Constancea 83, 2002
University and Jepson Herbaria
P.C. Silva Festschrift

Ontogeny v. Phylogeny:
The Strange Case of the Silicoflagellates

Phil Parkinson
Research Centre
Alexander Turnbull Library
P.O. Box 12349
Wellington, New Zealand


The silicoflagellates are a small group of protists, variously categorised as flagellates (by zoologists) and chromists (by botanists). By geologists and palaeontologists, however, they are classified as as ‘protist microfossils’ and the geologists have taken a greater interest in them than biologists. These three different and conflicting regnal perspectives pose difficult conceptual problems for everyone. What are silicoflagellates? In the 19th century they were usually considered as protozoans but in the 20th century they were recategorised as algae (explicitly a category of plants) leaving them with a bi-regnal ambiguity which is reflected in both the conventional classification schemes and in the nomenclatural systems in which they were placed. Until electron microscopy began to shed light on their ultrastructure nobody knew much about their biology. As they were of negligible medical or economic importance and, above all, as no coherent view of their life histories was available, they were studied by only a few scientists. Botanists have paid the silicoflagellates little attention and the group was not even mentioned in that phycological classic the Sylloge Algarum of De Toni (1889–1924).

Until a few decades ago the silicoflagellates were usually named in general texts as if they were animals. Mandra (1969), for example, treated ‘Silicoflagellata’ within ‘Protozoa: Mastigophora’. Although Ehrenberg first saw living silicoflagellates in 1839, most of his studies were based on fossils and the names he adopted all have fossil types. The orthodox taxonomy of the group is founded on the interpretation of structural efficiency in the assembly of so-called ‘apical’ skeletal elements within a previously formed ‘basal’ element. In some taxa the ‘apical’ element does not exist. Thus the mechanical efficiency of the skeleton is the basis of the taxonomy. In fact, however, the skeleton of living cells is lateral to the protoplast and does not exist during all phases of the life history. That a classification is founded on an ephemeral life history stage calls into question its utility. Over the last decade ultrastructural studies by Moestrup and Thomsen (1990), and by Henriksen et al. (1993), together with earlier studies by Van Valkenberg and Norris (1970) and by Van Valkenberg (1971a, 1971b, 1980) have confirmed that silicoflagellates have a distinctive body-plan and are most closely related to the pedinellids and actinomonads, segregates of the so-called Chrysophyceae, a group now regarded as polyphyletic and heterogeneous. While the higher classification remains a matter of contention they are now usually classified in the ‘stramenopiles’ of Patterson (see e.g., Patterson, 1994 p. 6) or the ‘Chromista’ of Cavalier-Smith and others. They form part of the the informally named ‘axodines’, part of the ‘stramenochromes’ within the ‘stramenopiles’ of Patterson (in his classification of Eukaryotes: see, a monophyletic assemblage which includes the Actinophryidae, Pedinellidae, Rhizochromulinidae, and Dictyochidae — to use the names adopted by Patterson (1999) in his paper ‘The diversity of eukaryotes’ and in the second edition of the Illustrated guide to the Protozoa (2002). Patterson's classification retains the term ‘Silicoflagellata’ and characterises the group in terms of its ultrastructure. The group to which silicoflagellates belong has been variously considered as a single genus (Dictyocha Ehrenberg, 1837) or ranked as a family, order or class (Class Dictyochophyceae Silva, 1980).

The classifications used in the present day are the legacy of the enthusiastic adoption, in the 1860s to 1880s, of evolutionary ideas which often turned out to be wrong. In the works of Ernst Haeckel, particularly, there was a strong emphasis on a doctrine— “ontogenesis is a brief and rapid recapitulation of phylogenesis” 1 which was to have an effect on not only the radiolarians — the taxonomy of which has never recovered from his taxonomic attentions — but also on some of the others group he studied, the phaeodarians and the silicoflagellates. Although the ‘law of recapitulation’ has long been discredited, its nomenclatural legacy survives to torment us today.

Apart from the problem of classification, however, there is the related problem of determining whether botanical or zoological names should be used. At present there appears to be a consensus (disregarded only by the Society of Protozoologists (Lee et al. 2002), responsible for IG-2) that the nomenclature of the silicoflagellates should be governed by the International Code of Botanical Nomenclature (hereinafter ICBN) rather than the International Code of Zoological Nomenclature (hereinafter ICZN). The current edition of the International Code of Botanical Nomenclature (Greuter et al., 2000; hereinafter ICBN 2000, the ‘St. Louis Code’) was adopted in 1999 and published in 2000. 2 It supersedes all previous editions and “The Rules of nomenclature are retroactive unless expressly limited.” (Principle VI).

Because of these changes in the Rules, practices followed in the past do not always apply today. Legal determinations, apparently correct at the time, may now be incorrect, following changes in the Rules. Major changes (for the better) introduced at St. Louis particularly affected the use of special purpose names for what are now termed ‘morphotaxa’. These changes and their effects on classification and nomenclature are discussed in Part II. Although the ICBN now claims jurisdiction over photosynthetic chromists most of their historical nomenclature has followed the older zoological conventions. Phycologists who come across living silicoflagellates are likely to distinguish only three pretended species and these are usually called Dictyocha fibula, Dictyocha speculum and Octactis pulchra. There is continuing doubt as to whether these are three distinct species, two or only one.

Part I: Silicoflagellate biology and palaeontology

The silicoflagellate body-plan

Silicoflagellates are photosynthetic chromists (or ‘stramenopile protists’, to use Patterson's terminology) with a multi-stage life-cycle, not all stages of which are known. The best-known stage consists of a naked cell body with a single anterior flagellum and numerous plastids contained within an external lateral skeleton. The cells of this ‘skeleton-bearing’ stage have a single colourless round or oval nucleus and clear cytoplasm which is extended to form pseudopodia (sometimes called ‘tentacles’). The ultrastructure of the protoplast has been described, by Van Valkenberg, as consisting of a centrally placed nucleus surrounded by a dense band of cytoplasm (‘perikaryon’) containing all the numerous dictyosomes but no plastids. Numerous cytoplasmic strands extend from the perikaryon to connect with peripheral globular masses of cytoplasm (pseudopodia) which contain the plastids but no dictyosomes. Mitochondria and various vesicles are observed in both the pseudopodia and the perikaryon. Moestrup and Thomsen (1990) give a detailed and profusely illustrated account of the ultrastructure of what they have chosen to call Dictyocha speculum.

The feature which conspicuously distinguishes the skeleton-bearing stage from other phytoflagellate protists is a skeleton composed of hollow beams of opaline silica. These are of various shapes and are variously ornamented with spines or other processes. The skeleton functions as a scaffold, supporting the ‘sunburst’ arrangement of pseudopodia and cytoplasmic strands during the photosynthetic, motile, vegetatively reproducing phase of the life history. Average generation time in culture was 49 hours according to Van Valkenberg and Norris (1970). Studies of silicoflagellates in culture, however, have suggested that the shape of the skeleton is highly plastic and varies in response to the physiology of the environment: “It is generally agreed that silicoflagellates are highly temperature-sensitive, a given strain tolerating only a few degrees of temperature variation from its normal range.” (Van Valkenberg, 1980 p. 335). Development of a skeleton (with or without mitosis of the protoplast) may take as little as two days. Some silicoflagellates in culture, however, do not develop skeletons at all, as, according to Van Valkenberg and Norris (1970 p. 51):

Cells without skeletons appear to have a tough periplast or envelope not observed on cells with skeletons. In some cases the periplast is loosely fitting, but at other times the periplast is completely filled by the protoplast. . . . Some cells without skeletons have been observed to retain their flagella, and to divide and continue swimming. More often, however, they settle to the bottom of the container where a few attach to the glass. Here they either remain as single unflagellated cells, occasionally dividing but remaining uninucleate, or undergo nuclear division without cytokinesis. The latter multinucleate organisms have a flexible periplast and may exhibit slow amoeboid movement. During such movements the discoid yellow-brown chromatophores appear to be held in a network of fine filaments close to the periplast, and to move with the periplast. Coenocyte formation occurs under the following conditions: (1) when a 10 C culture passes its maximum density, (2) frequently in 15 C cultures at any density, and (3) when cultures are maintained in continuous light. The addition of glycolic acid to the medium along with an increase in light intensity appears to increase the size of the coenocytes.

The existence of the coenocyte multinucleate stage in nature has been established on one occasion (Henriksen et al., 1993 p. 38). Skeletonless swimming cells grown in continuous light have been observed to form new skeletons when placed under 16-hr day conditions. (Van Valkenberg and Norris, 1970 p. 53). A distinct naked flagellate stage of the life cycle was first described by Moestrup and Thomsen (1990). The skeletal variation induced in living clonal cultures, raised under different environments, has been claimed to cross the conventional generic boundaries established for the silicoflagellate palaeotaxa. In Van Valkenberg's clonal culture (Van Valkenberg and Norris, 1970 p. 53), for example:

Two hundred skeletons from our cultures, picked at random and all originating from the same clone, were identified as belonging to the following taxa according to Gemeinhardt's monograph [Gemeinhardt, 1930]: Distephanus speculum (Ehrenberg) Haeckel [sic], 17 skeletons; Distephanus speculum var. septenarius (Ehrenberg) Joerg., 6; Cannopilus calyptra Haeckel, 1; Dictyocha fibula Ehrenberg, 14; Dictyocha fibula var. pentagona Schulz, 89; Dictyocha hexacantha Schulz, 50. Twenty-three skeletons were not identifiable with any of the above species. . . . When these same skeletons were examined in the light of the simplified scheme of Deflandre . . . who listed the only 3 living species as Dictyocha fibula, D. speculum, and D. octonaria [i.e., Octactis pulchra], we found them to be a mixture of D. fibula and D. speculum.

Others, noting that teratological skeletons are relatively infrequent in examination of sediments, suggest that these deformed skeletons are artifacts of the culture method and little reliance should be placed upon them until they are confirmed by other investigators. What can be said with certainty of the silicoflagellates, is that there is no consensus on how they should be classified, at either generic or specific rank.

Sexual reproduction in silicoflagellates is unknown, but asexual reproduction by cellular division — accompanied by replication of the skeleton as a mirror image of the ‘parent’-cell has been observed (Gemeinhardt, 1930; Henriksen et al., 1993, etc.). Such a pattern of reproduction has also been observed in fossil silicoflagellates (e.g., by Stöhr, 1880) although, at the time, it was not recognised for what it was. Several excellent illustrations of what is there called Distephanus pulchra (i.e., Octactis pulchra) are given by Hsin Yi Ling and Takahashi (1985, Plate 1: fig. 5, Plate 2: fig. 1–2) and show mother and daughter skeletons locked together. Rapid asexual reproduction by this method may lead to bloom conditions (Gran, 1935) and this explains why abundant localised silicoflagellate fossil deposits exhibit considerable structural consistency in the skeletal morphology: the replication process ensures that the structure is cloned repeatedly.

Nevertheless, as shown by Van Valkenberg, “mistakes” in skeletal replication do occur. The number of sides in a single clonal culture ranged from four to seven, with numerous irregular specimens as well. The precise process by which skeletal formation takes place, however, is not well understood. The suggestion by Van Valkenberg (1971a) that: “Skeleton formation may proceed from an initial series of dichotomous outgrowths of an elongate formative vacuole, which later fuse at junctures of the basal ring and supporting bars. The presence or absence of an apical ring may be a function of variation in the duration of pre-division among individuals within the same clone” accounts adequately for the observed variation.

Living skeleton-bearing silicoflagellates identified as Dictyocha fibula or as Dictyocha speculum have a flattish ‘abapical’ side (for palaeontologists, this is termed ‘basal’, as above) and, usually, a more or less peaked ‘apical’ side, the whole organism often resembling a stirrup or a wide-brimmed hat. The crown of the hat contains the nucleus and greater part of the cell body, while the cytoplasm and plastids are distributed over the whole surface. The silicoflagellate brim is often extended by spinous processes and, as pointed out by Lipps (1970) “silicoflagellates are morphologically adapted to function as inclined planes during swimming.” However, as pointed out by Moestrup and Thomsen, the skeleton is actually lateral to the living cell and the palaeontological descriptive terminology ‘abapical’ is therefore rather inappropriate, as are the terms ‘basal’ and ‘apical’. In the other living species, Octactis pulchra Schiller, 1925, the so-called ‘basal ring’ is without secondary processes, but the more fragile so-called ‘apical ring’ appears to grow by the extension inward of the clathrus (basket) which holds the protoplast. The skeletal details, now obvious from SEM images, were not apparent to either Schiller in 1925 or to Ehrenberg in 1843 (compare the images by Hsin Yi Ling and Takahashi (1985 Plate 1: fig. 1 (Schiller), Plate 1: fig. 2–3 (Ehrenberg's specimens redrawn by Locker), and Plate 2: figs. 5–6 (Ehrenberg's figures of the same specimens). As Hsin Yi Ling and Takahashi suggest, it is plain that the microscopes in the 1840s did not reveal what SEM can reveal now.

The swimming of live silicoflagellates was observed by Ehrenberg as early as 15 September 1839. He noted that “the living animalcule has cells filled with a green contents, in which small bubbles and very fine bodies can be seen. Movement from place to place was very slow, observed over a long period.” (Ehrenberg, 1840a, my translation). The not-very-vigorous motion of the flagellum may be more effective in maintaining the orientation of the cell towards the ambient light rather than in facilitating lateral motion. Silicoflagellates live in the upper part of the water column, and are adapted for life in tropical, temperate, and frigid waters. The same structures occur in all seas and oceans. The shape of the skeleton is related to the physical and, perhaps, the chemical nature of the environment in which they live, assisting with resistance to sinking, orientation, and locomotion in ocean water of widely differing temperature, chemical conditions and illumination. Ultrastructure of the protoplast, which provides the biological basis for the higher classification of the silicoflagellates, was described by Van Valkenberg (1971a, 1971b) and by Moestrup and Thomsen (1990). A cladistic analysis of the class including the silicoflagellates, the pedinellids and rhizochromulinids — which are currently ranked as separate orders — has been published by Daugbjerg (1996).

The taxonomy of silicoflagellate skeletons

The taxonomies adopted in the palaeontological disciplines have become complicated by the speculations of researchers who are convinced that the wide variation observed in the skeletal morphology merits phylogenetic interpretation. In this respect they are following in the steps of Ernst Haeckel and the idea that “ontogeny recapitulates phylogeny” (originally applied to the metazoa). But in the case of the silicoflagellates the question is whether a phylogenetic or an ontogenetic model is more appropriate, as we seek to understand the relationship between the silicoflagellate organism in biology and its palaeontological representation in geology. Studies of the living organism have suggested (Van Valkenberg and Norris, 1970; Van Valkenberg, 1971a) that the skeletal morphology of living silicoflagellates is highly plastic. Moreover the skeleton is not even an essential part of the silicoflagellate organism, but rather occurs during just one phase of the life cycle, associated with motile cells in populations reproducing by asexual division. If the supposedly diagnostic criteria for silicoflagellate taxa are ephemeral results of water temperature and nutrient status in the ocean, as some suspect, it follows that the classification of the silicoflagellates is an artificial classification, only befitting morphotaxa like acritarchs and coccoliths, rather than biological taxa as the silicoflagellates have been assumed to be.

