The origin of Darwin’s “abominable mystery”
Abstract
The phrase “Darwin’s abominable mystery” is frequently used with reference to a range of outstanding questions about the evolution of the plant group today known as the angiosperms. Here, I seek to more fully understand what prompted Darwin to coin the phrase in 1879, and the meaning he attached to it, by surveying the systematics, paleobotanical records, and phylogenetic hypotheses of his time. In the light of this historical research, I argue that Darwin was referring to the origin only of a subset of what are today called angiosperms: a (now obsolete) group equivalent to the “dicotyledons” of the Hooker and Bentham system. To Darwin and his contemporaries, the dicotyledons’ fossil record began abruptly and with great diversity in the Cretaceous, whereas the gymnosperms and monocotyledons were thought to have fossil records dating back to the Carboniferous or beyond. Based on their morphology, the dicotyledons were widely seen by botanists in Darwin’s time (unlike today) as more similar to the gymnosperms than to the monocotyledons. Thus, morphology seemed to point to gymnosperm progenitors of dicotyledons, but this hypothesis made the monocotyledons, given their (at the time) apparently longer fossil record, difficult to place. Darwin had friendly disagreements about the mystery of the dicotyledons’ abrupt appearance in the fossil record with others who thought that their evolution must have been more rapid than his own gradualism would allow. But the mystery may have been made “abominable” to him because it was seen by some contemporary paleobotanists, most notably William Carruthers, the Keeper of Botany at the British Museum, as evidence for divine intervention in the history of life. Subsequent developments in plant systematics and paleobotany after 1879 meant that Darwin’s letter was widely understood to be referring to the abrupt appearance of all angiosperms when it was published in 1903, a meaning that has been attached to it ever since.
The much-repeated term “Darwin’s abominable mystery” is derived from a private letter written in 1879 by Charles Darwin to his friend Joseph Hooker, Director of the Royal Botanic Gardens Kew. In his letter, Darwin stated: “the rapid development as far as we can judge of all the higher plants within recent geological times is an abominable mystery” (Darwin, 2019, pp. 336–337). When the letter was first published in 1903, the volume’s editors Francis Darwin and Albert Seward placed it under a page header “Evolution of Angiosperms” (Darwin and Seward, 1903, p. 21). “Darwin’s abominable mystery” has ever since been a commonly used appellation in the scientific literature for a range of unanswered questions about the origin and diversification of this group (e.g., Axelrod, 1952; Stebbins, 1965; Crepet, 2000; Frohlich and Parker, 2000; Davies et al., 2004; Berendse and Scheffer, 2009; Friedman, 2009). The scope of the mystery has been seen to encompass all aspects of the ancestry, age, environment, character, diversity, and diversification rate of the early angiosperms. Whether all these aspects address Darwin’s mystery is questionable (Friedman, 2009), and Bateman (2020) argues that some recent approaches are misguided. But over the past century, paleobotanists, palynologists, anatomists, molecular phylogeneticists, developmental biologists, paleoecologists, and genome biologists have all seen themselves as tackling the “abominable mystery” when they have researched angiosperm origins and diversification. Despite much progress in these areas, there is general agreement that Darwin’s “abominable mystery” remains intact, and the origin and diversification of the angiosperms remains one of the greatest open questions of the history of life.
However, it is easy to unintentionally impose current scientific definitions and questions upon Darwin’s 1879 statement, forgetting that views of plant taxonomy, systematics, paleobotany and evolution may have been different then than they are now. The term “higher plants” is particularly problematic as it assumes hierarchy, and its meaning would shift with changing ideas about plant relationships. Here, I build on previous research on the history of “Darwin’s abominable mystery” (Friedman, 2009; Buggs, 2017) and give detailed answers, based on contemporary literature, to the questions: What did Darwin mean by the “higher plants” in his letter to Hooker? What paleobotanical evidence was he referring to? From which taxonomic group were the “higher plants” thought to have evolved? Why was the mystery so abominable to him? Why was his letter interpreted as referring to all angiosperms when it was published in 1903?
PLANT SYSTEMATICS IN DARWIN’S LIFETIME
To understand what Darwin was referring to by “the higher plants” in his 1879 letter, we must understand how he and his contemporaries classified plants. Today, the major division commonly recognized among seed plants is between gymnosperms and angiosperms (e.g., Soltis et al., 2018). It is easy to imagine that this systematic division has been recognized since Robert Brown identified the morphological difference between gymnospermy and angiospermy in the 1820s (Brown, 1826), but this is not the case. Throughout Darwin’s lifetime, there was considerable uncertainty and disagreement about the systematic placement of taxa that we now consider to be gymnosperms, and the most important systematic division was perceived to be between the monocotyledons and dicotyledons of de Jussieu (1789). The monocotyledons and dicotyledons were seen as differentiated by their growth form, with monocots growing “by additions to the inside” (Lindley, 1830, p. xix) and dicotyledons growing “by the addition of successive layers of new matter to the outside” (Lindley, 1830, p. xix); thus, monocots were also commonly known as endogens, and dicots as exogens (de Candolle, 1813). Since the rise of molecular phylogenetics, the monocotyledons have been recognized as a monophyletic group within the angiosperms, but the group that used to be known as the dicotyledons is seen as paraphyletic: they are divided by the monocot clade into the eudicots and some smaller early-diverging groups (Soltis et al., 2018).
Many different systems of plant classification were used and published in Darwin’s time. These were “natural” systems, founded on the idea of “classing species according to the likeness they bear to each other” (Lindley, 1830, p. xi). They were not evolutionary, but Darwin and his followers interpreted them in evolutionary terms (see section on phylogeny below). I outline the systems of Darwin’s time below and in Table 1, with an emphasis on the placement of the gymnosperms and hence the usage of the term angiosperms.
|
Antoine Laurent De Jussieu 1789 )
Augustin Pyramus de Candolle 1813 )
Adolphe Brongniart 1828 )
John Lindley 1830 )
Stephano Endlicher 1836 –1840)
Adolphe Brongniart (1843)
Alphonse de Candolle (1844)
N.B. In Prodomus (1864) a new group was added: the Dicotylédones gymnospermes John Lindley (1846-53)
Alexander Braun (1864)
August W. Eichler (1876)
G. Bentham and J. Hooker (1862-1883)
|
Modern “gymnosperms” split between monocots and dicots
The system of Augustin De Candolle (de Candolle, 1813) split taxa that we now group as gymnosperms between monocots and dicots. Conifers were placed within the exogens (de Candolle, 1813, p. 218) and the cycads within the endogens (de Candolle, 1813, p. 219). Thus, there were no groups called angiosperms or gymnosperms. Darwin studied this system as a medical student in Edinburgh (Desmond and Moore, 1992), so it doubtless influenced his understanding of plant classification. From 1836 to 1840, Stephan Endlicher published a similar system (Endlicher, 1836). He included conifers and Gnetaceae in the “Gymnosperms,” placed within the Acramphibrya (a group approximately equivalent to the dicotyledons), but placed cycads along with ferns and clubmosses in the “Protophyta”, a grouping of equal rank with the Acramphibrya. The monocots were a separate group of equal rank, the “Amphibrya”. The term “angiosperm” was not used as the name of any group.
