In the last Newsletter I discussed the nemertean theory for the origin of the vertebrates as proposed in the late 19th century by the Dutch embryologist Ambrosius Hubrecht (1853–1915). Hubrecht proposed his theory in conscious contrast to the two leading hypotheses that derived the vertebrates either from an ascidian tadpole larva or from an annelid worm. By homologising the nemertean proboscis sheath (also known as the rhynchocoel) with the vertebrate notochord, and the nemertean proboscis with part of the pituitary gland, Hubrecht sought to determine the evolutionary origins of some of the most conspicuous chordate features, and at the same time reveal the evolutionary link between the invertebrate and vertebrate domains of the animal kingdom.
However, Hubrecht’s theory never enjoyed the same popularity as the competing annelid and ascidian tadpole theories, even though Hubrecht’s theory arguably necessitated less special pleading than the annelid theory, and represented a more penetrating analysis of the evolutionary origin of vertebrate characteristics than the mere comparison with other chordates would allow. But despite the absence of any significant following, the nemertean theory maintained a tenuous thread of continuity across three centuries through the works of three remarkable individuals who could not have been more different in professional orientation. This essay chronicles their advocacy of the nemertean theory, and shows how a once fruitful hypothesis about the evolution of animal body plans ultimately became transformed into asterile monstrosity that spent its last years of life within the confines of published or unpublished abstracts in conference proceedings, until it finally expired in the new millennium with the death of its last and most vociferous proponent.
As the 19th century drew to a close, efforts to reconstruct the nature of
distant ancestors were increasingly regarded as sterile exercises in mental
gymnastics, and concurrently the nemertean theory of vertebrate origins sank into quiet oblivion. However, after several decades of dormancy it was revived in no uncertain terms by an unlikely intellectual successor of Hubrecht. At the end of a productive career John Muirhead Macfarlane (1855–1943) published his book The causes and course of organic evolution. A study in bioenergics [sic] (1918). Bestowing praise in no uncertain terms, Macfarlane credits Hubrecht’s nemertean theory as “one of the most far reaching and brilliant steps ever taken in zoological science.” Not surprisingly, Macfarlane was a botanist.Born and educated in Scotland, Macfarlane occupied several different academic positions at the University of Edinburgh before he emigrated to the United States to assume a professorial chair at the University of Pennsylvania in 1893, which he held until retirement in 1920. Although he was primarily a botanist, Macfarlane’s interests ranged widely, and in addition to the book mentioned above his other works included The evolution and distribution of flowering plants (Apocynaceae. Asclepiadaceae), The evolution and distribution of fishes, Fishes the source of petroleum, and The quantity and sources of our petroleum supplies. Nevertheless, despite such an outpour of work, Macfarlane is certainly no household name in biology. His main source of fame is probably connected with his leading role in organizing and elaborating the botanical garden of the University of Pennsylvania. In the obituary for Macfarlane published in Science a colleague justly admired the “universality of knowledge” of “this beloved Scot” (Steckbeck, 1943) and it was therefore only fitting for Macfarlane to crown his career with The causes and course of organic evolution (1918), which was a wide ranging panoramic overview of the patterns and causes of animal and plant evolution. To celebrate his obvious admiration for Hubrecht’s ideas, Macfarlane dedicated a significant portion of this book to discussing and elaborating the nemertean theory for vertebrate origins. His main goal was to fill in some holes in Hubrecht’s scheme that Macfarlane thought would have led to a “rapid and wide acceptance” of Hubrecht’s ideas, would Hubrecht have thought of them himself. Thus, Macfarlane enthusiastically set out to expand Hubrecht’s list of possible homologies between vertebrate and nemertean structures, and he confidently pointed out the nemertean precursor structures of vertebrate teeth, eyes, ears, and even the uvula (proboscis stylet, eyes, cerebral organs, dorsal wall of proboscis, respectively).