Thus we need to consider whether the taxa of silicoflagellates should be considered as morphotaxa, as now defined in the ICBN Art. 1.2: “[. . .] a morphotaxon is defined as a fossil taxon, which, for nomenclatural purposes, comprises only the parts, life-history stages or preservational states represented by the corresponding nomenclatural type.” (ICBN St. Louis Art. 1.2). This is the position I adopt here. The living organisms described by Ehrenberg, Schiller, Moestrup, and Foreman and by Van Valkenberg and others are not representatives of morphotaxa (which by definition are fossils) and cannot be named as such; in effect they are non-descript neotaxa, with the sole exception of Schiller's Octactis pulchra. Whether the three nominal species of living silicoflagellates represent one, two or three biological species remains unsolved and it is not proposed to formally name and describe them here. For palaeontological purposes, since the ICBN permits the retention of the name Dictyocha for the morphotaxa — irrespective of the fact that non-fossil representatives of Dictyocha must bear a different generic name — the long established name Dictyocha can be retained. It does not have to be replaced by the name Octactis. Since the morphotaxa of silicoflagellates are taxonomically different from the neotaxa of silicoflagellates they need not necessarily be governed by the same classification, even though they must be governed by the same nomenclatural Code.

As has already been observed, variation in the geometry and architecture of the skeleton has been interpreted as evidence of evolutionary or phylogenetic development. This has produced (e.g., in the hands of Bukry and Monechi, 1985, 1987, and their co-workers, see bibliography) a special purpose phenetic classification in which silicoflagellates are portrayed as rapidly evolving organisms of considerable stratigraphic importance. This approach is exemplified in the publications of Bukry and Monechi, Locker and Martini and others associated with the Deep Sea Drilling Project (DSDP) and the Ocean Drilling Program (ODP). A more conservative approach has been taken by McCartney in the reports of the Ocean Drilling Program (1993, 1994, 1995, 1998). A mathematical approach to the descriptive taxonomy of the group has been attempted by McCartney and Loper (1989 p. 296, 1992 p. 92) but these workers have admitted that the interpretation of the morphologies generated remains problematical. Some of the morphologies of today are well represented in the fossil record throughout the Tertiary and beyond into the Upper Cretaceous. As McCartney says, the sparse evidence for silicoflagellate evolutionary history consists principally of serial extinctions and a gradual diminution of taxa and morphological diversity since their appearence in the Lower Cretaceous. In their discussion of ‘fibulid / fibuloid’ 3 and ‘asperid / asperoid’ morphologies of Dictyocha fibula, for example, McCartney and Harwood, (1992 p. 821), state:

A broad species concept is used here, as with Dictyocha aspera. Asperid and fibulid morphologies tended to be very similar and to occur together, indicating perhaps a closer biological relationship than the taxonomic separation would suggest. Locker and Martini (1986) have determined, from examination of Ehrenberg's original type specimens, that the term Dictyocha fibula may best be applied to large specimens with an asperoid bar. Though it is difficult to argue against conclusions based on type material, the general practice through nearly all of the silicoflagellate literature of this century has been to apply the [epithet] fibula to Dictyocha that have a ‘fibuloid’ bridge, that is, one that is parallel to the major axis of the basal ring. Silicoflagellates are extremely variable and there is uncertainty, at least by the authors of this paper, about whether fibuloid and asperoid morphologies represent separate and distinct species.

Elsewhere in the same paper (1992 Plate 1: figs. 9–11) they show mesocenid skeletons (simple ‘basal’ rings without cross-struts or secondary ‘apical rings’) and corbisemid skeletons (three-sided ‘basal’ rings with a trigonal apical element) which are clearly morphs of the same population assigned to different nominal genera. They acknowledge the improbability of the conventional classification used by geologists: “We acknowledge the problems associated with silicoflagellate taxonomy based on simple morphological characteristics that may be influenced by environmental controls. Many silicoflagellate morphologies, presently treated as distinct, may be the same taxon . . . Taxonomy is an art and we are all painters. To avoid complicating this situation by offering new names and combinations, in this paper we adhere to existing concepts of taxa and favour broad morphological groups.” (McCartney and Harwood, 1992 p. 819). McCartney has also adopted a conservative stance on nomenclature in his comments in the section on silicoflagellates (cowritten with Lipps) in the chapter ‘Chrysophytes’ (pp. 143–154) in Lipps's Fossil prokaryotes and protists (1993), where he writes: “Silicoflagellate skeletons exhibit remarkable variety that has spawned a taxonomy, particularly in palaeontologic usage, that may not indicate actual biologic relationships.” (p. 145) and (p. 150) “Exactly how many silicoflagellate species exist in the Recent or at times in the geologic past is a perplexing question. Most biologists believe that only two silicoflagellate species live today and that the four-sided dictyochids and six sided distephanids are conspecific.”

The third species, Octactis pulchra, (which is normally eight-sided and is identical to the morphotaxon Dictyocha polyactis of Ehrenberg) is not mentioned by McCartney except in his last paragraph: “In Recent waters six- and seven-sided distephanids predominate at high latitudes, while four-sided dictyochids with inclined bars predominate at lower latitudes. Middle latitudes usually contain a mixture of the two groups. The D. speculum populations move toward the equator in autumn. An unusual seven- to nine-sided form previously known as Octactis but now considered synonymous with D. speculum, is especially abundant in [R]ecent nearshore environments.” (p. 153). McCartney's scenario therefore implicitly recognises only two species in one genus, with Octactis pulchra relegated to synonymy with D. speculum.

Names and types of names for silicoflagellate taxa

A survey of the described fossil silicoflagellates by A.R. Loeblich III et al. (1968) documented the names validly (and invalidly) published for the group up to that time. Subsequently a critical review of the generic concepts was made by Lipps (1970), who incorporated available biological information to provide a better characterisation of the body plan and identity of true silicoflagellates. This also indicated that a number of supposed silicoflagellate genera should be treated as fragments of diatoms or radiolaria. More recently Lipps and McCartney, in the former's valuable textbook Fossil prokaryotes and protists (1993), have provided a cautious overview from a palaeontological perspective.

None of these studies, however, completely resolved the nomenclatural problems of the group at any level. Neither paid much attention to categories above the rank of genus. Loeblich and Loeblich eschewed making taxonomic judgements for taxa at any level. Lipps (1970) gave attention only to the genera. Disregard or ignorance of the legal requirements of the ever-imperfect Codes has been widespread, and even the efforts of Loeblich et al. (1968) to set the nomenclature on a sound legal footing were not successful. Sigurd Locker (1974) published the results of a critical re-evaluation of the original silicoflagellate material in Ehrenberg's collection, but the results of this work have not been generally accepted because they break with the established taxonomy to which the few specialists working with silicoflagellates have become attached. In particular Locker's lectotypification of the name Dictyocha fibula has been problematic for palaeontologists with a narrow specific concept (see above).

Locker's work of 1974 — while valuable for his re-illustration of poorly figured Ehrenberg types and his assessment of infraspecific and specific synonymy — is of doubtful worth as far as generic taxonomy is concerned. Locker classified his taxa from fossil material without regard to living material and applied the ICZN rather than the ICBN to determine the status of names and typifications of names, despite accepting that silicoflagellates were algae (i.e., plants). He even formed names for two new families in botanical rather than zoological orthographies. Later, in 1985, he adopted the ICBN explicitly but even then used a superseded edition (the Leningrad 1978 edition), rather than the then current Sydney 1983 edition. In 1987 he was still using the superseded Leningrad Code. Locker's work, like that of some others, continually confuses taxa with names of taxa, and thus his nomenclatural assertions are problematic. Hsin Yi Ling (1981) showed similar error and confusion, citing the even more obsolete 1966 Code (Edinburgh, 1964). Care is also required in dealing with the nomenclatural assertions of Dumitrica (1973), Bukry and Monechi (1985, 1987). Bukry and Monechi acknowledged (in 1987) that they, also, were using the 1966 edition of the ICBN, then three editions out of date.

Another sort of confusion is expressed by William Siesser in his account of ‘calcareous nannofossils’ in Fossil prokaryotes and protists (1993). Admittedly, calcareous nannofossils are not fossil silicoflagellates, but the same code applies to both, and they are both studied by the same specialists with comparable nomenclatural difficulties. Siesser writes “Calcareous nannoplankton are best classified under the International Code of Botanical Nomenclature rather than under the International Code of Zoological Nomenclature because the botanical code provides for two taxonomic concepts — the organ genus and the form genus. The organ genus recognises taxa that can be assigned to a family, whereas a form genus cannot be assigned to a family because its biologic relationships are unknown.” Taxa — whether morphotaxa (form-taxa) or neotaxa — are classifiable, but not by a Code; they are merely named in accordance with the rules given in those Codes.

The ICBN of 1966 (the Edinburgh Code) did retain a distinction between organ taxa and form taxa, and that distinction was retained in the Seattle Code of 1972 — this being the edition used by W.W. Hay (1977), in his classification of ‘Calcareous nannofossils’. But that distinction between organ-taxa and form-taxa was removed in the Leningrad Code of 1978, when organ genera were abolished. In the St. Louis Code of 2000 the distinction was was completely removed when the rules permitting the use of morphotaxa were improved.

The status of the names of silicoflagellate morphotaxa

The way in which the silicoflagellates are handled as biological taxa and the way they are handled as distinctive fossil taxa (formerly ‘form taxa’) had long been unacceptable. As the preface to the St. Louis Code states (page xi):

Fossil plant nomenclature underwent profound changes at St. Louis. It is hardly exaggerated to say that the accepted compromise solution, that an ad hoc group of specialists worked out while the Section met, provides for the first time ever a sound, workable formal basis for past and current practice in palaeobotanical nomenclature. For nomenclatural purposes, botanical fossils are now considered to belong in the first place to morphotaxa: taxa at definite ranks that comprise only particular parts, life stages, or preservational states but not the whole organism (Art. 1.2). Formal synonymy, and the operation of the priority principle are confines within the framework and boundaries of morphotaxa (Art. 11.7). The qualitative definition of morphotaxon categories is not regulated by the Code but is wisely left for the practicing palaeotaxonomist to decide. [. . .] As strict synonymy and therefore priority only operates among morphotaxa of the same kind, names of botanical fossils cannot logically compete with names based on a non-fossil type. By consequence the former Art. 11.7, which ruled that names of non-algal non-fossil taxa take priority over names of fossils, has been downgraded to the status of a Note (Art. 11 Note 4).

The nomenclature section of the St. Louis Congress altered the text in Art. 11.7 to replace the word “algae” with the word “diatoms”. A proposal was accepted on the motion of Skog “on behalf of an ad hoc group of palaeobotanists in attendance” (report by F.R. Barrie and W. Greuter, 1999: p. 783 notes 8 and 9), with the result that the Article, as finally modified, now reads:

11.7 For purposes of priority, names of fossil taxa (diatoms excepted) compete only with names based on a fossil type representing the same part, life history stage, or preservational state (see Art. 1.2).
[two examples, for bark fragments and for fossil cysts, follow]
Note 4. Names of plants (diatoms excepted) based on a non-fossil type are treated as having priority over names of the same rank based on a fossil (or subfossil) type.

Since no special exemption was given for silicoflagellates, names of silicoflagellate fossils no longer compete with names of living silicoflagellate organisms. Applying Note 4 gives Octactis (which was based on a non-fossil type) priority over Dictyocha Ehrenberg, 1837 (based on a fossil type) should Octactis and Dictyocha be considered congeneric. Because one of these names is the name of a morphotaxon while the other is the name of a neotaxon, in law they do not compete. All non-Recent silicoflagellates should now be considered as morphotaxa, establishing their taxonomy on the avowedly artificial but empirically demonstrable evidence of skeletal structures alone. This leaves about twenty nominal genera, and at least three dubious nominal families to be classified and named as palaeontological morphotaxa.

A later section of this review attempts to establish a legal basis for the generic nomenclature of all silicoflagellates, living (one genus) and wholly fossil (up to about 25 genera of morphotaxa), in accordance with the current ICBN. No new generic names are introduced and no alterations to specific and infraspecific names are made either. Generic and suprageneric concepts and the names for these constructs by earlier writers are reviewed and the taxonomic history of those concepts is set out with a provisional assessment of their taxonomic utility, drawing on the enhancements made possible by SEM studies. Further fiddling with specific and with infraspecific taxonomy of morphotaxa and their nomenclature has been set aside, except where necessary for legal rather than taxonomic reasons. Upon this foundation it may be practicable, in future, to make taxonomic assessments of the species of silicoflagellates and their nomenclature.

Biregnality and the legal history of the silicoflagellates

While silicoflagellates exist in the world of biology, they exist in two other worlds too. One of these is the legal world of the Codes and the other is the palaeontological world, in the literature of which silicoflagellates are most often seen. Most of the geologists, palaeozoologists and palaeobotanists who have studied the silicoflagellates appear to have been little acquainted with either of the Codes, let alone their various editions. This has amplified nomenclatural difficulties.

Within the world of historical biology, however, there is a further legal division. In the original International Rules of Zoological nomenclature of 1926 (International Commission on Zoological Nomenclature, (1926) p. 75) the first Article stated: “Zoological nomenclature is independent of botanical nomenclature in the sense that the name of an animal is not to be rejected simply because it is identical with the name of a plant. If, however an organism is transferred from the vegetable to the animal kingdom its botanical names are to be accepted in zoological nomenclature with their original botanical status; and if an organism is transferred from the animal to the vegetable kingdom its names retain their zoological status.”

The classification of the group in the post-biregnal era, however, can be expressed by nomenclature governed either by the ICBN or the ICZN. In common with most who write about the silicoflagellates I have chosen to follow the ICBN. The dispute over whether silicoflagellates are plants or animals may be solved in a tidy way by designating them as ‘zoonominate protists’ or alternatively ‘phytonominate protists’ — in the sense that they are things named as if they were animals or plants, respectively. Coccolithophores and dinoflagellates are also phytonominate protists but phaeodarians (which are touched on in a later section) are zoonominate protists. Orchard et al. (1996 p. 289) wisely observe that “Despite increasing evidence that the organisms traditionally treated as ‘fungi’ and ‘algae’ probably constitute several kingdoms, and that some at least of these are more closely allied to the traditional animals that to the plants, the inclusion of provisions for their nomenclature within the botanical Code is a pragmatic solution that, on the whole works well. The Codes should not be viewed as a definitive statement on phylogeny. They are simply a set of rules to govern the way names are manipulated.”

Fensome, Jansonius, and Skog (as specialists in the dinoflagellates governed by the ICBN) criticised the difficulty in applying the often “arcane technicalities” of the ICBN (Fensome et al., 1998 p. 695) and questioned whether the then current upheaval in a quest for a new “harmonised” nomenclature was worth the trouble. The International Botanical Congress at Missouri decided that a suggested “harmonisation” was not desirable or practical and decided to simplify the rules in a manner which has attracted much favour. A so-called ‘Phylocode’ based on cladistic principles (but, confusingly, using names of taxa as if they are names of clades!) has been promoted among cladists but has been criticised as conceptually muddled (see e.g., Stuessy (2000) the reply by K. de Queiroz (2000) and the short cogent reply by P. Jørgensen (2000). It need not concern us here.