Gymnosperms within the dicots
In 1830, when Darwin was a student at Cambridge learning botany from John Henslow, John Lindley published a highly influential system of plant classification in English (Lindley, 1830). This separated the “Vasculares or Flowering Plants” into dicotyledons and monocotyledons, and separated the dicotyledons into two tribes: the Angiospermae and Gymnospermae. The tribe Gymnospermae contained two orders, the Cycadeae and the Conifereae. Lindley (1830) expressed uncertainty over whether or not Gnetum was truly gymnospermous but did not place it within his system. Thus, while the gymnosperms were seen as a natural grouping within the dicots, the Angiospermae tribe comprised only dicots and no monocots.
In 1843, the paleobotanist Adolphe Brongniart published a classification that divided the seed plants into monocotyledons and dicotyledons and subdivided the dicotyledons into Angiospermae and Gymnospermae (Brongniart, 1843). In 1844, an updated version of the De Candolle system placed both conifers and cycads within the Dicotylédones, in the group monochlamydées (de Candolle, 1844). In a volume of de Candolle’s Prodome published in 1864, Alphonse de Candolle placed the orders Gnetaceae, Coniferae and Cycadaceae within the Exogènes and grouped together as the Dicotyledonae Gymnospermae (de Candolle, 1864, p. 345).
Gymnosperms split from angiospermous monocots and dicots
In 1828, Adolphe Brongniart published a system in which the Phanerogames were separated into three “grandes divisions”: first the “gymnospermes”, then the “angiospermes monocotyledones” and finally the “angiospermes dicotyledones” (Brongniart, 1828a). The cycads and conifers were included in the “gymnospermes”. Thus, he used the terms angiosperm and gymnosperm in a similar sense to how we use them today. However, in 1843, Brongniart published a new classification that placed the “Gymnospermae” within the dicotyledons (see above).
Brongniart (1828a) divisions gained little support except in Germany. Alexander Braun’s system, published in a flora of the Province of Brandenburg (Ascherson, 1864), followed Brongniart (1828a) divisions. Braun divided the anthophytes or phanerogames into the Gymnospermae (which included Cycadeae, Coniferae, and Gnetaceae), followed by the Angiospermae divided into the Monocotyledones and Dicotyledones (Ascherson, 1864). Following Braun, in 1876, August Eichler placed the Gymnospermae before, and ranked equally with, the Monocotyleae and Dicotyleae (Eichler, 1876), though he did not include the term angiosperm as a grouping or in the title of the latter two groups.
Gymnosperms between monocots and dicots
In 1846, John Lindley published a new classification (Lindley, 1846) that removed the gymnosperms from within the Dicotyledons (as in his 1830 classification, see above), making the “Gymnogens” one of five classes of flowering plants, of equal rank with the “Rhizogens”, “Endogens”, “Dictyogens” and “Exogens”. What he had previously called Monocotyledons were divided between the Endogens and Dictyogens. The Gymnogens were placed between the Dictyogens and the Exogens, thus fundamentally dividing the monocots from the dicots. In George Bentham and Joseph Hooker’s highly influential classification of 1862 to 1883, the “Gymnospermae” were placed on equal rank with, and between, “Dicotyledones” and “Monocotyledones” (Bentham and Hooker, 1883). Thus, neither the Lindley (1846) or the Bentham and Hooker systems had a group named the angiosperms.
What were “all the higher plants”?
Given these contemporary systems of plant classification, the term “all the higher plants” would not have been understood in 1879 as a simple reference to the group that we know today as the angiosperms. The majority of plant classifications of Darwin’s time, especially in Britain (and including that proposed by Hooker himself), did not include a group called the angiosperms and did not even see what today we call the angiosperms as a natural group. The term “angiosperm” was rarely used in isolation, and when it was, it was rather ambiguous and tended to refer to a subset of the dicotyledons and excluded the monocots. Only the obscure systems of Brongniart (1828a) and Braun (Ascherson, 1864) used the term angiosperm to define a natural group that contained monocots and dicots. Based purely on the plant systems of the time, it is likely that Darwin was referring to the group referred to variously as “angiospermous dicotyledons” (Brongniart, 1828a), “Exogens” (Lindley, 1846), “Dicotyleae” (Eichler, 1876), or “Dicotyledones” (Bentham and Hooker, 1883) in his 1879 letter. The case for this interpretation is made stronger if we consider the state of paleobotanical knowledge at the time.
PALEOBOTANY IN DARWIN’S LIFETIME
Today, no pre-Cretaceous fossils are known that unambiguously represent angiosperms (Herendeen et al., 2017; Corio et al., 2019; Bateman, 2020). By contrast, in Darwin’s early scientific career, there seemed to be credible evidence that dicotyledonous and monocotyledonous plants had originated in ancient, not recent, geological times. Darwin read the first edition of Charles Lyell’s Principles of Geology (Lyell, 1830) on the voyage of the Beagle. In it, Lyell reported (p. 147) that a few fragments of dicotyledonous wood had been found in Carboniferous strata, in addition to cryptogamic and monocotyledonous plants. Lyell argued that although rare, the dicotyledonous fragments were highly developed enough to be “fatal to the doctrine of successive development” of plants. In the 1837 edition of Principles of Geology, which Darwin had in his library, Lyell elaborated further. He cited evidence from the Fossil Flora of Great Britain (Lindley and Hutton, 1831) that several highly developed monocotyledonous and dicotyledonous plants were present in the Carboniferous.