However, far from smoothly extending Hubrecht’s ideas, Macfarlane adopted a radically different strategy of evolutionary inference than Hubrecht had. Hubrecht had built his case according to the principles of evolutionary morphology, i.e. on the basis of a detailed study of pure comparative morphology, without much explicit attention to contextual information such as ecological factors, habits, possible selection pressures, the adaptive value of new features, etc. In sharp contrast, Macfarlane was very much concerned with the ecological and environmental context of evolutionary changes in body plans. However, rather than aiding the reception of his ideas, Macfarlane’s consideration of contextual information may have contributed rather more to the rapid obsolescence of his book. The reason for this is that Macfarlane outlined his ideas within the conceptual framework of neo-Lamarckism at the eve of the terminal last phase of this evolutionary outlook on life. In the 1920s, the famous exposure of apparent fraud in the work of neo-Lamarckian zoologist Paul Kammerer on midwife toads extinguished the last hopes for systematically sought experimental support for the inheritance of acquired characters, and it created a scientific scandal that ended with Kammerer shooting himself in the head on an Austrian mountain path. Although these events signalled the end of neo-Lamarckism in experimental zoology, certain palaeontologists and field naturalists relinquished their faith rather more slowly (Bowler, 1983). Macfarlane’s 1918 book represents a prominent example.
Macfarlane’s outlook on the evolutionary panorama was significantly influenced by ideas previously elaborated by the premier member of the first important American evolutionary school: the neo-Lamarckian Edward Drinker Cope (1840–1899). Specifically, reading Macfarlane’s The causes and course of organic evolution, one cannot escape the conclusion that it owed an obvious debt to Cope’s The primary factors of organic evolution (1896), in particular Cope’s ideas about the importance of consciously chosen reactions of the animal to its environment, and his concept of “kinetogenesis” in determining the direction of evolutionary change. According to the first idea, animals adapt by consciously or protoconsciously reacting to changes in the environment, leading to the evolution of novel features. According to Cope’s concept of kinetogenesis, motion could exert direct and inherited effects upon an organism’s structure, a proposal explicitly accepted by Macfarlane (1918: 657). These concepts of the evolutionary importance of environmental change and the organism’s direct and inherited response are reflected in Macfarlane’s list of five “causes” of evolution, especially the second and the third: 1) heredity; 2) environment; 3) proenviron; 4) selection; 5) reproduction. Macfarlane proposed the second and third causes in conscious analogy to Newton’s law of action and reaction; each action imposed by the environment (cause 2) upon the organism leads to an immediate and inherited reaction, or “proenvironal response” (cause 3) by the organism, thereby determining the direction of evolutionary change.
Given the great importance of the environment in stimulating organisms to
change, Macfarlane thought the marine realm not to be challenging enough to
drive significant evolutionary progress. Instead Macfarlane considered the “struggle for existence” to be much more severe in the more challenging terrestrial and fresh water environments. “Accordingly proenvironal response and selective survival act more sharply to evolve new types” in terrestrial and fresh water environments (Macfarlane, 1918: 408). As a result “the writer is forced to the conclusion that the main and dominant lines of animal evolution have all originated in fresh water or on land, and that only side lines have assumed a marine life.” Macfarlane extended his heterodox ideas to the evolution of plants as well, and the main trunks of both plant and animal phylogenies in his book run through fresh
water from root to crown.
Additionally, the concept of kinetogenesis leads one to envision nature as an ambitious sculptress who incessantly chisels away at the form of animals. In line with such brute mechanical forces operating upon the shape of organisms, Macfarlane therefore expected armoured animals to be relatively immune to rapid evolutionary change. Such could be expected, for example, for arthropods with hard exoskeletons (Macfarlane, 1918: 538).
Naturally, the soft-bodied and pliable nemerteans could scarcely be expected to put up a fight against the relentless action of the elements, and after their fresh water origin, they were quickly moulded into a multiplicity of forms, ranging from the lowly cyclostomes to our exalted selves. Apparently, Macfarlane’s ideas about the causes and course of animal evolution were not generally considered the stuff of textbooks, and a long period of silence followed his championing of the nemertean theory of vertebrate origins. However, the theory’s mummified remains were once again disinterred in 1960, this time simultaneously and independently by an English cytologist specialized in tissue cultures, and an American comparative psychologist interested in the behaviour of Paramecium. However, these authors didn’t built on Macfarlane’s ideas, but rather were inspired more directly by the
initial papers of Hubrecht.