Stratigraphy and Taxonomy

It is far outside the scope and intentions of this paper to enter into a discussion of the relationships between silicoflagellate taxonomy and silicoflagellate stratigraphy. It is necessary, however, to comment briefly on the stratigraphic nomenclature and taxonomy in relation to the species concept used by geologists. Locker and Martini (1985 pp. 895–901) discuss the questions in depth, commencing: “Taxonomy is a special field of research in neontology and palaeontology that aims to establish taxa both of plant and animal organisms that are clearly defined and stable in nomenclature, and that are generally in accordance with biological and phylogenetic data. Some of the problems still present in silicoflagellate stratigraphy are influenced by taxonomic ideas, especially by the species concepts and synonymies hitherto not noted.”

They go on to outline such a new classification remarking that “The species concept employed to determine silicoflagellate taxa of specific or infraspecific rank may have great importance in applied stratigraphy. If the population polymorphism of skeletons and the resulting interspecific and intergeneric homeomorphisms (present especially in the genera Dictyocha, Distephanus, and Connopilus [sic]) are not taken into account, subdividing and correlating geological sections will be difficult.” They remark that “Homologous skeletal morphes [sic] occurring in some species of different genera may lead to stratigraphic misrepresentations if they are given the same name.” This is to say that taxa of identical form — but which are separated solely by stratigraphic criteria — may be taxonomically acceptable. It also implies that taxa of identical appearance may bear different names. Both assertions are objectionable.

Locker and Martini acknowledge that “intraspecific variation may be very great under special conditions” although they assert “normally it is limited” and that taxa may be discriminated on the basis of “statistical and biological considerations (see the distinction between Mesocena elliptica, tetragonal forms of M. diodon, and M. quadrangula, this paper.” They assert, moreover, that “in some cases synonyms and homonyms can mask conformities between stratigraphic successions of species in different areas. . . . It is to be hoped that confusion about the stratigraphic ranges of species caused by synonyms will soon be overcome by fixing the correct names, that is the oldest legitimate names.” Fifteen years later this ‘confusion’ still seems no closer to resolution, probably because it is based on a interpretation of ontogenesis as phylogenesis, after the manner of Haeckel.

In their 1985 paper Locker and Martini acknowledge that “because they are plants, the ICBN rules must be applied to discriminate and define taxa” but it is not the function of the ICBN to “discriminate and define taxa”; this is more often an error associated with cladists (see Stuessy, de Queiroz, and Jorgensen references in the bibliography). 4 The ICBN can only be applied to determine the legal status of a name; it cannot circumscribe the application of that name according to particular taxonomic beliefs. In the ICBN certain authorised infraspecific ranks of taxa are stipulated — these are subspecies, varietas, subvarietas, forma and subforma. Locker and Martini's species concept used the ranks of subspecies, variety and form only. It was, they say, “based on statistical investigations made to clarify the taxonomic position of the species introduced by Ehrenberg (Locker, unpublished data) and on published empirical data” but they say that the difficulties in the taxonomy “can be solved only if species are treated exclusively as biological entities . . . rather than as typological entities”.

This species concept rested on the assumption that silicoflagellates “possess temporary phases of sexual reproduction. Such phases have not yet been demonstrated but may be deduced from the relatively high stability of fossil and recent species.” The hypothesis of sexual reproduction (attractive as it might seem) still lacks proof fifteen years later, despite Locker and Martini's statement that: “The persistent homogeneity of silicoflagellate species, a homogeneity often threatened by strong and divergent variation, can only be understood if sexual processes are presumed.” There is a strong indication here of a priori assumptions. 5

Noting Ernst Mayr and Theodosius Dobzhansky's criteria for species discrimination in sexually reproducing animals, Locker and Martini, wished to apply some of these criteria to the case of silicoflagellates as well. A table presented by Locker and Martini (Table 4, page 900) defines the ranks selected both biologically and typologically. The application of Mayrian concepts of sympatry and allopatry — concepts assuredly difficult to apply in plankton! — and the discrimination of clones ranked as varietas and forma, mark the infraspecific taxonomy as typological. The authors write:

One of the most important categories of the species concept used here is the form. Forms are regarded as intrapopulation variants which can be fixed by features which vary discontinuously. . . . In terms of genetics these forms must be quoted as morphes. The concept of intrapopulation polymorphism agrees well with the empirical data . . . and also with statistical evaluation . . . Nevertheless, in some cases it may be difficult to apply the concept because similar morphes may be developed in different species of the same genus (homologous morphes). Making distinctions will be further complicated by populations containing, in addition to the normal variants, morphes which correspond in habitus to forms of other genera (analogous morphes).

This results in co-called ‘dictyochoid forms’ of Corbisema, ‘distephanoid and cannopiloid forms’ of Dictyocha and ‘dictyochoid and cannopiloid forms’ of “Distephanus” and ‘naviculoid and corbisemiform morphes’ of, for instance, Corbisema disymmetrica — as illustrated in their later paper (Locker and Martini, 1987 pp. 32–33). According to these authors: “The species concept proposed here is intended to establish precise criteria for distinguishing intraspecific taxa, criteria that are biologically based and the may be reproduced in fossil populations.” The applicability of such concepts to fossil silicoflagellates is unconvincing. Their classification has not been widely adopted by others.

Conclusion to Part I: A reality check on silicoflagellate classification

Van Valkenberg and Norris (1970 pp. 53–54) are surely correct in casting doubt on the recent palaeontological taxonomy of silicoflagellate morphotaxa:
the 200 skeletons from 1 clone which we examined would, according to Gemeinhardt, belong to more than 1 genus. The systematics of Deflandre would assign them to more than 1 species. Frenguelli's work would assign them to several forma. The practice of assigning ever lesser rank to differences in skeletal form as the plasticity of the skeleton is increasingly recognised seems to us to be equally incorrect and to evade the basic issue. Skeletal characters alone either are or are not sufficiently stable to be used as the taxonomic base. The fact that differences in such characters as number of spines and presence or absence of apical windows can occur among the progeny of a single individual in a growing period of only 21 days would suggest that perhaps another basis for the classification of living silicoflagellates needs to be devised.

Since the ‘silicoflagellate problem’ (as McCartney calls it) has remained insoluble and the morphotaxonomic classifications appear to have failed, I have looked at the historical origins of these classifications to explore a different kind of solution to ‘the silicoflagellate problem’. The errors made in the past have left a complex and exasperating legacy for people trying to deal with the living and the fossil silicoflagellates in the future. This approach is described below.

Part II: The classical taxonomy of silicoflagellates

First appearences

The first scientific reports of silicoflagellates were made by Ehrenberg (1837 p. 61, and 1838 p. 118) when he noted the presence of two species in ‘Polierschiefer’ (polishing shale), a diatomaceous earth from Oran in Algeria. The first of these papers (“Über eine briefliche Nachricht des Hrn Agassiz in Neuchatel über den ebenfalls aus mikroskopischen Kiesel-Organismen gebildeten Polierschiefer von Oran in Africa”, in Berichte K. Akad. Wiss. Berlin aus dem Jahr 1837, pp. 59–61) was presented to the academy on 13 April 1837. The second paper (“Über das Massenverhältniss in jetzt lebenden . . .” from Abhandlungen K. Akad. Wiss. Berlin 1838, pp. 109–135) was issued in 1838, but Loeblich et al. (1968) do not list this paper. The generic name was accompanied by a brief description of the genus, but the two specific names were without descriptions. In the Abhandlungen paper the name Dictyocha appears on page 118 in a statement that the genus contains three fossil species but no living forms. On a folded table inserted after page 132 two species are listed (as nomina nuda) from Oran, Algeria, and a third (also a nomen nudum) is recorded from Zante Island (Ionian Islands, Greece) — this last was named in 1839 as D. navicula. The Abhandlungen article was presented in two sessions (20 July and 3 August 1837) and printed with supplements in December 1837. 6

The next pertinent publication is Ehrenberg's Infusionsthierschen (1838), of which the Vorrede is dated July 1838, and in which, as a “Nachtrag zur Section der Desmidiaceen” Ehrenberg gives a somewhat expanded description of the genus and lists the three included species as nomina nuda (page 165). Finally, in the Abhandlung for 1838, read at a meeting held on 20 December 1838, Ehrenberg formally described the genus (pp. 128–129) and six species (p. 129) including the three mentioned above. The text of the Abhandlungen paper was first published as a separate under the title Die Bildung der europäischen, libyschen und arabischen Kreidefelsen und des Kreidemergels aus mikroskopischen Organismen, dargestellt und physiologisch erläutert (Berlin, 91 pp., 3 tabs., 4 pl.) in 1839. 7

The two species first described were subsequently observed in other deposits, including the famous diatomaceous ‘Bermuda Earth’ (a deposit not in fact from Bermuda, but from near Nottingham in Maryland). According to Loeblich et al. (1968 p. 16) “This rich material was first obtained from a mineralogical correspondent, Mr Tuomey of Petersburg, Virginia and was sent, in turn to Prof. J.W. Bailey of West Point. Bailey examined the material, made sketches of many of the species and sent these with part of the material to Ehrenberg, who described from it some nine new genera and 58 new species of diatoms and silicoflagellates. Some of these were also described and figured by Bailey himself (1843, 1845).” The Ehrenberg publications prior to 1844 and 1845, however, do not seem to draw on this Maryland material and so the earliest published accounts of silicoflagellates are based on material from Oran, Caltanisetta in Sicily and Aegina in Greece.

Initially Ehrenberg recognised only two nominal species of silicoflagellates. His remarks in the Berichte paper (1837) are very brief (they are translated rather literally): “Very striking are two recently seen perforated and scaffold-like shapes one of which comes close to the [living] Arthrodesmus truncatus 8 but which are quite distinct because of the netlike siliceous skeleton. These may be called Dictyocha speculum and D. fibula.” Later in the paper he mentions a third species “welche eine Navicula ähnlich ist” (which resembles a Navicula) observed in ‘Polierschiefer’ from Greece.

It was Ehrenberg's practice to introduce novelties in brief presentations which were published in the Berichte with German descriptions (often minimal), anticipating a formal treatment in more substantial papers in the Abhandlungen, in which formal Latin descriptions or diagnoses were given. The 1837 Berichte paper was (as usual) written in the third person by the Secretary of the academy and does not even have a caption title, and the volume in which it is contained lacks a table of contents or an index. The brief first comments in the Berichte have been taken by Loeblich et al. (1968) and by others as sufficient a description to legally validate the generic name Dictyocha (although not to validate the species names mentioned as nomina nuda). They scarcely constitute a ‘description’ in terms of ICBN Art 32.2 (“A diagnosis of a taxon is a statement of that which, in the opinion of its author distinguishes the taxon from others”) but the ICBN Art 32 is lax in its toleration of descriptions and diagnoses, permitting almost anything. Art. 34 is less tolerant, when it states that a name is not validly published if it is “merely proposed in anticipation of the future acceptance of the group concerned” but the rule is aimed at names proposed for hypothetical taxa, rather than taxa which are foreshadowed. It might be argued that, rather than being intended as a formal introduction of a name, the Berichte paper merely makes an incidental mention of a provisional group not yet definitely accepted and diagnosed (see ICBN Art. 34.1 (b)). This was the position taken by Locker (1974) in treating the generic name as dating from 1840 rather than 1837. The matter of intent is important. It is not sufficient that nomenclatural acts take place as a result of omission. Omissions certainly can invalidate nomenclatural acts (e.g., when an author omits to cite a basionym for a new combination) however the ICBN, in its present form, allows the expression “Very striking . . . perforated and scaffold-like shapes . . . which are quite distinct because of the netlike siliceous skeleton” as an adequate description, for better or worse.

In a further account of the silicoflagellates, in 1839 “Über die Bildung der Kreidefelsen und des Kreidemergels durch unsichtbare Organismen” in the Abhandlungen K. Akad. Wiss. Berlin for 1838, (published in 1839 as a separatum, although the volume appeared in 1840), Ehrenberg heads his account “Dictyocha Novum Genus. Netzschiebe” thereby indicating that he was here formally establishing the genus (and publishing the name with a Latin diagnosis) as proof of formal publication. In this account he includes no less than six nominal species, for all of which Latin names and diagnoses are provided. 9

The genus as circumscribed in this 1839 paper was still characterised entirely on the basis of fossil material. The six species are D. fibula and D. speculum (both seen first in the Oran ‘Polierschiefer’ sample; D. navicula (originally seen in the ‘Polierschiefer’ from Greece) and three new nominal species D. polyactis, D. ? stella, and D. triangula which are all described from a deposit found at Caltanisetta in Sicily. All of these names were lectotypified by Locker (1974), after study of extant material in Ehrenberg's own collection.

The so-called “D.? stella” is not in fact a silicoflagellate, but rather the internal siliceous skeleton of a dinoflagellate (Actiniscus) and was separated out from the silicoflagellates sensu stricto at subgeneric level (Ehrenberg, 1840a) and later at generic level (Ehrenberg, 1843a). It is not to be confused with Macrora Hanna, 1932 (Type: M. stella (Azpeitia) Hanna, 1932, = Pyxidicula ? stella Azpeitia, 1911) which is a synuracean with the same epithet but nomenclaturally unconnected.

D. triangula is a silicoflagellate but was separated subgenerically as Dictyocha subg. Mesocena Ehrenberg, 1840 (Ehrenberg, 1840b) and generically as Mesocena Ehrenberg, 1843 (Ehrenberg, 1843b). Locker (1974), however removed it to his new genus Bachmannocena. It was known from a single specimen. The nature of this taxon is further discussed below, as the nomenclature is complex.

D. polyactis is also a silicoflagellate and was re-illustrated by Locker (1974 Tafel IV, fig. 2, from the unique “holotype”) under the name Distephanus polyactis. It was also promoted to generic rank in due course as Paradictyocha Frenguelli, 1940. As will be shown below, however, this specimen is in fact a fossil specimen of Octactis pulchra Schiller, 1925.

D. navicula was based on fossil material from Greece (Upper Miocene of Placa di Furni, Zante Isl.). The description “D. navicula, cellulis octo [sic i.e., duo], rete oblongum cylindricum, obtusatum, septo medio. Naviculae simile formantibus.” (D. navicula, with eight [i.e., two] cellules, network oblong, cylindrical, obtuse with a septum in the middle. Resembles Navicula in shape.) indicates a typical naviculiform silicoflagellate of the morphology subsequently separated as the genus Naviculopsis Frenguelli, 1940, to which it was eventually transferred by Deflandre in 1950. The type from Zante is illustrated by Locker (1974 Tafel II, fig. 1). Naviculiform silicoflagellates are confined, apparently, to the geological period Palaeocene to Oligocene, although they have been reported from the Miocene (Loeblich et al., (1968) p. 19), and the type of D. navicula is said to be of Upper Miocene age.

This leaves the two original nominal species, both fossils, upon which the original generic concept of Dictyocha was principally based: D. fibula and D. speculum. They represent four-sided and six-sided specimens respectively.

D. fibula “cellulis quaternis inaequalibus, planis, totidem apiculis armatis” (D. fibula, with four cellules of unequal size, flat, with a spine at each apex) was named for the fancied resemblance to an antique brooch, (the fibula) with its pin backing the face of the ornament (the abapical surface). Although the account is based on the fossil material (Locker, 1974 Tafel I, fig. 6 illustrating the type specimen), Ehrenberg saw living material, which he assigned to D. fibula, the following year. In later works, however, Ehrenberg separated several fibuliform silicoflagellates as distinct species, notably D. crux Ehrenberg, 1841 and D. epiodon Ehrenberg, 1845 (see the same work, Tafel III, fig. 6, Tafel I, fig. 4 respectively).