Darwin soon became aware of contradictions to Lyell, Lindley, and Hutton’s views from continental European paleobotanists. On his return from the voyage of the Beagle, Darwin read a lecture given by the leading French paleobotanist Adolphe Brongniart in 1837 (Brongniart, 1838; Buggs, 2017) and wrote in his notebook on the transmutation of species “no dicotyledonous plants and few monocot in coal formation?” and “Says coniferous structures intermediate between vascular or cryptogram (original Flora) and Dicotyledonous which nearly first appear at Tertiary epochs” (De Beer, 1960). Brongniart was clearly claiming that dicotyledons, which in Brongniart’s 1828 system excluded gymnosperms (unlike other contemporary systems, see above) were absent in the Carboniferous and appeared in what we would call today the late Cretaceous.
Brongniart had been arguing for a recent origin of the higher plants even before Darwin embarked on the Beagle. In 1828, in his Prodrome d'une histoire des végétaux fossiles (Brongniart, 1828b), he presented evidence that the dicotyledons did not appear in the fossil record until the Quaternary (Fig. 1). That Lyell, Lindley, and Hutton disagreed with Brongniart was not simply because Brongniart was classifying cycads and conifers as outside of the dicotyledons (see above and Table 1). It was also because the British paleontologists believed that fossil Stigmariae (today known to be the rooting structures of giant lycopsids) in the Carboniferous were allied to the Cactaceae or Euphorbiaceae (Lyell, 1837, p. 231) and thus were within what both they and Brongniart would class as dicotyledons.
During Darwin’s lifetime, the continental paleobotanists gradually began to believe in an earlier origin of the angiospermous dicotyledons as more fossils were found in the Cretaceous, and the British paleobotanists gradually began to believe in a later origin of the angiospermous dicotyledons as they realized that they were mistaken in their identifications of some Carboniferous fossils. Eventually, all agreed that angiospermous dicotyledons appeared suddenly and in great diversity in the Cretaceous.
…ou manquer complètement, ou ne s'annoncer que par quelques indices rares, douteux et très différents de leurs formes actuelles, signalant, du reste, plutôt la présence de quelques Monocotylédones que celle des Dicotylédones angiospermes. (Brongniart, 1849, p. 93)
[Translation: …either to be completely missing, or to be announced only by a few rare, doubtful indications and very different from their current forms, indicating, moreover, rather the presence of some monocotyledons than that of angiospermous dicotyledons.]
He listed 15 monocotyledons species in the Carboniferous but noted they were “très douteuses et imparfaitemant connues” (very doubtful and imperfectly known; p. 97). In the Triassic and Jurassic, he listed a few monocot species, but noted that all were “douteuses” (doubtful; pp. 102–103). He listed two reliable monocot species and several angiospermous dicotyledon species in the Cretaceous, then many from both groups in the Eocene, Miocene and Pliocene.
Le caractère dominant de cette dernière transformation de la végétation du globe, c'est l'apparition des Dicotylédones angiospermes, de ces Végétaux qui actuellement constituent plus des trois quarts de la creation végétale de notre époque, et qui paraissent avoir acquis celte prédominance dès l'origine des terrains tertiaires. Pendant longtemps j'avais pensé même que ces Végétaux ne commençaient à se montrer qu'après la craie, avec les premières couches des formations tertiaires; mais des recherches plus récentes ont constaté que des couches appartenant au terrain crétacé en présentaient déjà quelques exemples bien positifs. (Brongniart, 1849, p. 109)
[Translation: The dominant character of this last transformation of the vegetation of the globe, it is the appearance of the angiospermous dicotyledons, these plants which currently constitute more than three quarters of the vegetable creation of our time, and which seem to have acquired this predominance from the origin of the tertiary. For a long time, I had even thought that these plants only began to show themselves after the Cretaceous, with the first layers of tertiary formations; but more recent research has found that layers belonging to the Cretaceous already presented some very positive examples.]
[Translation: offering all the characteristics resulting from the predominance of dicotyledonous and monocotyledonous Angiosperms]
Les debris de végétaux monocotyledons sont généralement très difficiles à rapporter a leurs familles; car, si l'on excepte les Palmiers et un petit nombre de plantes dont les feuilles ont des caractères très particuliers, ces organes, les plus fréquents à l'état fossile, n'offrent que des caractères différentiels peu importants. Les fruits, qui sembleraient devoir nous conduire plus facilement à une détermination, manquent trop souvent de caractères de structure interne, et alors leur forme extérieure n'est qu'un indice assez vague. (Brongniart, 1849, p. 85)
[Translation: Fragments from monocotyledonous plants are generally very difficult to bring back to their families; for, if we except the Palm trees and a small number of plants whose leaves have very particular characters, these organs, the most frequent in the fossil state, offer only unimportant differential characters. The fruits, which would seem to lead us more easily to a determination, too often lack characters of internal structure, and then their external form is only a rather vague indication.]
Thus, Brongniart and his contemporaries may well have considered that, if monocots had been present in the Carboniferous, a few doubtful fragments were all that could be realistically expected as evidence for their existence. Others were willing to report Brongniart’s findings with fewer caveats than he himself used. For example, in 1853, in the 9th edition of Principles of Geology, Charles Lyell cited Brongniart as an authority and stated “It is remarkable that none of the exogens of Lindley (dicotyledonous angiosperms of Brongniart), which comprise four-fifths of the living flora of the globe, and include all the forest trees of Europe except the fir-tribe, have yet been discovered in the coal measures, and a very small number—fifteen species only—of monocotyledons.” (Lyell, 1853, pp. 133–134). In 1876, William Williamson published a paper arguing that one of Brongniart’s 15 carboniferous monocotyledons (Myeloxylon) was in fact a fern (Williamson, 1876).
One of the leading popularisers of paleontology of the mid-19th century was Hugh Miller. In 1857, in The Testimony of the Rocks (Miller, 1857, p. 8) he published a simple diagram of the plant fossil record (Fig. 2) summarizing current knowledge. This diagram showed monocotyledons from the Carboniferous onward, and uncertainty about the presence or absence of dicotyledons in the Oolitic (= modern Jurassic), with certainty of their presence in the Cretaceous. At the foot of the figure, he contrasted the order of Lindley’s natural system of plant classification with the order of the fossil record. In a footnote to this figure, Miller speculated that monocotyledons could in future be discovered in the Old Red Sandstone (= modern Devonian), which would give them an even earlier origin and bring the fossil record into closer agreement with Lindley’s system.
