhaps not more than one in a hundred species, or perhaps not more than one in a thousand, or even fewer may be expected to exhibit it (1905, p. 687).
But why should a species enter a rare mutable period, and why should most species be stable nearly all the time? What triggers a transition into this exceedingly uncommon state of evolutionary promise? On this crucial point of his entire system, de Vries fell almost eerily silent, for he could offer nothing precise. He supposed (1909a, volume 1, pp. 206-207) that some external trigger of environmental change — isolation by colonization of new areas, for example — must initiate phyletic lability, but amounts and directions of mutation must be attributed to internal states of pangenes (1905, p. 691). He offered a few general words about pangenes becoming mutable, or moving to a position of high changeability, or arising de novo with such a propensity. But he still couldn't cite anything physical, or propose anything testable. De Vries' distress and unease about this crucial subject even inspired a rare burst of [Page 436] almost religious romanticism, an odd rhetorical strategy (and smokescreen) for such a severe rationalist: “The view of permanency represents life as being surrounded with unavoidable death, the principle of periodicity follows in the same way the idea of resurrection, granting the possibility of future progression for all living beings. At the same time it yields a more hopeful prospect for experimental inquiry” (1905, p. 693).
The Structure of Evolutionary Theory Page 69