In describing Dictyocha varia Locker said that he had observed more than a hundred skeletons assigned to the species in the Ehrenberg collection from Caltanisetta and he distinguished the species on a statistical evaluation, comparing it with what he called Dictyocha epiodon (both being segregates from Dictyocha fibula sensu lato). Within the species three nominal forms were distinguished on the basis of trivial differences. Locker (1975 p. 101) speaking of the species acknowledged that the taxonomy and nomenclature were in a chaotic state (“zur Zeit in einem chaotischen Zustand befinden”) and it was necessary to explicate the identity of the most important species. Whether these trivial differences merit taxonomic discrimination at any level, much less at specific rank remains an open question.

D. speculum, the other species, was so called from the resemblance to the shape of an ancient hand mirror, the description being “cellulis senis, calathi formam, spinulis inaequalibus senis radiatam referentibus. Primo intuiti speculi antiqui formam addiceres.” (with six cellules, and shaped like a calathus [i.e., a sewing basket or shallow wine cup], with six radiating spines. At first impression you'd think it was an ancient mirror.) The type from Oran and another specimen from the same collection are illustrated by Locker (1974 Tafel III, fig. 2 (lectotype) and Tafel III, fig. 1 (heptagonal variant, subsequently called Dictyocha septenaria in 1845).

Again, this initial description is based on fossil material. Soon afterwards. However, Ehrenberg observed living cells attributed to D. speculum from the Baltic Sea near Kiel, the North Sea (Oslo Fjord) and elsewhere. In his account of these (Ehrenberg, 1840a) he adds to the diagnosis the point that all the ‘cellules’ are of the same size (which distinguishes it from D. fibula) and says (rather literally): “Living silica-animalcules (‘Kieselthierchen’) very different from the chalk-forms [i.e., coccolithophores] were first observed near Kiel on 15 September 1839, which from their appearance recalled an antique mirror. They differ in the relative length of the girders (‘Stacheln’) in the size and strength of the cell wall [i.e., the basal ring], as well as in the above character.” He goes on to describe the colour, internal organelles and movement (already quoted above).

Fibuliform and speculiform silicoflagellates, although described from fossil material, are well known as living taxa, and over time have come to be regarded as generically distinct, conventionally separated as “Dictyocha” and “Distephanus”, respectively. As pointed out above, however, neither of these names is applicable to living taxa (the only available name for these is Octactis). If all fibuliform and speculiform silicoflagellates morphotaxa should be regarded as belonging to a single genus the name Dictyocha (Type: Dictyocha fibula Ehrenberg, 1839) should be used.

As is clear from the original publications, Dictyocha fibula itself was described from middle Miocene fossil material from Oran, Locker citing the type preparation as “Lectotypus, Prap. ‘Oran 3c w’” in the Ehrenberg collection, figured by him (Tafel I, fig. 6). Two trivial variants from the same Oran collection were later described as Dictyocha tripyla and Dictyocha pons (both in 1845) and all three are shown in the same plate. The fossil character of the type of this name and of the name Dictyocha and of Dictyochaceae, Dictyochales and Dictyochophyceae makes all of these name “nomina palaeomorphosium” or names of fossil morphotaxa. However Dictyocha fibula was applied to living material within months of its valid publication as a name of a fossil and has been so applied ever since.

This creates an unexpected problem, as any name for a silicoflagellate at the rank of family, order or class and based on a non-fossil type, must be preferred to a name at such rank based on a fossil type. Dictyochaceae, Dictyochales and Dictyochophyceae (all established names for morphotaxa) must be replaced in non-fossil phycological nomenclature by names based on those of living silicoflagellates hitherto not published ( ‘Octactinaceae’, ‘Octactinales’ and ‘Octactinophyceae’) as a consequence of Art. 11.7, Note 4. Octactis is not a ‘fossil taxon’ and thus the conditions for the validation of botanical names of neotaxa apply. 10 Publication must occur elsewhere.

1. “Distephanus”: a taxonomic mistake

Stöhr (1880) distinguished a separate genus of silicoflagellate under the name Distephanus saying (my translation): “In certain examples of Dictyocha speculum a little siliceous ‘hat’ lies against the base, replicating it. It develops in this manner into a separate complete shape and marks the [evolutionary] progress to the Sphaeridia [i.e., the sphaerellarian polycystine radiolaria].” This new genus was based on a single species “Distephanus rotundus Stöhr, 1880” itself based on a single specimen (“Sehr selten, nur ein Exemplar.” (Very rare, only one example) from the Miocene of Sicily. The name used, Distephanus (two crowns), aptly described the appearance.

This is actually a double specimen of Dictyocha speculum in the process of replicating itself, as was soon pointed out by Haeckel (1887): “The genus Distephanus was founded in 1880 by Stöhr (loc. cit.) for a single twin-piece of the skeleton of Dictyocha speculum. . . . No doubt this was a mistake and the apparent little sphere was one of the above mentioned twin forms composed of two separate hexagonal truncated pyramids which were loosely connected by their basal rings. I have often seen such twin pieces of Dictyocha speculum and of other species (Pl. 101, fig. 12, Pl. 114, fig. 8) and was always able to separate the two loosely connected halves of the bivalve shell by slight compression. The genus Distephanus of Stöhr, therefore, is nothing other than the Dictyocha of Ehrenberg.” Haeckel was right.

In botanical nomenclature the name Distephanus is preoccupied homonymously by Distephanus Cassini, 1817 (a junior synonym of the name of a daisy, Vernonia Schreber, 1791). In 1980 Silva considered what to do over the question and there remarked: “While there is skepticism among those who study living organisms as to the biological validity of Distephanus (e.g., Van Valkenberg and Norris, 1970), the genus is firmly entrenched in present day palaeontological nomenclature. It seems advisable, therefore, to consider conservation of this name, and I shall publish a formal proposal elsewhere. If Distephanus is conserved Distephanaceae will become available.” (Silva, 1980 p. 56). No such proposal was ever submitted and I would not support it if it were. Although Distephanus has become widely established in the literature, and could be proposed for conservation on that basis, there is no point in trying to save the name so as to distinguish four-sided from six-sided specimens of what might as easily be considered the same “species”. It is more convenient to treat both four- and six sided skeletons as congeneric, using Dictyocha and distinguish the four- and six sided morphs by informal epithets. There is no merit in trying to save the illegitimate name Distephanaceae Locker, 1974.

2. Haeckel's misinterpretation of the silicoflagellates

Although his observation on the identity of “Distephanus rotundus”, was correct, otherwise Haeckel's treatment of the silicoflagellates rests on a major interpretational error, which is elucidated below. In the prodromus for his report on the Challenger expedition radiolarians (1882, but possibly issued as a separatum in November 1881) Haeckel lists genus “588. Dictyocha” without explanation in “Subfamilia Cannorhaphida” in “Familia Phaeocystida” in “Ordo Phoeodaria.” He is writing here of a concept more or less equivalent to the zoologically-named Class Phaeodaria (deep-sea radiolarians) of the present day. He gave a more detailed account of this group as he saw it under the name “Tripylea” in “Sitzungsberichte der Jenaischer Gesellschaft fur Medic. und Naturw. 1879” issued on 12 December of that year (Haeckel, 1879). His thinking about the group emerges most clearly from his comprehensive account in the Challenger report v. 18 (Haeckel, 1887).

In the Challenger report Haeckel classified the silicoflagellates as a subfamily “Dictyochida” of his family “Cannorrhaphida” of the Order “Phaeocystina”, having adopted the view of Richart Hertwig (1879) who had (erroneously) taken siliceous skeletons of Dictyocha for isolated spicules analogous to those found in the calymma of radiolarians. Haeckel's account of the history of the group and of his previous interpretations makes this quite clear: “The genus Dictyocha had already been founded by Ehrenberg in 1838 [sic] with the following definition: ‘Lorica simplex, univalvis, silicea, laxa reticulata aut stellata’ Abhandl. d. K. Akad. d. Wiss. Berlin, 1838 p. 128”. Haeckel said that Ehrenberg had found only scattered pieces of the skeleton fossil in Tertiary rocks. He placed them among the Bacillaria (= Diatomaceae), but added, that they may be possibly scattered spicules of sponges (“forsan Spongiarum ossicula”). Haeckel says:

In 1859 I myself observed similar forms of Dictyocha at Messina, and first recognised them as true Radiolaria. But I placed them at that time among the Acanthodesmida, beside Prismatium, supposing that a small spherical body which I had sometimes seen in the cavity of the pileated pieces (probably a phaeodellum) was the small central capsule (Monogr. d. Radiol. 1862 p. 271 Taf. xii figs. 3–6). The complete body of Dictyocha was not described until 1879 when R. Hertwig gave a full description of its peculiar structure, and especially of the great central capsule, which resembles that of the other Phaeodaria. He first stated that the singular pileated pieces described by Ehrenberg were not complete shells, but isolated pieces of the skeleton, which are scattered in the jelly envelope around the central capsule in a mode similar to the spicula of Thalassoplancta, Thallassosphaera, and Sphaerozoum [three genera of polycystine radiolaria]. Hertwig also first recognised that the thin rods which composed the reticular pileated pieces of the skeleton in Dictyocha are not solid bars but thin hollow tubules similar to the hollow rods of Aulacantha and of other phaeodaria. Numerous complete and well-preserved specimens of Dictyocha which I found in the collection of the Challenger, have convinced me that the accurate description of R. Hertwig is correct in every respect and that these remarkable bodies are true Phaeodaria, most closely related to Cannobelos (= Thalassoplancta) and to Aulacantha (compare Pl. 101, fig. 10). I now regard them as representatives of a peculiar subfamily of Cannorrhaphida, which I call Dictyochida. To the same subfamily also belong the small circular bodies which Ehrenberg described in 1841 as Mesocena (loc. cit. p. 401), and the elegant, more complicated reticular and pileated bodies which Stöhr figured in 1880 under the name Distephanus (Palaeontogr. vol. xxvi, p. 121).

This misinterpretation of the evidence — Haeckel had evidently forgotten or not noticed Ehrenberg's description of the living silicoflagellates he had seen in September 1839 — is expressed in the splendid illustration of “Dictyocha stapedia Haeckel, 1887” Pl. 101, fig. 10 of the Atlas to his work (also reproduced as Pl. 22, fig. 1 in Loeblich et al., (1968)). This is described as “A complete specimen, observed living at Ceylon. In the centre is visible the large spheroidal tripylean central capsule with its three openings, containing a large nucleus with numerous nucleoli. Its oval half is covered with the dark phaeodium . . .” The organism illustrated is in fact a specimen of the phaeodarian Phaeodina tripylea which has been accreting epiodontiform specimens of Dictyocha fibula, skeletons of which lie on its surface. Haeckel even says (1887 p. 1544) “In the specimen of Phaeodina tripylea which I observed living, it exhibited exactly the same shape as the figure of Dictyocha stapedia in Pl. 101, fig, 10. The only distinction in this latter being indicated by the pileated pieces of the skeleton on the surface.” Phaeodina is one of a group of so-called “naked phaeodarians” which accrete siliceous cases from any siliceous detritus which happens to be available, including spicules, frustules and silicoflagellate skeletons.

Hence Haeckel diagnoses “Dictyocha” as “Cannorrhaphida with a skeleton composed of annular pieces which are fenestrated by one or more arches, arising on one side of the simple ring but do not form a truncated pyramid (therefore no apical ring)” and distinguished it accordingly from his misappropriated application of the name Distephanus — “Cannorhaphida with a skeleton composed of pileated pieces, each of which is a small truncated pyramid, with a girdle of meshes (the apical ring being simple.)”

For Haeckel the Dictyocha organism is (conceptually speaking) a Phaeodina which has been accreting fibuliform silicoflagellates, while Distephanus is a Phaeodina which has been accreting speculiform silicoflagellates. Haeckel also records Mesocena in a similar way, referring to one specimen observed covered with mesocenid rings, here described as Mesocena stellata gathered from 2925 fathoms in the Central Pacific and rather implausibly figured by him. Another of his phaeodarian genera (Catinulus) appears to be yet another Phaeodina which has been accreting centric diatoms!

This might not matter very much, but, finally Haeckel describes a new genus of this alliance, Cannopilus, for “Cannorraphida with a skeleton composed of pileated pieces each of which is a small truncated pyramid with two girdles of meshes (the apical ring being fenestrated)”. Conceptually this is a further kind of phaeodarian accreting silicoflagellates. No such organism had actually been seen: its existence was inferred from the presence of isolated ‘pileated pieces’ which could not be assigned to a phaeodarian known from a ‘complete’ specimen. One of the five nominal species assigned to Cannopilus, his C. hemisphaericus (Ehrenberg) Haeckel, 1887 (= Dictyocha hemisphaerica Ehrenberg, 1844), was known to Haeckel only from Ehrenberg's account. The other four species, which he had actually seen, are merely overdeveloped or teratological fibuliform or speculiform skeletons of Dictyocha fibula. The result of the error is that we have inherited a clutch of silicoflagellate morphotaxa with names more appropriate for phaeodarians. These names persist: Van Valkenberg and Norris (1970) say that one skeleton from their cultured silicoflagellates corresponded to that of “Cannopilus calyptra Haeckel”.

Cannopilus has been treated as a junior synonym of Halicalyptra Ehrenberg, 1847 (LT: H. virginica Ehrenberg, 1854, lectotypification by Campbell, 1954), but this is problematical. In his original account of Halicalyptra Ehrenberg gave a generic diagnosis in tabular form but named no species and provided no illustrations. Subsequently (1854) he illustrated three species but gave no descriptions of them. Descriptions were eventually provided by Haeckel (1862 p. 289). The identity of these three nominal species has been obscure but two of them (virginica and depressa) were figured by Locker (1974) and shown to be silicoflagellates from a Miocene deposit in Richmond, Virginia, while the third (fimbriata) is actually a polycystine radiolarian from the Eocene of Barbados — this representing the material upon which the genus was originally based in 1847.

Halicalyptra has subsequently been treated as a radiolarian by both Haeckel (1887) and by Campbell (1954), but Campbell's ‘lectotypification’ of the name on H. virginica was a mechanical one following the strict prioritarianism of zoological practice governed by the ICZN. Campbell's lectotype, however, is actually a silicoflagellate (and so governed by the ICBN) although he considered Halicalyptra a radiolarian (governed by the ICZN). The holotype of Halicalyptra virginica, figured by Locker (1974 Tafel IV, fig. 6) is indeed a curious creature.

The reasonableness of the lectotypification of the generic name Halicalyptra on H. virginica is thus very questionable. The selection was rejected under the ICZN as unacceptable by Locker (1974) who pointed out that the species had not been unequivocally included in Halicalyptra by Ehrenberg. Locker lectotypified the name on the original figured radiolarian species H. fimbriata Ehrenberg, 1854 instead. This lectotypification is accepted here with the result that Halicalyptra is returned to the polycystine radiolaria. Locker regarded the other two species of Halicalyptra as representative of “Cannopilus” (i.e., isolated silicoflagellate skeletons) naming them as Cannopilus virginica (Ehrenberg) Locker, 1974 and as Cannopilus hemisphaericus (Ehrenberg) Haeckel, 1887.