Darwin corresponded with continental paleobotanists whose discoveries helped to date the origin of the dicotyledonous angiosperms to the Cretaceous. One of these was Oswald Heer, Professor of Botany at the University of Zurich. According to Joseph Hooker in a letter to Darwin on 10 July 1862, his “old friend,” Heer, had “showed me a leaf apparently Dicotyledonous from the Lower Lias in Jura” (Darwin, 1997, p. 310). This fossil leaf appeared to push dicotyledons back to the Jurassic, but it was a short-lived identification. In 1875, Heer wrote to Darwin with a gift of the newly published volume III of his Flora Fossilis Arctica highlighting that within it he reported the discovery of the “oldest known dicotyledonous plant yet (naturally with the exception of the gymnosperms which are very different from true Dicotyledonae)”, a single species of “angiospermous Dicotyledoneae”, being a Populus. It was found in Greenland in an “older Cretaceous flora, which probably falls in the Urgonian [i.e., Barremian]” (Darwin, 2015, p. 525).
The interval from the Devonian to the Cretaceous is immensely long, and, as far as we know until now, the vegetable kingdom then consisted only of cryptogams, conifers & cycads and a few monocots. In the upper Cretaceous, however, the flora suddenly underwent a great transformation and … there appear for the first time the (angiosperm.) Dicotyledonae … It took an immensely long time for the first Dicotyledonae to arise, and as soon as they appeared, they developed at great speed". (Darwin, 2015, p. 525)
The sudden appearance of so many Dicotyledons in the Upper Chalk appears to me a most perplexing phenomenon to all who believe in any form of evolution, especially those who believe in extremely gradual evolution, to which view I know you are strongly opposed. The presence of even one true Angiosperm in the Lower Chalk makes me inclined to conjecture that plants of this great division must have been largely developed in some isolated area, whence owing to geographical changes, they at last succeeded in escaping and spread quickly over the world. But I fully admit that this case is a great difficulty in the views which I hold. (Darwin, 2015, p. 96)
Taken out of context, Darwin’s 1875 sentence mentioning “even one true Angiosperm” could be misinterpreted by modern eyes as referring to both monocotyledons and dicotyledons. Indeed, when this letter was first published in 1903 (in the same volume as Darwin’s 1879 letter to Joseph Hooker), the editors added a footnote: “No satisfactory evidence has been brought forward of the occurrence of fossil Angiosperms in pre-Cretaceous rocks. The origin of the Monocotyledons and Dicotyledons remains one of the most difficult and attractive problems of Palaeobotany" (Darwin and Seward, 1903, p. 239). This editorial gloss obscured Darwin’s meaning. He was simply slipping between the use of “dicotyledon” and “angiosperm” to refer to what Heer had called “the (angiosperm.) Dicotyledonae”. Heer’s letter, which was not published until 2015, clearly stated that monocots existed before the Cretaceous. Darwin did not dispute this claim in his reply, so he must only have been thinking of the dicotyledonous angiosperms when he referred to “one true Angiosperm”. Besides, the angiosperms in the modern sense were not a “great division” (Darwin, 2015, p. 96) in the major plant systems of the time (see above), but the dicotyledons were.
Heer’s findings regarding the overall patterns of the plant record were likely already known to Darwin as they were published 10 years before in his 1865 book Die Urwelt der Schweiz. Darwin possessed the 1872 French edition of this work in his library at Down House (https://www.biodiversitylibrary.org/docs/DarwinsLibraryBibliography.pdf). In this book, Heer clearly described “The entire absence of the Angiospermous Dicotyledons” until the mid-Cretaceous, and “their sudden abundant development in the lowest stage of the Upper Cretaceous (the Cenomanian)” (quoting from the 1876 English translation: Heer, 1876, vol. 1, pp. 243–245). He called this “one of the most important phenomena in the history of the development of nature.” (Heer, 1876, vol. 1, pp. 243–245). Later in his book he even used the word “mystery” to describe the alteration of old sets of species for new ones (Heer, 1876, vol. 2, p. 279).
… my ideas on the flower imply a long succession of changes from which the class of Dicotyledons would have come; but their sudden appearance about the bottom of the Cenomanian overturns all the calculations and brings us face to face with an unknown whose limits elude us. (Darwin, 2018, pp. 509–510)
Il n'y avait pas d'abord d'Angiospermes. Les Monocotylédones sont venues les premières, longtemps très-faibles numériquement, obscures et subordonnées. Puis, les Dicotylédones ont apparu à leur tour, et dès leur apparition…elles ont lutté en nombre et en importance, et n'ont depuis cessé dé croître et de se développer, tandis que les monocotylédones les suivent, sans atteindre jamais au même degré d'expansion, tout en accomplissant, de leur coté, un progrès analogue: tel est, en gros, l'ordre successif et la marche de la végétation en Europe, je puis ajouterdans le monde entier… (de Saporta, 1878, pp. 13–14).
[Translation: There were no Angiosperms at first. The monocotyledons came first, for a long time very weak numerically, obscure and subordinate. Then the dicotyledons appeared in their turn, and as soon as they appeared…they fought in number and importance, and have not ceased to grow and develop since then, while the monocotyledons follow them, without ever reaching the same degree of expansion, while making, for their part, a similar progress: such is, roughly, the successive order and progress of vegetation in Europe, I can add in the whole world.]
Thus, the expert paleobotanists with whom Darwin corresponded in the 1870s were convinced that dicotyledons originated in the Cretaceous, but monocotyledons had been in existence long before.
During the secondary period endogens are found in fossil deposits, few in number and obscure in their affinities; but the appearance of the higher type of exogenous plants is not disclosed by direct evidence until about the middle of the cretaceous period. (Ball, 1879, p. 579)
I have just read Ball’s essay. It is pretty bold. The rapid development, as far as we can judge, of all the higher plants within recent geological times is an abominable mystery. (Darwin, 2019, pp. 336–337)
Therefore, from published paleobotanical works of the time, as well as Darwin’s own correspondence, it is clear that in his view “the rapid development as far as we can judge of all the higher plants within recent geological times” was the first appearance of the angiospermous dicotyledons in the Cretaceous fossil record and their rapid diversification.