In 1960 Edward Neville Willmer (1902–2001) was reader of histology at the University of Cambridge, and in his tissue culture laboratory he studied
tissue growth and cell division with several collaborators, including the Nobel prize winner Peter Medawar. After Willmer had published previous books on tissue culture and the structure of the retina in relation to colour vision, he published the first edition of Cytology and evolution. In this textbook of comparative histology nemerteans only appear on page 334, to introduce a few pages dedicated to “the nemertine as prototype,” that is, as a prototype vertebrate. However, this situation changed dramatically with the publication of the second edition of Willmer’s book in 1970. The nemerteans as a vertebrate prototype had been promoted from a cameo role to a leading character, and the nemertean theory of vertebrate origins functioned as the main organizing theme in the book. In the book and a 1974 paper, Willmer attempts to identify precursor cell types and tissues in nemerteans of as many vertebrate cell types, tissues and organs as he can, including chloride-secreting cells, the pineal organ, the urogenital system, and the liver. In addition, Willmer placed his study of comparative histology in the context of a functional evolutionary scenario that specifies the changes in habit and habitat accompanying the evolutionary transition of nemertean to vertebrate, in particular the evolution of filter-feeding and swimming that were concomitant with a proposed escape of the carnivorous nemerteans from the benthos into the pelagic realm of the ocean. Willmer also presented a consideration of the selection pressures potentially involved in this change in body plans. As an example of his approach, Willmer suggested that it would be beneficial for nemerteans to leave the benthos because overcrowding “with a probable plethora of floating eggs, larvae, and unicellular organisms, would make it advantageous to escape into the supernatant water-phase (just as the insects and birds have escaped into the air)” (Willmer, 1974: 328).
Quite independent of whether one accepts Willmer’s carefully crafted functional scenarios or not, his work represents the most comprehensive and detailed attempt to date to defend an evolutionary link between vertebrate characters and precursor structures present in nemerteans in terms of both the structure and function of cells. Willmer’s ideas even stimulated some research into nemertean ultrastructure (e.g. Ling, 1969) to test the nemertean theory. It should be noted, however, that Willmer did not propose to derive extant vertebrates from extant nemerteans, but merely from some as yet unidentified extinct nemertean-like worm. As a result, when Yunnanozoon was described, Willmer was quite struck with its similarities to nemerteans, as he confided to fellow Cambridge-based nemertean worker Janet Moore. In fact, Willmer was not the only one with that idea. The palaeontologist Jerzy Dzik recently pointed to the similarities of nemerteans and fossils such as Yunnanozoon as perhaps providing tentative support for the “rather unorthodox” hypothesis that links nemerteans to chordates (Dzik, 2000:
139).
Willmer did not publish more work on the nemertean theory after the early 1970s. When he submitted a manuscript on this topic to be included in the proceedings of the meeting on “Recent advances in nemertean biology” to be published in 1988 in Hydrobiologia, nobody could be found to review it. The manuscript was never published. And when Janet Moore confronted Willmer with genetic data that might refute his ideas, Willmer simply responded that genes were recipes for the phenotype, and he wasn’t interested in recipes.
The last and in many ways the most extreme chapter in this story also opens in 1960. Two years after earning his PhD in comparative and experimental psychology for a thesis on “Behavioral effects of feeding, fission, and ultraviolet microbeam irradiation in Paramecium aurelia,” Donald D. Jensen published a short paper in Nature titled “Hoplonemertines, myxinoids and deuterostome origins” (Jensen, 1960). Inspired by Hubrecht’s ideas, and independently of Macfarlane and Willmer, Jensen elaborated a list of possible homologies shared between nemerteans and vertebrates. He proposed that the hoplonemerteans (a group of nemerteans possessing a proboscis armed with teeth) gave rise to early hagfishes. Hagfishes in turn formed the ancestral stock for all vertebrates, as well as all other deuterostomes, including tunicates, cephalochordates, hemichordates, and echinoderms, which arose by varying degrees of degeneration. One of the strikingly controversial aspects of Jensen’s theory was his derivation of all living deuterostomes from within the extant nemerteans. In his last contribution to the published literature, Jensen took this idea to its extreme. He in effect proposed that all bilaterians have evolved from within the nemerteans, with nalaeonemerteans representing the elusive Urbilateria, heteronemerteans giving rise to protostomes, and hoplonemerteans evolving into the deuterostomes (Jensen, 1999). In summary, during a period of 40 years Jensen defended and elaborated his nemertean theory in publications that were without exception invited book chapters, conference proceedings, or abstracts (e.g. Jensen, 1963, 1983, 1988, 1990, 1999).