But, contrary to the views of Locker (1974), the name Cannopilus — proposed as a zoological name for a hypothetical phaeodarian represented by skeletal material, mostly composed of large speculiform specimens of Dictyocha— is itself plainly mislectotypified on C. hemisphaericus. This was something which Haeckel had never actually seen and which he only included on the basis of a description of a ‘single pileated piece’. The specimen from Ehrenberg's collection which is the stated “Lectotype” of Dictyocha hemisphaerica Ehrenberg, 1844c was figured by Locker (1974 Tafel IV, fig. 1). Dictyocha hemisphaerica Ehrenberg, 1844c is the earliest name for a member of the so-called ‘cannopiloid group’ of Dictyocha and the lectotype (in “Prap. Bermuda 5 d w”) was one of the specimens isolated from ‘Bermuda earth’. Other specimens of Dictyocha hemisphaerica from various locations are figured by Locker in the same plate (figs. 4–5, 7–8) as Cannopilus hemisphaericus: all are speculiform silicoflagellates, not phaeodarians or parts of them.

ICBN 2000 (St. Louis Code) Art. 9.13 states: “The holotype (or lectotype) of a name of a species or infraspecific taxon of fossil plants (Art 8.5) is the specimen (or one of the specimens) on which the validating illustrations (Art. 38) are based. When, prior to 1 January 2001 (see Art. 38.2), in the protologue of a name of a new taxon of fossil plants of the rank of species or below, a type specimen is indicated (Art. 37.1) but not identified among the validating illustrations, a lectotype must be designated from among the specimens illustrated in the protologue. This choice is superseded if it can be demonstrated that the original type specimen corresponds to another validating illustration.” There was no validating illustration for Dictyocha hemisphaerica provided by Ehrenberg in 1844 and none by Haeckel in 1887 when the entity described by Ehrenberg was re-named as Cannopilus hemisphaericus. Yet under the type method used in the ICBN it is the identity of the specimen of Ehrenberg, rather than Haeckel which determines how the name Cannopilus must be employed.

But this applies to the lectotypification of a species name rather than a generic name and the typification of the name Cannopilus must be determined under Art. 10.5 of that Code: “The author who first designated a type of a name of a genus or subdivision of a genus must be followed, but the choice may be superseded if (a) it can be shown that it was in serious conflict with the protologue and another element is available which is not in serious conflict with the protologue, or, (b) that is was based on a largely mechanical method of selection.” Is Cannopilus a silicoflagellate or is it a phaeodarian? Haeckel's description, quoted above, indicated that what he had seen were skeletons of Dictyocha which he believed to be parts of a discrete organism. Can therefore the concept of the ‘epitype’, (see ICBN 2000 Art. 9.7) be used?

This article states “An epitype is a specimen or illustration selected to serve as an interpretive type when the holotype, lectotype or previously designated neotype, or all original material associated with a validly published name is demonstrably ambiguous and cannot be critically identified for the purposes of the precise application of the name of a taxon” and Art. 8.5, stating: “The type, epitypes (Art. 9.7) excepted, of the name of a taxon of fossil plants of the rank of species or below is always a specimen (see Art. 9.13). One whole specimen is to be considered as the nomenclatural type (see Rec. 8A.3).”). An epitype is not applicable in the present case because there is no problem over the identity of the type of Cannopilus hemisphaericus.

The lectotypification of the generic name Cannopilus on C. hemisphaericus was made by Frenguelli (1940) and it was in serious conflict with the generic protologue because, when Haeckel created the nominal genus, it was assumed to be a genus of phaeodarians. It would be nice to be able to lectotypify Cannopilus on C. cyrtoides Haeckel, 1887, (by nominating the specimen figured in Pl. 114, figs. 11–12 of the Atlas to the Challenger report as nomenclatural type of both the specific name and the generic name Cannopilus) but this cannot be done because of Frenguelli's prior selection. Although the drawing of Cannopilus calyptra is (typically for Haeckel) idealised and schematic, it plainly represents a doubly fenestrated example of Dictyocha speculum (alternatively ‘Distephanus aculeatus’) such as was illustrated by Locker (1974) in his revision of the silicoflagellates in Ehrenberg's collection (Locker, 1974 Tafel III, fig. 6, the holotype of Dictyocha binoculus Ehrenberg 1845, here described by Locker as a “cannopiloid Variante”). Splitters may wish to regard this as “Distephanus aculeatus” as Locker did; but this taxonomy is nugatory. Two other teratological specimens figured by Locker from Ehrenberg's collection (Tafel VI, figs. 4–5) are there called Cannopilus hemisphaericus and differ in having three or four ‘apical windows’ rather than the normal one (‘Distephanus speculum’ or two ‘Distephanus aculeatus’).

To prevent the name Cannopilus ever being disinterred from the grave of synonymy I also propose its formal rejection under ICBN 2000 (St. Louis Code), Art. 56: “Any name that would cause a disadvantageous nomenclatural change (Art. 14.1) may be proposed for rejection. A name this rejected, or its basionym if it has one, is placed on a list of nomina utique rejicienda (App. IV). Along with the listed name, all combinations based on them are similarly rejected and none is to be used.” Given the conceptual error which led to the proposition of the name Cannopilus, it seems highly desirable to dispose of the unnecessary and confusing name permanently and the addition of the name to the list of nomina utique rejicienda is justified. A formal proposal will be made elsewhere.

3. Taxa of “cannopiloid” or “cannopilid” silicoflagellates

Several silicoflagellate genera and species have been assigned to Cannopilus subsequent to its establishment by Haeckel. These are poorly studied and all are based on a small number of fossil specimens. McCartney (in Lipps, Fossil prokaryotes and protists p. 148) provides a concise summary: “The cannopilids [or cannopiloids] have an apical ring that is subdivided by additional elements to form multiple windows. The basal ring varies from a distephanid-like hexagonal ring and six basal spines to a nearly circular ring and many windows that can, in extreme cases, give it a radiolarian appearance. Skeletons with less than six basal sides are relatively rare. The six-sided and few-windowed forms are considered as members of Distephanus by some workers on the basis of associations with one-windowed types and from evidence supplied by double skeletons. This genus is called Caryocha by some workers.”

The cannopilids ought to be critically examined afresh; they appear to fall into three groups. The first group comprises speculiform silicoflagellates with a (usually) hexagonal to octagonal basal ring with radial spines and a saucer-shaped irregularly fenestrated cap on the apical side. Examples of nominal taxa in this group are: Cannopilus calyptra var. spinosus Lemmermann, 1901; C. cyrtoides var. nonaria Schulz, 1928; C. haeckelii Lemmermann, 1901, C. hemisphaericus var. polyommata Schulz, 1928, C. schulzii Deflandre, 1962 and C. schulzii forma longispinus Bachmann, 1962. Dictyocha hemisphaerica Ehrenberg, 1844 is the earliest name for a member of this group. Similar to these are the specimens illustrated by Locker and Martini (1987 Tafel 5, figs. 38–49) under the name “Distephanus speculum subsp. notabilis n. subsp.” divided into a number of ranked “forms”.

The second group comprises forms with a prominent subglobose apical cap which is heavily fenestrated with a much less prominent basal ring beneath, some forms having spiny processes at the margins of the basal ring or on the cap: all of these are Miocene in age. Nominal taxa included here are: Cannopilus tetraceros Deflandre, 1962 (from Spain); C. ichikawae Bachmann, 1964 (from Austria); C. jouseae Bachmann, 1964 (from Japan); C. latifenestratus Bachmann, 1964 (from Austria) C. iidaensis Bachmann et al., 1967 (from Japan) C. virginicus (Ehrenberg) Locker, 1974 (= Halicalyptra virginica Ehrenberg, 1854), and C. depressus (Ehr.) Locker, 1974 (= Halicalyptra depressa Ehrenberg, 1854). The last two of these are both from Richmond, Virginia and their types have been figured by Locker “Prap. Richmond 6a w” and “Prap. Richmond 11b r” respectively (Locker, 1974 Tafel IV).

Very similar to the second group are two nominal species with more globose caps and emergent apical spines, superficially resembling radiolarians: Cannopilus picassoi Stradner, 1961 and C. sphaericus Gemeinhardt, 1931 both reported from the Miocene of Austria. All of these Miocene taxa are quite similar, and a taxonomist inclined to lump might well consider all of them conspecific.

Finally there is a singular small species with a fenestrated cap but inconspicuous basal ring, which has been generically separated as Clathropyxidella similis Deflandre, 1938 (from the Eocene of Barbados) the type of the generic name Clathropyxidella Deflandre, 1938. McCartney overlooked the genus of Deflandre but it seems probable that the generic name Clathropyxidella should be retained and Halicalyptra depressa (including all the above named “cannopilid” species of group two, with the exception of C. minima) should be transferred to it. This must, however result in a new name: Clathropyxidella depressa (Ehrenberg) Parkinson, to be published elsewhere, as the name cannot be not effectively published herein.

Unfortunately, yet another name has been proposed for this group. This is Caryocha Bukry et Monechi, 1985 T: C. depressa (Ehrenberg) Bukry & Monechi, 1985 (= Halicalyptra depressa Ehrenberg, 1854). The concept of Caryocha was created essentially as a substitute for Halicalyptra / Cannopilus in its silicoflagellate sense as the outcome of the earlier work of Locker (1974). The flatter cannopilid silicoflagellates (C. hemisphaericus) previously considered to exemplify Cannopilus were excluded from Caryocha and “Because the type species of Cannopilus is more readily classified in genus Distephanus as Distephanus speculum hemisphaericus (Ehrenberg), than with silicoflagellates having globular apical structures, the new genus Caryocha is proposed for a more circumscribed classification.” (Bukry and Monechi, 1985 p. 378).

In view of the previous erection of Clathropyxidella, however, Caryocha is conceptually superfluous as treated here. Bukry and Monechi transferred to Caryocha six nominal species formerly included in Cannopilus (depressa including sphaericus), ernestinae, ichikawae, jouseae, latifenestratus, and picassoi, (see above) but not tetraceros or iidaensis (both from Japan) or Clathropyxidella minima. Consequently there is no reason not to use the name Caryocha if one considers it generically distinct from Clathropyxidella, but I see no basis to justify separating the two nominal genera Clathropyxidella and Caryocha. I consider, moreover, that all the nominal species assigned to Caryocha by Bukry and Monechi are probably conspecific.

Further palaeontological study will be required to clarify the relationships of these doubtful taxa. In comparing the illustrations of them provided by their authors one cannot help but notice that the illustrations themselves impart a great deal of the ‘generic’ character to organisms with broadly similar body plans. Before declaring firm opinions on the possible synonymy of these nominal taxa it will be necessary for the original material to be compared. Reliance on illustrations of varying quality — some of which appear to be idealised reconstructions — is not sufficient.

Most often these ‘cannopilid’ silicoflagellates are classified by biologists as minor taxa within Distephanus speculum — as by McCartney and by Locker. There appears to be a consensus among geologists in favour of retaining a generic separation between Dictyocha and Distephanus and the question must be asked whether there is a legally available name to replace Distephanus embracing all those with normally hexagonal basal rings and a so-called ‘apical window’?

But I feel that a new name is unnecessary, as it would be coined purely to satisfy the apparent wishes of geologists for something to replace Distephanus as a name for normally six-sided morphotaxa. Let somebody else do this, and then face a move to restore Distephanus by conservation. I do not believe that the morphotaxon Dictyocha is generically distinct from the morphotaxon Hexadictyocha. I prefer to stick with the classification and names Ehrenberg started with.

4. Mesocena and its ‘relatives’ (Bachmannocena, Septamesocena, and Paramesocena)

The first silicoflagellate referable to Mesocena (as originally characterised) was Dictyocha triangula Ehrenberg, 1839 — a species with a prominent hole in the middle rather than a bridging structure. Mesocena (conceptually speaking) was first established as a subgenus of Dictyocha — “Dictyocha (Mesocena) Ehrenberg 1840” with described with three included nominal species: D. (M.) circulus Ehrenberg, 1840, D. (M.) elliptica Ehrenberg, 1840 and D. (M.) triangula Ehrenberg, 1839, the sole original species. All are described with Latin diagnoses under genus Dictyocha but unfortunately no diagnosis of Dictyocha subg. Mesocena was given and the name appears here (Ehrenberg, 1840b) as a nomen nudum. The name Mesocena was subsequently validated at generic rank in 1843 with reference to the three original species of 1840 and two new species (added provisionally with a sign of doubt). Combinations under the generic name Mesocena appeared only in 1844. The position of each of the nominal species of 1840 therefore needs to be considered.

The three ‘original’ nominal species are all defined by their general shape (triangular, elliptical or circular) and are named accordingly as D. triangula “cellula triangularis unica, trianguli lateribus asperis, apicibus mucronatis” from the Miocene of Caltanisetta, Sicily; D. elliptica “cellula unica elliptica obscure quadrangula, spinis quatuor in formam quadratum dispositis, marginalibus. Diam 1/52'' fossilis in marga insulae Zacynthi (Zante)”; and D. circulus “unica circulari margine dentata. Diam. 1/48''. Fossilis in marga Graeciae.”

As noted above, Mesocena was validly published as a generic name in 1843 (Ehrenberg, 1843b p. 401). Although Ehrenberg implicitly included all the three earlier species, he did not provide binomial combinations for them in this paper, although he explained his intention to do so clearly enough: “A genus from the Family Bacillariae in a new section Lithotheciae. Lorica simple, univalve, siliceous, a small ring, circular or angulose, often with spines on it, (= Dictyocha, lacking the cellules in the middle.” (from the Latin, my translation). He continues (in German, my translation): “This category was included under the name Dictyocha in 1840. The absence of the middle cellules is the chief grounds for separation.” He then includes two new nominal species, both marked with question marks to show that he was not certain that they really belonged here. These are “Mesocena ? heptagona, annuli denticulis externis septem” and “Mesocena ? octogona, annuli denticulis externis octo” both of which are given small and inadequate figures and were described as “Recent” from Lima, Peru, having been found in Pleistocene or Holocene guano. These two species are clearly late additions to the concept of Mesocena — in fact they are fossil or subfossil examples of Octactis. Ehrenberg supplied binomials for the three original species in a further paper issued in 1844 (Ehrenberg, 1844a).

McCartney (in Lipps, 1993) calls Mesocena (following a subsequent tradition) ‘Bachmannocena (forms without apical structures)’ and writes: “Three- to many-sided bachmannocenids consisting only of a basal ring without an apical structure, were, until recently, placed in the genus Mesocena. The three- to six-sided forms may be quite similar in size, spine length and ornamentation to other silicoflagellates that are associated with them in the same rock. These may be ecophenotypic variants of Corbisema or Dictyocha. The many-sided forms, however, are distinct and are not similar to any other types: they are sometimes referred to in the separate genus Paradictyocha.” (p. 148). However, as I have shown above, Paradictyocha is merely the fossil state of Octactis. Leaving aside Octactis for the time being let us consider the concept of Mesocena in order to establish the typification of that name.

5. The lectotypification of Mesocena

The lectotypification of the name Mesocena is highly problematic. Three selections have been made in the past and I will now consider each of these in turn. The nomenclatural type specimens of all three nominal species have been illustrated by Locker (1974 Tafel II, fig. 4) for M. elliptica, Tafel II, fig. 10 for M. triangula and Tafel II, fig. 11 for M. circulus).

Mesocena triangula (Ehrenberg) Ehrenberg, 1844 (= Dictyocha triangula Ehrenberg, 1839) is the third of the original species. The species was well illustrated by Ehrenberg (1854b) and has also been well refigured by Locker (1974 Tafel II, fig. 10). The only drawback to its selection as lectotype of Mesocena is that it is represented in Ehrenberg's collection by a single specimen and this singular specimen is unidentifiable (it could even be an ‘aberrant’ corbisemid).