PLANT PHYLOGENY IN DARWIN’S LIFETIME
Every attempt that we make to gain a knowledge of the pedigree of any small or large group of organisms related by blood must, in the first instance, start with the evidence afforded by the existing “natural system” of this group. For although the natural system of animals and plants will never become finally settled, but will always represent a merely approximate knowledge of true blood relationship, still it will always possess great importance as a hypothetical pedigree. It is true, by a “natural system” most zoologists and botanists only endeavour to express in a concise way the subjective conceptions which each has formed of the objective “form-relationships” of organisms. These form-relationships, however, as the reader has seen, are in reality the necessary result of true blood relationship. Consequently, every morphologist in promoting our knowledge of the natural system, at the same time promotes our knowledge of the pedigree, whether he wishes it or not. (Haeckel, 1876, vol. 2, pp. 77–78)
As discussed above and shown in Table 1, the placement of gymnosperms, monocotyledons, and dicotyledons in the natural systems of the time varied a great deal. The majority placed gymnospermous trees next to wind-pollinated angiospermous dicotyledonous trees, as they were superficially similar in their reproductive structures and exogenous growth. Many argued that the Gnetaceae were intermediate between these groups. Lindley argued that Gnetum and Ephedra formed intermediates between these groups saying that in the Gnetaceae “we find precisely the structure and habit that would be wished for by a theorist searching for evidence to bring Gymnogens into communication with true Exogens” (Lindley, 1846, p. 232). When Joseph Hooker described Welwitschia for the first time, placing it in the Gnetaceae, he stated: “it is easy to suppose that we have in Welwitschia a transition in function, as well as in structure, between the gymnospermous and angiospermous Dicotyledons” (Hooker, 1863, p. 24). Darwin described this in a personal letter to Hooker as “a most grand case to connect two such groups” (Darwin, 1997).
The intermediacy of Welwitschia became commonly accepted, but there were differences of view among evolutionists about the order of branching of groups. In 1872, Eduard Strasburger, Professor of Botany at the University of Jena, argued for an evolutionary lineage in which angiospermous dicotyledons evolved from Gnetaceae, which had evolved from conifers (Strasburger, 1872). In his 1873 presidential address to the Linnean Society at its anniversary meeting, George Bentham argued that Strasburger was mistaken. He claimed that Welwitschia (Hooker, 1863) was “the nearest approach to (the least modified amongst the descendants of) the original type or parent stock of Dicotyledons which has reached recent geological periods” (Bentham, 1873, p. 18). He argued that this original type had given rise to two lineages, one leading to the conifers and the other to the “higher Dicotyledons”.
Thus, in the 1870s, there seemed to be a clear case for the evolution of dicotyledonous angiosperms from gymnosperms. But what of the monocotyledons? These proved much harder to fit in. In the 1860s, Ernst Haeckel made one of the first attempts to build a universal phylogenetic tree, or “pedigree” (Haeckel, 1868). Haeckel stood in the German tradition of Alexander Braun (see above), which saw a fundamental systematic divide between gymnosperms and angiosperms. Based on this system and the fossil record, he first proposed that angiosperms first evolved as monocots. In the first edition of his book Natürliche Schöpfungsgeschichte, Haekel published a fold-out diagram of a monophyletic pedigree of the vegetable kingdom (Fig. 3), that showed monocots evolving from cycads just before the Triassic and dicots evolving from the monocots just before the Cretaceous (Haeckel, 1868).
However, by the 1870s, Haeckel was persuaded that the angiosperms had arisen from the Gnetaceae and that dicots appeared before monocots. In an updated version of his pedigree diagram for his 1874 edition, he depicted monocots and monochlamydeous dicots evolving from the Gnetaceae in the Triassic, with other dicot groups evolving from the monochlamydeous dicots in the Cretaceous and Eocene (Haeckel, 1874). The 1876 translation of Natürliche Schöpfungsgeschichte into English under the title History of Creation, had the same fold-out monophyletic pedigree diagram as the 1874 German edition (Fig. 4), and stated “It is extremely probable that the Dicotyledons descended out of the Gnetaceae, but that the Monocotyledons descended later out of a branch of the Dicotyledons” (Haeckel, 1876, vol. 2, p. 112). In the 1874 German edition (p. 404), he also showed a diagram (Fig. 5) where the monochlamydeous dicots evolved first from Gnetaceae and gave rise to two separate lineages: the monocots and the dichlamydeous dicots. Haeckel could cite no evidence from the fossil record for his pre-Cretaceous lineages of Gnetaceae and monochlamydeous dicots, but assumed they must have existed, because they must have occurred before the arrival of monocots. Thus, his new phylogenetic hypothesis was more out of step with the fossil record than was his 1868 hypothesis.
Therefore, in the 1870s, there were two major possibilities for the evolution of the angiospermous dicotyledons, given that both monocotyledons and gymnosperms were thought to have much longer fossil records. The shared angiospermous character of the monocots and dicots recognized by a minority of continental systematists favored the evolution of dicots from monocots. This fit the order of the fossil record as it was understood at the time. But the morphological affinity that many systematists saw between dicotyledons and the gymnosperms, and especially the Gnetaceae, suggested that dicotyledons had evolved from gymnosperms, even though this fit the fossil record poorly if it was taken to imply that monocotyledons had evolved from dictoyledons. Thus, combining knowledge of plant paleontology and systematics posed difficulties for early plant phylogenetic hypotheses. Whichever hypothesis was true, the sudden appearance of diverse dicotyledons in the Cretaceous was a morphological leap of puzzling magnitude.
IMPACT OF THE MYSTERY IN DARWIN’S LIFETIME
To understand why the mystery was “abominable” to Darwin, we must consider the various contrasting inferences that were being drawn at the time from the sudden appearance of dicotyledonous angiosperms in Cretaceous rocks. Both Darwin’s followers and his critics recognised it as a crucial test case for his views. It provided a challenge to gradualism within evolutionary thought, but more importantly, a challenge to evolution itself, as some paleobotanists saw it as evidence for a divine hand at work.
Challenge to gradualism
Darwin knew that evolution must be a highly gradual process if natural selection is the main driver of the origin of new types. Many whom he persuaded of evolution, including Thomas Henry Huxley (Huxley, 1893), did not share his conviction that evolution must be gradual. The evolution of dicots seemed to be a case in point (Friedman, 2009). As Darwin commented in his 1879 letter to Hooker, the French paleobotanist Gaston da Saporta (see above) suggested “an astonishingly rapid development of the high plants,” (Darwin, 2019, p. 337) driven by coevolution between plants and insects. Darwin was not unfriendly to this idea (Friedman, 2009), but preferred the gradualist hypothesis that “perhaps there was during long ages a small isolated continent in the S. hemisphere, which served as the birthplace of the higher plants” (Darwin, 2019, p. 337).