In the end the nemerteans became Jensen’s Nemesis. What had initially
started out as a fruitful theory of animal body plan evolution in the late
19th century had become transformed into a personal dogma that was zealously
advocated, and aggressively shielded from criticisms. Although Jensen was
a good field naturalist and a very friendly colleague (Dr Janet Moore, pers.
comm.), he was regarded as somewhat of an oddball in the nemertean community
because of his rather heterodox ideas. As a result, and despite repeated airing of his ideas at seminars and meetings, his ideas mostly fell on deaf ears, and his theory was never critically discussed (Sundberg et al., 1998 represents the only exception). The inability of either morphological or molecular evidence to support a close relationship between nemerteans and vertebrates effectively doomed Jensen’s ideas.fÌff
But Jensen never relented. Jensen ended his unpublished abstract for the 5th International Conference on Nemertean Biology in 2000 by writing “recent
attempts by this writer to apply methods of biochemical systematics to available biochemical data will be discussed.” Unfortunately that was the last on nemerteans by Jensen. While teaching a course on scientific approaches to parapsychology, Jensen was hospitalised on Thanksgiving Day of 2003. He died three weeks later.
What can we learn from this nemertean affair? Should the opinions of Hubrecht,
Macfarlane, Willmer and Jensen simply be dismissed with a chuckle as the misguided efforts of some feeble minds? Or can we learn something more instructive from the shared commitments over a period spanning three centuries of a Dutch embryologist, a Scottish botanist, an English cytologist, and an American psychologist? I think we can, and we should. Hubrecht, Macfarlane, Willmer and Jensen all placed ancestors centre stage in their evolutionary epistemology, and this approach represents no oddity in evolutionary inference. As recently summarized by Mayr and Bock (2003: 175) “The study of phylogeny was traditionally considered to be, so to speak, a backward looking endeavour, the search for and study of common ancestors. The starting point in such an analysis is a particular taxon and the student of phylogeny attempts to infer the properties of its ancestors.” As Dayrat (2003) recently pointed out in a perceptive paper, this central role of ancestors in phylogeny reconstruction is strikingly illustrated in the work of the first great phylogenizer, Ernst Haeckel. The trunks of Haeckel’s trees represent a linear evolutionary succession of morphological stages exhibited by the species at the top of the trunk. This strategy necessitates one to make direct pronouncements about the nature of distant ancestors in order to explain the observed morphology of a chosen taxon. This backward looking approach to phylogeny reconstruction, this immediate infusion of the arrow of time into their evolutionary speculations, is exactly what unites the efforts of Hubrecht, Macfarlane, Willmer, and Jensen. This strategy of recovering the past by looking back really became only unpopular with the advent of cladistics. Ancestors have no central role in cladistics. They are merely the by-products of studies of comparative morphology, and they only vaguely flicker in ghostly outlines at the internal nodes of cladograms. The construction of a cladistic data matrix typically involves no assumptions at all about the arrow of time, just an assessment of organismic variation.
Only when the cladogram is rooted do the hypothetical ancestors appear.
With the exception of the later papers by Jensen, the nemertean theory of vertebrate origins was defended before the cladistic revolution had gathered sufficient force. And Jensen himself had serious misgivings about cladistics. Both Willmer and Jensen can be criticized for ignoring conflicting evidence, or at least not being overly receptive to it. But their direct focus on ancestors was current practice for many systematic biologists for a very long time. I think that for most of us it just feels natural to include the arrow of time right from the beginning of our phylogenetic theorizing. That explains why we sometimes cannot help ourselves. Even the most devoted hardcore cladists sometimes let their intuitions take over, with the result that cells in a data matrix are filled with nothing more concrete than unsupported expectations and suspicions based on the acceptance of an a priori hypothesis about the pattern or process of evolution. The scoring of the presence of an orthogonal, or ladder-like nervous system for taxa that lack any trace of such a nervous system in morphological cladistic data matrices published over more than a
decade is a revealing example (Jenner, 2004).