Mesocena circulus (Ehrenberg) Ehrenberg, 1844 (= Dictyocha circulus Ehrenberg, 1839) is the first of the original species and has also been well refigured by Locker (1974 pl. II fig. 10).

Mesocena elliptica (Ehrenberg) Ehrenberg, 1844 (= Dictyocha elliptica Ehrenberg, 1839) is the second original species and is discussed in more detail below.

The first lectotypification was made by Frenguelli (1940 p. 69) who selected M. elliptica (Ehrenberg) Ehrenberg, 1844. This choice was rejected by Loeblich et al. (1968) on the basis that this species was “not included in the publication wherein Mesocena was validated by description.” That statement is incorrect, however, because the species was implicitly included on account of the direct reference to the 1840 paper in which it was described. Procedurally, therefore, there can be no objection to Frenguelli's selection.

In lectotypifying Mesocena in 1940 Frenguelli says, however: “I propose therefore that the name of Mesocena Ehrenberg be reserved for the forms of [my] third type, which correspond most exactly to the original diagnosis (‘individual libera solitaria, annulum circularem aut angulosum saepe spinescentem referentia’) and which includes the first form described, the name of which is Mesocena triangula Ehrenberg, 1838.” (my translation from Spanish). The ‘third type’ means ‘angulosum saepe spinescentem’ and refers explicitly to M. elliptica. The basis for Frenguelli's preferring M. elliptica and M. triangula over M. circulus was therefore the fact that there were spines at the angles of the angular species elliptica and triangula but not on the third species, M. circulus. It seems to me that Frenguelli really preferred M. triangula rather than M. elliptica on the grounds of priority. But after discarding M. circulus he was was forced to select M. elliptica. This lectotypification must stand under ICBN Art. 9.13.

The second lectotypification of Mesocena was by Glezer (1966 p. 282) who selected M. circulus (Ehrenberg) Ehrenberg, 1844 as type, and this selection was also rejected by Loeblich et al., for the same reasons as led him to reject M. elliptica (and equally incorrectly). Glezer does not indicate why he lectotypified Mesocena on M. circulus but it was probably because it was the first species mentioned, (following the zoological custom not approved by the ICBN). No question of “page priority” arises since all three nominal species were described on the same page of Ehrenberg's work (Ehrenberg, 1840b p. 208). This might be regarded as unfortunate, as M. circulus would have been a more convenient choice, but Frenguelli's earlier choice still stands.

Loeblich et al. (1968 p. 54) then formally lectotypified Mesocena a third time on M. octogona Ehrenberg, 1843, one of the two named ‘included species’ of the paper in which Mesocena was validated. This selection is itself objectionable on account of the fact that the species was not “definitely included” (ICBN Art. 10.2). It was also a bad choice on account of the question mark used by the author of the protologue and that selection is here formally rejected.

In 1974 Locker stated that Frenguelli had lectotypified Mesocena on M. elliptica but remarked: “Das Taxon Mesocena ist vom Jahre 1841 an als gültig publiziert zu betrachten, da es in Verbindung mit mehreren verfügbaren Arten eingeführt wird (IRZN, Artikel 16a-v). Gleichzeitig kann Mesocena elliptica als korrekt festgelegte Typusart angesehen werden, weil sie zu den drei ursprünglich eingeschlossenen Arten gehört (IRZN, Artikel 69a-i).” (My translation: ‘The taxon Mesocena has shown to have been made lawful in 1841 [i.e., 1840] as it was established with several species [International Rules of Zoological Nomenclature] Art. 16 (a)-(v). At the same time M. elliptica can be shown to be the correct type species under the rules IRZN Art. 69 (a)-(i)’). Even if the Locker analysis had been correct under the ICZN, however, it is not correct under the ICBN. It was not until ten years later (1985) that Locker and Martini adopted the ICBN rather than the ICZN, rejecting the ICZN published in that year (Ride et al., 1985). 12

With the selection of M. elliptica as type of Mesocena the matter should have been settled, but this was not to be. Locker (1974 p. 634) excluded all other species of Mesocena to another genus (Bachmannocena Locker, 1974) therein erected. The basis for the drastically altered circumscription was that, since internal crosspieces had been observed in some specimens of M. elliptica, it could not be genuinely congeneric with the authentically ‘open-middled’ species — for which Mesocena was originally created. In those circumstances Locker might have had a case for rejecting the lectotypification on M. elliptica as inconsistent with the protologue under ICBN 2000 (St. Louis Code) Art. 9.17 or an earlier version (“The author who first designates a lectotype . . . must be followed but his choice is superseded if . . . it can be shown that it is in serious conflict with the protologue. . .”). But that option was not to his inclination.

6. Bachmannocena

On accepting the lectotypification of Mesocena on M. elliptica Locker adopted a new generic name Bachmannocena for “skeletons with a basal ring only” assigning the genus to the family Dictyochaceae, along with Dictyocha and Naviculopsis. He simultaneously segregated Mesocena into a new family Corbisemaceae Locker, 1974 (correctable to ‘Corbisemataceae’).

Bachmannocena was based on B. diodon (Ehrenberg) Locker, 1974 = Mesocena diodon Ehrenberg, 1845, a simple ring with opposite marginal spines. It is also illustrated by Locker (1974 Tafel II, fig. 9). The differences between the circular forms “Bachmannocena circulus” of Locker and the elliptical “Bachmannocena diodon” forms and the quadrangular “Mesocena elliptica” forms are minor and certainly do not merit separation at the family level. Three-sided specimens were placed in another genus, Corbisema. The generic concept Bachmannocena was subsequently altered by Bukry (1987 p. 403) in such a way as to create even more confusion. Bukry says: “The original genus Bachmannocena [Locker, 1974 ] was limited to ring silicoflagellates with two or more spines. That restriction is herein removed and emended to include all natural ring silicoflagellates that lack apical structures, whether or not they have spines. Mesocena has been the most widely used generic name for such ring silicoflagellates, but Locker's (1974) restudy of the Ehrenberg type materials led him to recognise that the designated type species of Mesocena [i.e., M. elliptica (Ehrenberg) Ehrenberg, 1844] is more closely related to the Distephanus-like genus Octactis.” This gives the conventional modern circumscription of Bachmannocena.

This Bukry statement of 1987 is erroneous, however, since Locker (1974 pp. 633–635) actually assigned Mesocena to the new family ‘Corbisemaceae’, Octactis to the family Vallacertaceae, and placed Distephanus in yet another new family ‘Distephanaceae’. Bukry seems to have thought that the typification made by Loeblich et al. for Mesocena (i.e., the selection of Mesocena ? octogona Ehrenberg, 1843) had been accepted by Locker, whereas Locker had quite properly rejected it. The effect of Bukry's diagnostic alteration to Bachmannocena was to expand Bachmannocena to the original Ehrenberg limits for Mesocena (i.e., admitting any silicoflagellate with an aperture, but without a transverse septum. The collapse of this taxonomy began when it was found that some ring silicoflagellates had secondary more fragile structures in their rings which Ehrenberg had not been able to see in his specimens from Peruvian guano. This can be seen upon a comparison of Locker (1974) Tafel II, fig. 5–8, named by Locker as “Octactis ? bioctonaria” with Hsin Yi Ling and Takahashi's (1985) Plate 1–2 there named as “Distephanus pulchra” and which show essentially the same images.

Mesocena” as now modified by Bukry had become conceptually equivalent to Paradictyocha Frenguelli, 1940. The identity of the living Octactis and fossil Paradictyocha had been apparent earlier to Dumitrica who, according to Bukry, wanted to avoid the ‘mistypification’ of Mesocena adopted by Loeblich and favoured M. elliptica. Apparently Dumitrica also “recommended changing the type species of Paradictyocha Frenguelli, 1940 from the original Distephanus polyactis (Ehrenberg) to Mesocena circulus var. apiculata Lemmermann, because of Dumitrica's solutionary theory about the derivation of different mesocenid specimens.” Bukry disregards the fact that the selection of M. elliptica (by Frenguelli, 1940) over M. circulus (by Glezer, 1967) has priority under ICBN (Art. 8.1). He also objected to the lectotypification of Mesocena on M. elliptica on the basis that “Dictyocha circulus Ehrenberg (1844 p. 65) had page priority” but page priority is an ICZN concept, not recognised by the ICBN. Bukry was then using a version of the ICBN which was three editions out of date (1966 edition) and clearly had a poor grasp of its implications.

Ultimately Bukry seems to advocate abandoning the name Mesocena and using Bachmannocena instead: “The genus Bachmannocena Locker, 1974, with B. diodon (Ehrenberg) as the type species, is accepted as the first practically and procedurally correct taxon for ring silicoflagellates. Conservation of the widely used name Mesocena could be considered . . . In any event, use of the emended Bachmannocena is accepted here as an unambiguous means to deal with the ring group. Historically, Mesocena has been a minor silicoflagellate genus; Loeblich et al., (1968) recorded only 46 valid names in the genus. About 20 additional taxa have been added from 1968 to 1985. Because of the manageably small number of taxa involved (smaller still when synonyms are identified) conservation is not a practical necessity.”

The primary objective for Bukry seems to have been the establishment of a distinctive generic name for all “ring silicoflagellates” (those without apical structures) while avoiding the forced adoption of the old and widely used name Mesocena. Mesocena was emerging as the inconvenient prior name for the synonyms Octactis and Paradictyocha. But is Mesocena really the same as Paradictyocha / Octactis? The answer is “No” for Mesocena is a morphotaxon, whereas Octactis is a neotaxon. They have different circumscriptions and positions; they belong as it were, in parallel universes.

The solution to this problem is relatively simple, once the misapplications of irrelevant nomenclatural codes are cleared away. The selection by Loeblich et al. (1968) of “Mesocena ? octogona” as type of Mesocena certainly cannot stand. The typification of Mesocena on Mesocena elliptica by Frenguelli can (and should) stand — unless the alternative (M. circulus) is formally conserved against it (see ICBN Art. 10.5, Art. 14.1, and Art. 14.5) in the interests of nomenclatural stability. In my opinion the selection of Mesocena elliptica by Frenguelli should stand. Conservation of the name Mesocena to stabilise its typification is entirely unnecessary.

Bachmannocena Locker, 1974 was typified on B. diodon (Ehrenberg) Locker, 1974 (Mesocena diodon Ehrenberg, 1844) lectotypified on “Piscataway 2a w” in Ehrenberg's collection, from the Miocene of Piscataway, Maryland, USA. This I propose to regard as a species of Mesocena, thus disposing of the generic name Bachmannocena into synonymy. That clarifies the identity of Mesocena but still leaves the problems of its segregates (now placed in a different family!). There are three of these: Septamesocena, Paramesocena, and Paradictyocha all based on fossil types. These are considered next.

7. Septamesocena

Septamesocena Bachmann, 1970 (?T: S. elliptica (Ehrenberg) Bachmann, 1970 (= Dictyocha elliptica Ehrenberg, 1840) had been established as a generic segregate for M. elliptica following the mistypification of the name Mesocena by Loeblich et al. (1968). Internal cross bars in some mesoceniform silicoflagellates are well illustrated, for example, in Loeblich et al. (1968, Pl. 16, figs. 1–8) in Dictyocha formosa Bachmann, 1964 from the Miocene of Japan. I attach no importance to these: in many fossils they are probably lost. In any event the type specimen of Septamesocena elliptica does not have them. Locker rejected the name Septamesocena as falling into the synonymy of Mesocena (Locker, 1974 p. 634) but the name is actually illegal as a “superfluous name” under ICBN (Art. 52.1–51.2) and is accordingly disposed of automatically.

Locker, after examining Ehrenberg's original material of Mesocena elliptica said (my translation) “In various skeletons of the type population internal septa have been demonstrated. The genus Septamesocena established by Bachmann (1970b) on the basis of the internal partitions therefore falls into synonymy with Mesocena.” (Locker, 1974 p. 634). Curiously, however, he used it again in a late publication (1996 p. 114) for two nominal species, Septamesocena apiculata (Schulz) Bachmann, 1970 (= Mesocena apiculata Schulz, 1928) and Septamesocena oamaruensis (Schulz) Bachmann, 1970 (= Mesocena oamaruensis Schulz, 1928). These two nominal species are here restored to Mesocena (q.v.).

8. Paramesocena

Next in this group of unnecessary names we have Paramesocena Locker and Martini, 1985 T: P. apiculata (Lemmermann) Locker & Martini, 1985 (= Mesocena circulus var. apiculata Lemmermann, 1901). This was created “to solve the many questions related to discrimination of the mesocenoid apiculata-circulus group from the genus Distephanus. As several authors have indicated, this species group represents a special evolutionary line which certainly originates from the D. polyactis group . . . If the ancestor is included in a separate genus covering this line, Paradictyocha Frenguelli, 1940 can be employed. But if one maintains the opinion that D. polyactis still belongs to Distephanus, another genus must be used, because D. polyactis has clearly been indicated as type of the genus Paradictyocha (Frenguelli, 1940). According to Art. 48.1 of the ICBN, it is not permitted to designate a new type species for Paradictyocha, as Dumitrica (1973b) did, solely to preserve the genus name. Because no genus was available to comprise this group, the new genus was introduced.” (Locker and Martini, 1985 p. 909).

Paramesocena was defined as “Skeleton consisting of a basal ring which is wide, polygonal or circular with numerous radial spines or nodes; basal pikes are generally absent. Distinguished from Mesocena and Septamesocena by the high number of radial spines displayed in all species.” (Locker and Martini, 1986 p. 909). The genus was based on three nominal species — P. apiculata, P. circulus (Ehrenberg) Locker and Martini, 1985 (their renaming of Dictyocha circulus Ehrenberg, 1840); and Paramesocena dumitricae (Perch-Nielsen) Locker and Martini, 1985 (their renaming of Paradictyocha dumitricae Perch-Nielsen, 1975, for citation reference see bibliography). These species are figured by Locker and Martini, (1985) plate 9 fig. 7.

Locker and Martini essentially re-invented Mesocena in a restricted sense for silicoflagellates without either apical structures or angles. The essential shape of these skeletons is circular, as represented in Mesocena as originally conceived and based on Mesocena circulus. Paramesocena is a redundant renaming of Paradictyocha — the fossil expression of the neotaxon Octactis. The name Paramesocena is therefore disposed of.

9. Octactis and Paradictyocha— the seat of the problem

Octactis Schiller 1925 was based on a single newly described species — Octactis pulchra Schiller, 1925. The protologue provides a Latin diagnosis (with German translation) of the description of the genus, used for the species also (Schiller, 1925). The Latin version says (my translation): “Basal ring with eight sides and spines radiating out, the sides of the basal ring with short obconical processes, on which eight thin arches sit, forming an apical ring.” The German version says (my translation) “Basal ring eight sided, broad with eight radiating spines. On each side of the basal ring is a short conical process from the eight of which there project into the open space very slender and fragile vaulted bows which form an apical ring which is very wide. Plasma bodies very seldom seen, frothy, loose, with three – five pale yellowish chromatophores.”