Nothing is more extraordinary in the history of the Vegetable Kingdom, as it seems to me, than the apparently very sudden or abrupt development of the higher plants. I have sometimes speculated whether there did not exist somewhere during long ages an extremely isolated continent, perhaps near the South Pole. (6 August 1881, Darwin Correspondence Project, “Letter no. 13277,” accessed 11 August 2020, https://www.darwinproject.ac.uk/letter/DCP-LETT-13277.xml)
I have been so astonished at the apparently sudden coming in of the higher phanerogams, that I have sometimes fancied that development might have slowly gone on for an immense period in some isolated continent or large island, perhaps near the South Pole. (12 August 1881, Darwin Correspondence Project, “Letter no. 13288,” accessed 11 August 2020, https://www.darwinproject.ac.uk/letter/DCP-LETT-13288.xml)
Within the evolutionary camp, there was a debate to be had about the speed of evolution (an issue explained more fully by Friedman [2009]). But it is likely that the reason why Darwin saw the mystery as “abominable” was not merely due to the disagreements it generated with fellow evolutionists.
Challenge to evolution
At a broader level, Darwin’s contemporaries saw the sudden appearance of dicotyledonous angiosperms in Cretaceous rocks as a challenge to evolution as a natural process for the generation of novel types (Friedman, 2009). John Ball, who described himself as a “true disciple” (Ball, 1879, p. 565) of Darwin wrote: “To my mind there is no alternative between abandoning the doctrine of evolution and admitting that the origin of the existing types of flowering plants is enormously more remote than the period as to which we have direct evidence.” (Ball, 1879, p. 580). For no less a figure than Oswald Heer (see above), the sudden appearance of new forms, with the development of the dicotyledons as a supreme example, was evidence against evolution and for creation.
The flora and fauna of the present day represent the most highly organized forms. Consequently in approaching a more perfect state of the vegetable and animal kingdoms, an advance in the organization of living creatures took place, and a definite progress may be traced in their development … There is therefore unmistakably, on the whole, a progressive development from the more simply constructed towards the more complicated and consequently more highly organized types. (Heer, 1876, vol. 2, p. 276)
He also held that new types of organisms derived “from the organic [rather] than from the inorganic world” so that “a genetic connexion exists through the whole organic world” (Heer, 1876, vol. 2, p. 282). In other words, he believed in common ancestry.
However, unlike Darwin, Heer believed that sudden appearances of new forms pointed to punctuated instances of large change in the history of life. He called this the “remoulding of species” (Heer, 1876, vol. 2, p. 289). He suggested that this remolding was somewhat analogous to metamorphosis, but was “a secret, an enigma, in the explanation of which may be exercised the talents of divination, but which has not been fully and entirely solved either in the known phenomena of nature, or by the application of established physical laws” (Heer, 1876, vol. 2, p. 290). He referred to episodes of remolding as “times of creation” (Heer, 1876, vol. 2, p. 291). He also believed that “there was a primaeval epoch during which the first species were brought into being…an act of creation” (Heer, 1876, vol. 2, p. 291).
An advanced knowledge of Nature leads to a profound conviction that the enigmas of the natural world and of human life can only be solved by a belief in an Almighty Creator, and in the creation of the heaven and the earth by Divine wisdom according to an eternal and preconceived plan. (Heer, 1876, vol. 2, p. 293)
Darwin was aware of Heer’s disagreement with him. As quoted above, Darwin noted in an 1875 letter to Heer “I know you are strongly opposed” to “extremely gradual evolution” (Darwin, 2015, p. 96). Years before, in an 1863 letter to Asa Gray, Darwin mentioned “Heer’s view of species arising suddenly by monstrosities” (Darwin, 1999, p. 465), immediately noting that this view would be impossible to explain by chance alone and would require creationism.
The case for a divine hand in the origin of the dicotyledons in Cretaceous rocks was made in the 1870s more forcibly, and closer to home, by William Carruthers, Keeper of Botany at the British Museum, and a leading paleobotanist of his day (Fig. 6). Carruthers devoted his presidential address to the Geologist’s Association to this topic on 3 November 1876 at a meeting held in the library of University College London. This address was published in the Proceedings of the Geologists’ Association (Carruthers, 1877a) and reprinted as a pamphlet entitled Fossil Plants and their Testimony in Reference to the Doctrine of Evolution (Carruthers, 1877c; Fig. 7). In it, Carruthers argued that the appearance of “Vascular Cryptogams and seed-bearing Gymnosperms” in Devonian rocks, and the appearance of monocotyledons in the Carboniferous, occurred in the absence of “an innumerable series of gradually advancing steps” (Carruthers, 1877c, p. 28), and was therefore strong evidence against evolution. But the “testimony” of the “higher or Dicotyledonous division of the flowering plants” was “more important” (Carruthers, 1877c, p. 29) evidence against evolution for three reasons. (1) They have “higher organization” that is “sharply separated” from monocotyledons and gymnosperms; (2) they have “numerous differences” that allow their “systematic classification”; and (3) they appear in young strata that are well-known (Carruthers, 1877c, p. 30).
Carruthers devoted the final third of his address to this topic, arguing that there was “no evidence whatever” that lower forms of dicotyledons had existed in “the Triassic and Jurassic periods” (Carruthers, 1877c, p. 31), despite abundant evidence for a rich plant fossil record spanning those periods. He also pointed out that when the dicotyledons do appear in the Cretaceous they appear as a wide diversity of forms that could be readily identified as belonging to extant genera.
Taking the willow genus (Salix) as an extended example, Carruthers described how fossil evidence supported the common ancestry and evolution of the numerous willow species: “it is easy, then, to construct an exact phylogenetic tree of the genus Salix” (Carruthers, 1877c, p. 33). Beyond the genus level, it is not difficult to imagine “a generalized form” between Salix and Populus, but “there is no record of such a form” in the rocks. Indeed “the two genera appear together among the earliest known Dicotyledons” (Carruthers, 1877c, p. 34). If we want to trace an evolutionary tree further back than this, “every step in this phylogenetic tree must be imagined” (Carruthers, 1877c, p. 34). It is “unsupported by a single fact” (Carruthers, 1877c, p. 34). There is no time available in the fossil record for the evolution of willows by “slow and imperceptible changes” from a “generalized Angiosperm” (Carruthers, 1877c, p. 34).