There are enough other examples that reveal our tenacious tendencies to infuse
our thinking about evolution with time’s arrow to fill a sizeable book. Just consider the extremely common fallacy of equating character states in species-poor taxa as ancestral with respect to those found in its more species-rich sister group (Crisp and Cook, 2005). The prevalence of this nasty habit of thought recently inspired an editorial in Systematic Entomology (Krell and Cranston, 2004). Evidently, we haven’t cleaned up our acts as well
as we should, yet…
References
BOWLER, P.J. 1983. The eclipse of Darwinism. Johns Hopkins University Press.
COPE, E.D. 1896. The primary factors of organic evolution. Open Court Publishing.
CRISP, M.D. and COOK, L.G. 2005. Do early branching lineages signify ancestral traits? Trends in Ecology and Evolution 20: 122–128.
DAYRAT, B. 2003. The roots of phylogeny: how did Haeckel build his trees? Systematic Biology 52: 515–527.
DZIK, J. 2000. The origin of the mineral skeleton in chordates. Evolutionary Biology 31: 105–147.
JENNER, R.A. 2004. Towards a phylogeny of the Metazoa: evaluating alternative phylogenetic positions of Platyhelminthes, Nemertea, and Gnathostomulida, with a critical reappraisal of cladistic characters. Contributions to Zoology
73: 3–163.
–––– 1963. Hoplonemertines, myxinoids, and vertebrate origins. In E.C. Dougherty et al. (eds), The lower Metazoa: comparative biology and phylogeny, pp. 113–126. The University of California Press.
–––– 1983. Hoplonemertean ancestry of vertebrates – an old head and new brain for craniates. American Zoologist 23: 925.
–––– 1988. Hubrecht, Macfarlane, Jensen and Willmer: on the nature and testability of four versions of the nemertean theory of vertebrate origins. Hydrobiologia 156: 99–104.
–––– 1990. A phylogenetic epistemology: the evolution of sensory organs, nervous systems, and effectors from protozoan to primate. In G. GREENBERG and E. TOBACH (eds.), Theories of the evolution of knowing,
pp. 123–133. The T.C. Schneirla Conference series, Vol. 4.
–––– 1999. Paleonemerteans as urbilaterians and hoplonemerteans as rudimentary vertebrates. American Zoologist 39: 11A.
–––– 2000. Abstract in abstract book of 5th International Conference on Nemertean Biology.
KRELL, F.-T. and CRANSTON, P.S. 2004. Which side of the tree is more basal? Systematic Entomology 29: 279–281.
LING, E.A. 1969. The structure and function of the cephalic organ of the nemertine Lineus ruber. Tissue & Cell 1: 503–524.
MACFARLANE, J.M. 1918. The causes and course of organic evolution. A study in bioenergics. Macmillan, New York.
MAYR, E. and BOCK, W.J. 2002. Classifications and other ordering systems. Journal Zoology, Systematics and Evoloutionary Research 40: 169–194.
STECKBECK, W. 1943. John Muirhead Macfarlane. September 28, 1855 – September 16 1943. Science 98: 487–488.
SUNDBERG, P., TURBEVILLE, J.M., HÄRLIN, M.S. 1998. There is no support
for Jensen’s hypothesis of nemerteans as ancestors to the vertebrates. Hydrobiologia 365: 47–54.
WILLMER, E.N. 1960. Cytology and evolution. 1st edition. Academic Press.
–––– 1970. Cytology and evolution. 2nd edition. Academic Press.
–––– 1974. Nemertines as possible ancestors of the vertebrates. Biological Reviews 49: 321–363.
Ronald Jenner Section of Evolution & Ecology, University of California, Davis,
CA 95616, USA
Created by Mark D Sutton on the 2006-02-23. (Version 2.0)