Schiller amplified this diagnosis with observations (my translation): “Basal ring broad, oval or circular, each of the eight spines constructed (from optical observation) of two different substances in layers. The eight spines are in pairs directly opposite and some pairs are clearly larger than others. Apical ring very delicate, beautifully constructed of eight vaulted bows, the eight bow windows open. Apical opening very large. Cytoplasm very delicate with little pale yellowish coloured chromatophores. Flagellum not seen. Diameter including spines 50–60 mu. [Seen] Summer, spring. [Place] Central and Southern Adriatic, 0–25 metres presumably a tropical form. Found here and there singly, quite at random. The above-mentioned form is regarded as allogenetic for the Adriatic [i.e., is foreign to the Adriatic] They occur at a time of considerable warming of the Adriatic water, rare and scattered in the surface water. Possibly it is one of the forms from the Mediterranean and tropical parts of the Atlantic and Indian Oceans which now, like many other plankton, come into the Adriatic with the current and become completely acclimatised.” In fact these organisms had been found in the Mediterranean long before; they are present in the Miocene samples obtained from Sicily by Ehrenberg in 1839. Octactis pulchra is the neotaxon corresponding to the fossil morphotaxon Dictyocha polyactis Ehrenberg, 1839, described from Caltanisetta and well figured by Locker (1974 Tab. IV fig. 2). The type of D. polyactis differs from most specimens only in having nine sides, rather than the usual eight. Ehrenberg himself mistakenly described it as ten-sided. D. polyactis Ehrenberg, 1839 is the nomenclatural type of Paradictyocha Frenguelli, 1940 (Frenguelli, 1940 p. 59, 69) regarded by its author as a mesoceniform derivative of Dictyocha. His description (translated from Spanish) reads “Paradictyocha, ordinarily a small ring formed by the close juxtaposition of two rings, the one basal and the other apical, of polygonal outline with numerous sides and a corresponding number of short radial horns without an apical apparatus in the cavity of the ring.”

Locker (1974 p. 637) assigned this species to Distephanus (as Distephanus polyactis) but treated Octactis as a proper genus in a different family (1974 p. 636). Here he placed emphasis on what he described as a thin apical disk situated between the bows. This suggested to him a connection with the ancient pentagonal Vallacerta (q.v.), a genus of silicoflagellates known only from the Cretaceous. This “apical disk” was not reported by Schiller, nor by Frenguelli, who both describe the apical ring as open. Locker, however, figured several specimens (including the types of Mesocena bioctonaria Ehrenberg, 1845 and Mesocena binonaria Ehrenberg, 1845) which were described as fossils from guano in Peru (see figures in Locker, 1974 Tafel II, figs. 5 and 6). These specimens undoubtedly conspecific, one representing an eight-sided and the other a nine-sided morph, of Dictyocha polyactis.

Locker calls the fossils Octactis bioctonaria and suggests that two other nominal species of similar appearance, both also from Peruvian guano — Mesocena octogona Ehrenberg, 1843 and Mesocena heptagona Ehrenberg, 1843 — might be better assigned to “Octactis ? bioctonaria (Ehrenberg, 1846) comb. nov.” despite the non-survival of the ‘apical disk’ (see figures in Locker, 1974 Tafel II, figs. 7 and 8). All are of these are simply fossils of Octactis pulchra being eight sided and seven sided morphotaxa respectively.

Not mentioned by Locker, but helpfully figured by Loeblich et al., (1968) is the type of Dictyocha rotata Glezer, 1967 said to be of Upper Eocene age from the USSR. This is a heptagonal speculiform Dictyocha demonstrating seven internal slender bows forming an apical ring united by a thin perforated siliceous “apical disk”. A somewhat similar structure seems to be indicated in the photographic illustration of “Dictyocha fibula var. octagona Tsumura” in Loeblich et al., (1968 pl. 13 fig. 6).

Hsin Yi Ling and Takahashi (1985 p. 78) have reinvestigated the so-called ‘loop structure’ reported by Schiller and the so-called ‘apical plate’ or membrane. They have stated that they doubt such plates exist but suggest that they are optical illusions just like the loop structure reported by Schiller. I accept their explanation. The redrawn figures (by Locker after Ehrenberg) for Mesocena ? octogona Ehrenberg, 1843 and Mesocena ? heptagona Ehrenberg 1843 (see Locker, 1974 Tafel II, fig. 7 and 8) and Hsin Yi Ling and Takahashi's reproduction of the original figures for Mesocena bioctonaria Ehrenberg, 1854 and Mesocena binonaria Ehrenberg, 1854 (1985 pl. 2), show the same skeletal structures. Other octagonal specimens of Octactis usually assigned to Mesocena (e.g., from the Miocene of Japan) might represent a separate species of this genus in which the more delicate apical apparatus of ring and disk has not survived, leaving only the stubs of the apical ring e.g., the so-called “Mesocena polymorpha var. bioctonaria forma fossilis Tsumura, 1963”. Further study of these is required.

Assembling these different observations, I conclude that Octactis and Paradictyocha would be congeneric were it not for the fact that one is a neotaxon and the other is a morphotaxon. They are formally distinct taxa occupying separate positions in the classification of silicoflagellate biota. As morphotaxa and neotaxa can be named separately the fossil specimens can be assigned to Paradictyocha but in cases where the fossils are named as congeneric with Octactis the name Octactis must be used (ICBN 2000 Art. 11.7). Octactis and Paradictyocha are not, however, congeneric with the morphotaxon Mesocena since the latter does not exhibit (so far as can be told from fossil material) the characteristics of the living Octactis, an ‘apical apparatus’ of a distinctive structure.

Irregular spiny processes, internal bars or perforated disks (as are thought to be seen in Vallacerta) will all serve to strengthen simple rings, just as simple crossbars will suffice in naviculiform morphs, and just as stapediform (stirrup shaped) or fibuliform (brooch-like) crossbars in four-sided morphs and just as hexagonal speculiform (hand-mirror like) structures in six-sided morphs. When a morph is more than six-sided the classical speculiform apical structure is no longer structurally efficient and more diverse structures are used to extend the skeleton including the morphologies termed ‘cannopilid’ and ‘polyommatid’ (those with multiple apical windows). But all of these trivial structural variations are plastic and ephemeral and they probably have no phylogenetic significance.

10. Corbisema and Phyllodictyocha

Corbisema Hanna, 1928 (Type: C. geometrica Hanna, 1928) was established for a silicoflagellate of more or less triangular shape, with a trigonal septum. It is evidently one of a group of allied taxa known only from fossils ranging from the Cretaceous to (doubtfully) the lower Miocene. McCartney (in Lipps, 1993 p. 146) comments of ‘Corbisema (three-sided forms)’ : “This group has a basal ring with three sides and three struts. These struts attach to the basal ring at or very near the midpoint of each side and meet at their opposite ends at a single point or, rarely, form an apical ring or plate. The apex of the apical structure is positioned directly above the centre of the basal ring. Spines may be at the corners of the basal ring. Most corbisemids have basal rings that are roughly equilateral, though C. hastata has one side that is somewhat shorter than the other two, giving it a bilateral symmetry. Although the corbisemids have a basal ring with three corners by definition, two and four-cornered variants are known. The corbisemids are generally dominant in the Cretaceous and much of the Palaeogene. Their importance declines in the Oligocene and only the relatively small C. triacantha is found in any abundance in the Miocene.”

Corbisema can be characterised as a silicoflagellate with a more or less equilaterally triangular shape, sometimes with the apices extended into spines, and with either flat or convex sides. While some have been reported as ‘recent’, particularly by Lemmermann (1901), these are all from benthic ooze and consequently are probably of Tertiary age. The so-called “Corbisema triacantha” (= Dictyocha triacantha Ehrenberg, 1845 figured by Locker (1974 Tafel I, fig. 10) however, may be just a three sided specimen of a fibuliform Dictyocha such as Dictyocha varia Locker, 1975 (compare Locker and Martini, 1985 plate 2 figs. 1–2, 3–4).

The classification within the Corbisema complex and the naviculoid variants of Corbisema merit only brief mention here (below). Among these is Crassicorbisema Hsin Yi Ling, 1981 — a morphogenus not listed by McCartney and which is naviculopsid rather than corbisemid.

11. Deflandryocha

While there is no doubt that Deflandryocha Jerkovic, 1963 (T: D. cymbiformis Jerkovic, 1963) is a genuine silicoflagellate, it is not clear what its affinities are. Specimens are said to vary considerably, although most are naviculiform. The supposedly distinguishing character of rounded or spatulate apices is a trivial characteristic. Three nominal species, doubtless all conspecific, were originally found together in a single diatomite deposit of Miocene age in Yugoslavia (Sveta Nedalja). A further nominal species from the same deposit was added two years later. Some specimens resemble Dictyocha crux Ehrenberg, 1841. I prefer to regard the nominal species of Deflandryocha (D. cymbiformis Jerkovic, 1963, D. intercalaris Jerkovic, 1963, D. naviculoidea Jerkovic, 1963 and D. spathulata Jerkovic, 1965 as conspecific and probably conspecific with Dictyocha crux Ehrenberg, 1840, thus disposing of all these names into synonymy, with Deflandryocha becoming a junior synonym of Dictyocha. McCartney (in Lipps, 1993) does not mention Deflandryocha.

12. Distephanopsis and the cruciform ‘distephanids’

Dumitrica established the nominal genus Distephanopsis (T: Distephanopsis crux (Ehrenberg) P. Dumitrica, 1978 (= Dictyocha crux Ehrenberg, 1841) for a Miocene fossil. The combination intended was not validly published because the place of publication of the basionym was not cited by Dumitrica. Most workers treat the nominal genus as a junior synonym of Dictyocha. McCartney (in Lipps, 1993) does not even mention it.

13. Naviculopsis and other ‘naviculopsids’

Naviculopsis Frenguelli, 1940 (Type: N. biapiculata (Lemmermann) Frenguelli, 1940 = Dictyocha navicula var. biapiculata Lemmermann, 1901) was composed of two nominal species: N. biapiculata (Lemmermann) Frenguelli, 1940 and N. navicula (Ehrenberg) Frenguelli, 1940 (= Dictyocha navicula Ehrenberg, 1840). Ling (1972 p. 87) added N. quadratum (Ehrenberg) H. Y Ling, 1972 ‘quadrata’ (= Dictyocha quadratum Ehrenberg, 1845) as a transfer and Locker (1974 pp. 635 and 642, Taf. II fig. 3) illustrated the type, at the same time synonymising the species with Naviculopsis iberica Deflandre, 1950 and Dictyocha navicula var. rectangularis Schulz, 1928 which had previously been recombined as Naviculopsis rectangularis (Schulz) Frenguelli, 1940. Several other nominal species have been added; a list is given by Locker (1996 p. 114). McCartney (in Lipps, 1993 p. 148) comments: “Naviculopsids are very elongate with the minor axes connected by a bridge. [. . .] The apical structure consists of a strutless bridge or plate that arches across the breadth of the skeleton. The apical bridge varies in width and stratigraphic position and so is commonly used as a taxonomic character. The group is locally very abundant in Eocene rocks but declined through the Oligocene and became extinct in the early Miocene.” He also writes: “Some silicoflagellates have the naviculopsid shape but are considered within the genus [sic] Corbisema and Dictyocha. Dictyocha sometimes occurs without struts and strutted Naviculopsis are also found. Some workers consider such unusual skeletal morphologies as separate genera, but these morphologies are generally similar to standard ones that co-occur in the same rocks.” (p. 148).

14. Eunaviculopsis

Eunaviculopsis was described as a segregate from other naviculiform silicoflagellates on account of the absence of “flattened plate-like structure[s] at apices on major axis” (Ling, 1977 p. 212). Ling sought to distinguish the new genus, based on Dictyocha navicula Ehrenberg, 1840 ( = Naviculopsis navicula (Ehrenberg) Frenguelli, 1940) from specimens such as the type of Dictyocha ponticulus Ehrenberg, 1844, which were illustrated together as conspecific by Locker (1974 Tafel II, figs 1, 2). Eunaviculopsis was said to represent ‘the original naviculiform silicoflagellates’. Ling had earlier (Ling, 1972) suggested that Naviculopsis was polyphyletic: “Further investigation of this group of microfossils now substantiates the fact that those with the platelike structure or so-called ‘endflächen’ of Stradner (1956, 1961) at the apices of major axis [sic] are structurally distinctly different from and have a much more limited geographical range than those possessing definite radial spines at the apices and tubular basal body of the genus Naviculopsis; therefore, differentiation of these two groups is here proposed.” Unfortunately, however, although a ‘type species’ was indicated, Ling failed to comply with Article 33.2 of the ICBN, concerning new combinations, by not giving the requisite “full and direct reference” to the place of valid publication of the basionym for Eunaviculopsis navicula “with page or plate reference and date”, and consequently the intended new combination was not validly published.

15. Neonaviculopsis

Neonaviculopsis Locker & Martini, 1985 (1985 p. 908) is based on a single nominal species Neonaviculopsis neonautica (Bukry) Locker and Martini 1985 (= Dictyocha neonautica Bukry, 1981). As constructed by these authors it consisted of two subspecies (“ssp. neonautica” and “ssp. praenautica”) also (apparently in error) named as “forms” at one point (as “Neonaviculopsis neonautica (Bukry) ssp. neonautica. f. praenautica”). The species is described as consisting of a basal ring and an apical apparatus but is, in effect, a naviculiform silicoflagellate with apical pikes like Dictyocha quadratum Ehrenberg, 1845 (= Naviculopsis quadratum (Ehrenberg) H. Y Ling, 1972 ‘quadrata’) and Naviculopsis biapiculata (Lemmermann) Frenguelli, 1940 (= Dictyocha navicula var. biapiculata Lemmermann, 1901). Some of these specimens have a simple transverse bar linking the two sides (forms called “neonautica”) and others have a more fibuliform structure (forms called “praenautica” — see Locker and Martini, 1985 Plate 10: figs. 1–12). Locker and Martini subsequently described biapiculated naviculiform specimens as Naviculopsis robusta Deflandre, 1950 and non-apiculated specimens as “Corbisema disymmetrica” and “Corbisema inermis” depending on whether they were naviculiform or corbisemiform (see Locker and Martini, 1987 Tafel 2, fig. 15–16 and Tafel 3, fig. 23–24, 26–27 and Tafel 6, figs. 50 –51). The separation of these two or three sided skeletons at familial level (Locker, 1974 p. 633) was absurd and was abandoned by Locker and Martini in 1985 and in 1987. McCartney (in Lipps, 1993) does not mention Neonaviculopsis.

16. Crassicorbisema

In 1981 Hsin Yi Ling described Palaeocene silicoflagellates discovered during the Deep Sea Drilling Expedition (Hsin Yi Ling, 1981 pp. 1–12). Most of these, whether trigonal or naviculoid, were classified in the nominal genus Corbisema (as C. inermis, C. naviculoidea, C. cuspis, and C. constricta). One nominal species, however, was segregated as genus Crassicorbisema Ling, 1981 (T: C. disymmetrica (Dumitrica) Ling, 1981, based on the previous Corbisema inermis disymmetrica Dumitrica, 1973). The genus, as circumscribed by Ling, contained one species with two subspecies. The typical subspecies was segregated from Dumitrica's nominal taxon ‘Corbisema inermis’ pro parte.