Carruthers concluded that “the facts of palaeontological botany are opposed to evolution” but like Oswald Heer, he agreed that they did “testify to development, to progress from lower to higher types” (Carruthers, 1877c, p. 35). But this fit better with creation than evolution: “development is not the property of the evolutionist; indeed the Mosaic narrative—the oldest scheme of creation—which traces all nature to a supernatural Creator represents the operations of that Creator as having been carried out in a series of developments” (Carruthers, 1877c, p. 35).
This challenge to evolution itself came from an acknowledged expert on plant fossils at the heart of the British scientific establishment. Carruthers was one of the few full-time, paid paleobotanists in Britain, responsible for its greatest collection of fossil plants, and a prolific researcher. By 1876, he had published 62 research articles (Anon., 1912). He had examined many fossils from pre-Cretaceous rocks that other researchers had suggested as putative monocots or dicots and shown them not to belong to these groups (e.g., Carruthers, 1870, 1869, 1871, 1872a, b). He had also identified some fossils in Jurassic rocks as monocotyledons (Carruthers, 1867, 1868). He is best known today for describing and naming the Bennettites (Carruthers, 1870). Carruthers was, like Darwin, a Fellow of the Royal, Linnean and Geological Societies. He served on the Council of the Royal Society from 1877 to 1879. He would go on to be elected President of the Linnean Society in 1886.
Unlike Oswald Heer, Carruthers did not have a friendly correspondence with Darwin. He was a colleague of one of Darwin’s leading critics, Richard Owen, head of the Natural History Collections at the British Museum. In the early 1870s, Carruthers stymied an attempt by George Bentham and Joseph Hooker to have the British Museum herbarium moved to Kew (Carruthers, 1872c), which did not endear him to Darwin’s circle of friends. In 1874, Carruthers opposed changes to the Linnean Society bylaws introduced by George Bentham when Bentham was President, contributing to Bentham’s resignation; this episode led Hooker to describe Carruthers to Darwin as an “Owenised Scotchman (of old Crawford “d–d Scotchman” type)” (3 March 1874, Darwin Correspondence Project, “Letter no. 9331,” accessed 11 August 2020, https://www.darwinproject.ac.uk/letter/DCP-LETT-9331.xml), with Darwin replying “what an odious man he seems to be” (4 March 1874, Darwin Correspondence Project, “Letter no. 9333,” accessed 11 August 2020, https://www.darwinproject.ac.uk/letter/DCP-LETT-9333.xml).
There can be no doubt that Darwin was aware of Carruthers’ use of the sudden appearance of dicotyledons in the Cretaceous as evidence against evolution. Carruthers’ 1876 presidential address to the Geologists’ Association was reported at length in The Times and Darwin was in the fixed habit of reading this newspaper after lunch every day (Freeman, 1978). Opening The Times on Monday, 6 November 1876, Darwin would have seen on page 6 the headline “Plant Evolution”, followed by a 1300-word article recounting Carruthers’ argument in detail. A detailed resume of the lecture was provided to the Gardener’s Chronicle by Carruthers and printed in its 18 November 1876 issue (pp. 644–645). Darwin kept a clipping of this article (CUL-DAR205.9.45-46; van Wyhe, 2002). Darwin also kept a clipping (CUL-DAR205.9.49-50; van Wyhe, 2002) of a lengthy response to Carruthers published in Nature on 29 March 1877 by J. Starkie Gardner (1877). Gardner’s response relied heavily on arguments that the fossil record was incomplete. He also cautiously suggested that leaves from the Dakota formation might represent early dicotyledons, not modern genera, and had the potential to provide a new angle on the debate. Darwin appears to have been unsatisfied with Gardner’s response as he did not refer to it in his subsequent letters on the subject.
Carruthers’ 1876 address was also published in The Geological Magazine (Carruthers, 1876) and in The Contemporary Review (Carruthers, 1877b). In 1877, when the paper based on his address was published in the Proceedings of the Geologists’ Association (Carruthers, 1877a), Carruthers sent copies of the reprint (Carruthers, 1877c) pamphlet to colleagues. The copy held in the library at Royal Botanic Garden Kew (Fig. 7) was sent to Myles Berkeley, a mycologist who worked closely with Joseph Hooker. Berkeley had earlier identified the fungal samples collected by Darwin on the voyage of HMS Beagle.
In 1879, Darwin may have felt that the plant fossil record had been very publicly weaponized against him by an established expert, whom he had previous reason to dislike. An abominable predicament indeed.
EVOLUTION OF THE MYSTERY: 1879 TO 1903
Darwin’s 1879 letter to Hooker lay unpublished until 1903. During those 24 years, important developments occurred in plant systematics and paleobotany. This progress led to the “abominable mystery” being understood, from its publication onward, in terms of the sudden appearance of all angiosperms (in the still-current sense of the term) in the mid-Cretaceous.
By 1903, it was commonly accepted by systematists that angiosperms versus gymnosperms was a fundamental division and that the monocotyledons and dicotyledons were groups within the angiosperms. This development is shown in Table 2, in the post-1879 systems of August W. Eichler (1886), Frank L. Ward (1885), Adolf Engler and Karl A. Prantl (1887–1915), Charles E. Bessey (1894), and Hans Hallier (1905). The Engler and Prantl system was particularly influential and widely adopted. This shift in systematics was based on a more explicitly evolutionary approach but occurred after Darwin’s letter to Hooker and after Darwin’s death.
August W. Eichler (1886)
Frank L. Ward (1885)
Adolf Engler and Karl A. Prantl (1887–1915)
Charles E. Bessey (1894)
Hans Hallier1905)
|
Discussion of this change in plant systematics can be seen in the literature of the time. For example, in his 1885 book, Sketch of Paleobotany, Lester F. Ward described the “prevailing system” as one where the group called angiosperms were a subset of the dicotyledons, and he suggested his own “modified system” that placed the monocotyledons and dicotyledons within the angiosperms. In 1894, Charles Bessey felt it necessary to add as a footnote to a paper: “I have not deemed it necessary to refer to the positions accorded to the Gymnosperms in most of our systematic works. It is so manifest an error as to need no discussion here. The flowering plants referred to in this paper are the Angiosperms, both Monocotyledons and Dicotyledons” (Bessey, 1894, p. 244). Some older botanists remained skeptical of the newer systems. Even in 1892, the sixth edition of George Bentham’s Handbook of the British Flora, revised by Joseph Hooker, still placed the Coniferae within the Dicotyledons (Bentham, 1892). In 1929, Raymond Pool wrote that Engler and Prantl system “has been the dominant scheme of plant classification in most of the world since 1900” (Pool, 1929, p. 146).
There was also a shift in understanding of the plant fossil record between 1879 and 1903, regarding whether or not monocotyledons could be found before the Cretaceous. In 1885, Ward’s Sketch of Paleobotany informed readers that monocotyledons had been found in the Carboniferous. The same year, Gaston de Saporta and A.-F. Marion suggested that many supposedly pre-Cretaceous monocots were actually mere precursors to the angiosperms and called them “proangiosperms” (de Saporta and Marion, 1885). They saw this as a clear shift from Brongniart’s views and de Saporta’s previous views. However, they still held that a few true monocots could be found in the Jurassic. Starkie Gardner reviewed this evidence in 1886 and concluded that though many supposedly pre-Cretaceous monocot fossils were unconvincing, “the evidence tends to show that Monocotyledons (?) of some sort existed as far back as the Trias…and that during the Jurassic decided Monocotyledons…flourished” (Gardner, 1886, p. 204). By 1891, de Saporta argued that no pre-Cretaceous fossils were truly monocots, and those that had appeared to be should be seen as “proangiosperms” (de Saporta, 1891).
In the 1890s, Albert Seward undertook a comprehensive review of all Jurassic and Lower Cretaceous plant fossils held at the British Museum (Natural History). This huge project convinced him that, contrary to previous reports, no monocot fossils existed from the Jurassic or Lower Cretaceous. In the second volume of his review, published in 1895, Seward noted: “Many of the supposed oldest monocotyledonous plants have been shown to be either inorganic fossils, or to belong to some other class of plants” (Seward, 1895, pp. 170–171). At the end of this volume, he concluded: “we search in vain among the abundant samples of the Wealden [Lower Cretaceous] vegetation for any fragments of monocotyledonous or dicotyledonous plants” (Seward, 1895, p. 240). In 1896, he published a paper in Annals of Botany critiquing evidence for pre-Cretaceous monocots, concluding that there was no proof that they existed (Seward, 1896). In the fourth volume of his British Museum review, published in 1904, Seward argued that no monocot or dicot fossils had been found in the Jurassic or Lower Cretaceous, with the possible exception of a Jurassic leaf from Stonesfield in Oxfordshire that appeared to be dicotyledonous (Seward, 1904a). In his comments on this leaf, he quoted Darwin’s 1879 letter in the context of “the suddenness with which the Dicotyledons took their place in the vegetation of the world” (Seward, 1904a, pp. 154–155).
This state of knowledge concerned Albert Seward. In 1903, he published a rallying-call in the New Phytologist for a “quest for evidence bearing on the ancestry of the Angiosperms” (Seward, 1903, p. 244) in paleobotany. In recent years, this area had been “almost completely neglected”, he claimed. This neglect was partly because the area was perceived as a seriously difficult one for research, he said, and partly because people had been persuaded that “Proangiosperms” had been found. “These so-called Proangiosperms are, I believe, in many instances of no botanical value and their designation by so alluring a title is not justified by the facts,” Seward warned (p. 244). Workers were urgently needed for the “laborious task” of “an organised exploration of the later plant-bearing strata” (p. 244).
The same year, at the main meeting of the British Association for the Advancement of Science in 1903, as President of the Botany Section, Seward described the global lack of angiosperm fossils from beneath the mid-Cretaceous and the “amazing rapidity” with which they assumed “the leading role” thereafter. He quoted Darwin’s “abominable mystery” letter. Toward the end of his speech, he told the assembled scientists that “One of our most pressing needs” was more paleontological research on the origin of the angiosperms. “This is a task which is sometimes said to be impossible or hardly worth the attempt…but it is at least a praiseworthy aim not to say a duty…” (Seward, 1904b, p. 847) he urged them.
Thus, when Albert Seward (together with Francis Darwin) edited Charles Darwin’s 1879 letter to Joseph Hooker (Darwin and Seward, 1903), he was newly convinced that no angiosperms or clear precursors to angiosperms—either monocot or dicot—had been found in the fossil record, and he was keen to encourage fresh research effort in this area. The publication of Darwin’s 1879 letter to Hooker fit with this concern perfectly. Even if Darwin had meant the dicots when he referred to the “higher plants”, he would surely have agreed that the mystery encompassed all angiosperms had he been alive in 1903. Albert Seward and Francis Darwin could no doubt have explored Darwin’s meaning more contextually, but this would have been cumbersome and distracting, and they were editing the book of his letters as active scientists excited about current problems, not as historians. From then on, Darwin’s abominable mystery was inextricably identified with the origin of all angiosperms.
CONCLUSION
Darwin’s “abominable mystery” was to him in 1879 the sudden appearance of angiospermous dicotyledonous fossils in great diversity in the Cretaceous. It was not about all angiosperms, because angiosperms were not seen as a natural group in the great majority of systems of plant classification employed at the time, and monocotyledons were thought to occur much earlier in the fossil record. The apparently long fossil records of both monocotyledons and gymnosperms presented both groups as possible direct progenitors of the dicotyledons, the majority view in the 1870s being that the dicotyledons had evolved from the gymnosperms, specifically from the Gnetaceae. The sudden appearance of diverse dicotyledons was a challenge to Darwinian gradualism, but more importantly, to evolution itself, encouraging prominent paleobotanists such as Oswald Heer and William Carruthers to argue that the plant fossil record gave evidence for divine intervention in the history of life. This concatenation of circumstances led Darwin to exclaim to his friend and ally Hooker that “the rapid development as far as we can judge of all the higher plants within recent geological times is an abominable mystery” (Darwin, 2019, p. 336). Between 1879 and the publication of Darwin’s letter in 1903, plant systematics and paleobotany developed in a way that made it natural to read Darwin’s letter as referring to the abrupt appearance of all angiosperms when it was published, and it is this meaning that has been attached to it ever since.
ACKNOWLEDGMENTS
I thank Richard Bateman for insightful conversations and helpful suggestions on the text of a previous version of this manuscript. I thank four anonymous reviewers and American Journal of Botany editors for helpful comments and suggestions.