In the generic protologue Ling states: “It may be argued that their unusually thick skeleton and corrugated surface microstructure are due to some environmental conditions, but this is considered to be unlikely, at least at the present time, because abundant specimens of the present species are observed with other ‘normal’-formed silicoflagellates in the same slides.” (Ling, 1981 p. 5). Ling characterised the genus on the basis of the thick skeleton and noted as diagnostic characters the ‘basal body ring, either well-shaped trilobate, bilobate or widely irregular [in] outline;’ (i.e., both corbisemiform and naviculiform shapes were seen).

As so often among silicoflagellate specialists, Ling refers to the ICBN as the nomenclatural authority for the group but cites an obsolete edition of the Code (the Edinburgh Code of 1966) rather than the then current Leningrad Code. Moreover, (as shown in the nomenclature adopted in his paper) his author citations are a mixture of ICBN and ICZN conventions, as is the style used for parenthetical citations of authorship. Bukry (1976) recognised three segregates from within ‘Corbisema inermis’ (his subspecies angulata, communis, and disymmetrica) but Ling divided each of these again under the names Corbisema cuspis Busen & Wise, 1977 Corbisema constricta (Busen & Wise) Ling, 1981 and Crassicorbisema disymmetrica (Dumitrica) Ling, 1981. Ling explains (but misinterprets) Art. 24 of ICBN 1966, (currently ICBN 2000 Art. 11.2) concerning epithets of names not having priority outside the rank in which they were published, but nevertheless the names adopted by Ling are legitimate. Contrary to Busen and Wise (as cited by Ling, 1981 p. 4) ‘Corbisema disymmetrica angulata’ does not have priority over C. cuspis and this is because the names are applicable to different ranks.

17. Nothyocha

Described by Deflandre in 1949 this nominal genus has seldom been referred to again. It was based on a single nominal species Nothyocha insolita Deflandre, 1949 (1949 p. 674 figs. 1–2) from the Miocene of Sicily (Helvetian and Sarmatian).

18. Bukryella

The most recently described genus of silicoflagellate morphotaxa is Bukryella T.V. Desikachary and P. Prema, assigned to the ring silicoflagellates along with Mesocena, Paramesocena, and Septamesocena. This was based on an Upper Eocene sample from the Norwegian-Greenland Sea, the type species being named as Bukryella eocaenica T.V. Desikachary & P. Prema, 1996. The species name was invalid on account of the use of an iconotype being used for the fossil, the authors having written “Holotype (iconotype): Martini, E. and Muller, C. 1976. Init. Repts. DSDP 38: pl. 10, fig. 3, a, b, c.” According to ICBN Art. 8. 5 the type of a name of a taxon of fossil plants at the rank of species or below is always a specimen, not an illustration. ICBN Art. 9.13 states, furthermore, that the type of such a name is the specimen or specimens on which the validating illustrations are based. Since an illustration was cited, rather than a specimen, the name Bukryella eocaenica T.V. Desikachary & P. Prema, 1996 was not validly published.

According to Desikachary and Prema the Bukryella skeleton is “a simple triangular basal ring” without spines but with “prominent swollen and spherical bulbs, structures, connected on two sides to the triangular basal ring.” (1996 p. 167). The spherical ‘bulbs’ were thought to be “mesocenoid parallels to dictyochoid Hannaites Mandra.” In their theoretical approach to the ring-silicoflagellates they acknowledge “a clear morphological bias based on consistent and dominant structures of individual forms. We have segregated genera with only basal rings into a separate family Mesocenaceae, as against those that have other accessory skeletal features such as apical bar [sic], apical rings and lateral struts, in much the same way as, and parallel to the importance given to the poles of a bipolar skeletons [sic] forming the basis for separation of different segregates of naviculoid forms.” (pp. 167–168).

At the same time they acknowledge that there is “a need to clear the conflicts in taxonomy that have crept in because of the use of, or the stress on, abnormal forms (also cultural variants in extant forms) and criteria other than morphological.” In essence, Bukryella is a three-sided morph of Hannaites, lacking the fibuloid apical structure of that genus, which is also of Eocene age. Whether Bukryella is really distinct from Hannaites (they are assigned to different families) is uncertain as Bukryella may simply represent a three-sided morph of Hannaites, suggesting another instance of confusion between ontogeny and morphology. The separation of Bukryella from Hannaites was indicated as provisional (“For the present, we would separate the triangular forms with spherical protuberances at angles into a separate genus” p. 168), and is an artefact of the familial classification.

Part III: The Archaic silicoflagellates

Finally, the five nominal genera of pre-Tertiary silicoflagellates need to be considered. They are not well studied and their nomenclature and taxonomy are relatively uncomplicated.

19. Cornua

Cornua Schulz, 1928 (Type: C. trifurcata Schulz, 1928) was established for a rather poorly characterised fossil of the upper Cretaceous and the genus has subsequently been expanded to include a large number of ill-defined nominal species. These include silicoflagellates assignable to Corbisema as well as undoubted fragments of ebriid skeletons. Lipps (1970) considered the identity of Cornua trifurcata Schulz doubtful, and suggested in might be a sponge spicule. Only re-examination of the original material seems likely to resolve this question. McCartney (in Lipps, 1993 p. 149) characterised cornuids as ‘triradiate skeletons without basal rings’ but notes: “It is very poorly known and rarely found. Quite possibly these are aberrant corbisemids whose basal elements do not connect at the corners to form a ring and their rare occurrence in coastal waters may support this conclusion. Cornua is almost unknown from deep-sea sedimentary rocks, which is also where aberrant silicoflagellates are less abundant. The recent discovery of Variramus in the Early Cretaceous, however, might indicate that Cornua is an intermediate between Corbisema and more primitive morphologies.”

20. Vallacerta

Vallacerta Hanna, 1928 T: V. hortonii Hanna, 1928 (‘hortoni’) is a distinctive early silicoflagellate, known only from fossils and characterised by Hanna as “A convex sculptured pentagonal disk with a heavy siliceous spine at each corner.” Some specimens described by Deflandre under the name V. hannai Deflandre, 1944 — from the same formation V. hortonii — appear to be merely teratological examples of that species. V. simplex Zhuze, 1949 is apparently distinct, while the other described species (V. ravnii Wetzel, 1940) is of doubtful status and may not even be a silicoflagellate. Dictyocha siderea Schulz, 1928 (Upper Cretaceous (Senonian) W. Prussia, together with its var. quadrata Schulz, 1928 from the same deposit, is plainly also a Vallacerta (Vallacerta siderea (Schulz) Bukry, 1981) and has been further discussed by Bukry (1985 pp. 131–132) who noted 4, 6, 7 and 8 sided forms of this species in addition to the usual pentagonal forms. Intact paired skeletons were not uncommon in his samples, including non-pentagonal morphs.

McCartney (in Lipps, 1993 pp. 148–149) characterises Vallacerta as “basal rings that have apical domes without portals or windows” and notes that “the apical structure, however, is different from that of any Tertiary silicoflagellate in having a solid domal plate without portals or windows. Typical late Cretaceous morphologies have a basal ring with five sides, although variants with more or fewer sides occur. Late Cretaceous Vallacerta also lack basal pikes. The group is known to occur in the Early Cretaceous, where seven sided forms with basal pikes appear to be prevalent.”

21. Hannaites

Hannaites was described as a new genus from Eocene fossils and was nomenclaturally treated as a name of an animal (Mandra, 1969 p. 2). A single species was reported—Hannaites quadria Y.T. Mandra, 1969.

The morphology is distinctive with a more or less squared basal ring and a fibuliform apical structure, clearly differing from Dictyocha in the absence of spines and in having the apices of the basal ring formed as rounded bosses with small projections. It has seldom been collected and is not even mentioned by McCartney (in Lipps, 1993).

22. Variramus

Variramus is the most recently described genus of silicoflagellate morphotaxa (McCartney et al., 1990) and was suggested as being the oldest known silicoflagellate (Lower Cretaceous) and potentially ancestral to other early silicoflagellates (Lyramula, Cornua, Vallacerta).

When it was described two nominal species were assigned to it, but the name used for the “type species” (Variramus aculeiferus (‘aculeifera’) was not validly published because of failure to comply with the strict letter of the ICBN (Berlin Code unchanged in the Tokyo Code 1994), Art. 33.2. This is because the authors (who were geologists and not botanists) cited inclusive pagination for the article in which the protologue of the basionym appeared rather than the precise page on which the name appeared as the Code demanded). Consequently the name of the nominal species was not published even though the generic name and the name of the other nominal species were published (see ICBN 2000 (St. Louis Code) Art. 32.1: ‘In order to be validly published, a name of a taxon must: [provisions (a)-(c) are met] and (d) comply with the special provisions of Art. 33-45 [. . .]’).

I had considered the desirability of validating the name as: Variramus aculeiferus. The other species was validly named as Variramus loperi McCartney et al., 1990. However the name of the former nominal species has not yet been validly published and has not received a date of publication. Under Art. 45.1, it seems to me, the name Variramus aculeiferus (Deflandre) McCartney ex [whoever] is still not validly published a decade after it appeared in print. In Index Nominum Algarum, the invalidity of the combination is affirmed. Since the present on-line publication does not count as an effective publication (ICBN 2000 Art. 29.1) the intended name will be published elsewhere.

McCartney (in Lipps, 1993 p. 149) writes: “The general morphology of the skeleton, which can include spines and pikes, is often similar to the apical structures of four- or more-sided Dictyocha, although more complex morphologies also occur and may be conspecific with the simpler skeletons. The genus is characterised by extreme variability, even by silicoflagellate standards; indeed, no two specimens of this group appear to be alike.”

23. Lyramula

Lyramula is an Upper Cretaceous siliceous microfossil which McCartney (in Lipps, 1993) calls “commonly considered a silicoflagellate because of its tubular construction and its resemblance to an incomplete basal ring.” Described by Hanna (1928) from two similar morphospecies from the same formation (Moreno Formation, Fresno County, California), skeletons consist of a horseshoe shaped tube with (L. furcula Hann, 1928) or without (L. simplex Hanna, 1928) a spine at the bend, the long skeletal members are termed ‘limbs’. A third nominal species was described from fossil deposits of unspecified age in Japan (L. tenuipertica Kokubo & Tsumura, 1963). A variety of the last was described from California and a variety of the ‘type species’ L. furcula Hanna, 1928 was described as L. furcula var. minor Deflandre, 1940, later raised to specific rank as L. minor (Deflandre) Deflandre, 1950. McCartney observes (in Lipps, 1993 p. 149): “Specimens with more than two limbs occur uncommonly and this may be aberrant. This group can be extremely common in siliceous sedimentary rocks of the later Cretaceous.” Lipps (1971) earlier suggested that Lyramula was in fact fused setae of a diatom, but this view is not generally shared following SEM examination of material. Lyramula furcula has also been described and illustrated by Locker (1996 p. 119 Plate 2). The Locker caption for his Plate 1: fig. 1 is mislabelled and the age of the material is stated as being Miocene-Pliocene, an age which must be considered suspect, as Lyramula is otherwise held to be of the Upper Cretaceous. Several other species have been described by Bukry (see Bukry, 1985 p. 131 for further information).

24. Arctyocha

Established in 1985 for two species, Arctyocha quadralta and an unnamed species (Bukry calls it ‘Arctyocha sp. A Plate 9.1, fig. 1–3’) this generic name was typified by the statement, on page 130, “Type species: Dictyocha quadralta Hanna, 1928 p. 261, pl. 41. fig. 3” but the formal recombination was made on page 134 with the words “Arctyocha quadralta (Hanna) Bukry, n. comb. Basionym: Dictyocha quadralta Hanna (1928) p. 261, pl. 41, fig. 3”. The description states that the skeleton lacks basal pikes but that “The basal ring outlines, typically, have four, five or six corners with moderate length spines. The interspine sides of the ring are typically indented but may be straight. Various patterns of bars and struts compose the apical structures which rise above the plane of the basal ring.” Its author speculates that “Arctyocha may have developed from Corbisema but it lacks basal pikes. Arctyocha is separated from Dictyocha by several structural and lineal barriers.” Whereas Dictyocha was described from the Miocene, Arctyocha is from the Late Cretaceous and it was thought that despite some superficial resemblance “there is no direct generic connection between Late Cretaceous specimens of Arctyocha and the more dominant and continuous members of Dictyocha found in the Middle Miocene and higher.” Some of the specimens tended to be pentagonal (Bukry's fig. 1, p. 132) but there is clearly some structural irregularity in the skeleton, resembling that manifested in the ‘deformed’ skeletons cultured by Van Valkenberg (1971a).”

Conclusion to Part III: A rationalisation of the classical taxonomy

A sound taxonomy of the silicoflagellates must rest upon a realistic appreciation of their biological characteristics as ‘axodine eukaryotes’ (to use Patterson's term) and of their nomenclatural governance by a single nomenclatural Code, the ICBN. The conventional taxonomies are of morphotaxa rather than of neotaxa. These taxonomies are now widely recognised as artificial and of doubtful value. The elucidation of the historical taxonomy and the associated nomenclature surveyed here is intended as groundwork for a more realistic taxonomy of the silicoflagellates. No advantage will be gained, for either morphotaxonomists or neotaxonomists, if either group continues to ignore the conceptual differences between the world of biology and the world of the stratigraphers and micropalaeontologists. The historical taxonomy is the third key player in the development of a rationalised approach to silicoflagellate identity. The present contribution now turns to a provisional reclassification of the silicoflagellates as two separate classifications of neotaxa and morphotaxa, governed by a single Code.

Part IV: A provisional dual classification of silicoflagellate neotaxa and morphotaxa

The classification of silicoflagellate neotaxa proposed and named here is very simple and minimalist. It recognises a single nominal genus Octactis with a single nominal species Octactis pulchra Schiller, 1925 with three morphs which are not accorded separate subspecific rank or subspecific names. If, in the future, somebody wishes to distinguish the fibuliform and / or speculiform morphs of Octactis pulchra from the ‘typical’ octactiform morph, they will have to propose entirely new species names for them; these species cannot be named as avowed recombinations ‘Octactis fibula’ (after Dictyocha fibula Ehrenberg) and ‘Octactis speculum’ because the ICBN prohibits the use of names for morphotaxa as if they were names for neotaxa. Use of similar-sounding names would run the risk of creating parahomonymy under Art. 61.5 (confusingly similar names).

In the case of the established silicoflagellate morphotaxa (which have been at various times been separated into some twenty-five nominal genera and five nominal families), a large number of redundant morphogeneric names can be sacrificed, as can all but one of the family names. The morphofamily Dictyochaceae Haeckel, 1887 (pro Dictyochida) — irrespective of the facts that Haeckel thought that the family was composed of phaeodarians — is an established name and is retained. The family name was re-established by Lemmermann (1901 p. 92, footnote) according to Silva (1980 p. 56). The later familial name Corbisemaceae Locker, 1974 (better corrected to Corbisemataceae, and including Corbisema, Cornua, Lyramula, Mesocena, and Phyllodictyocha) is herein considered a junior synonym of Dictyochaceae. This familial name was not mentioned by Silva (1980 p. 56) because his work was explicitly concerned with living algae. Cornuaceae Krieger, 1954 (see bibliography) is here also considered a junior synonym of Dictyochaceae Haeckel, 1887. Desikachary and Prema (1996) rearranged the genera into three orders (Vallacertales, Cornuales, Dictyochales) the first two containing, respectively, Vallacertaceae (with only one genus, Vallacerta) and, in the newly coined Cornuales, two families (Cornuaceae, with Cornua and Variramus; Lyramulaceae Tsumura, 1963 [Yokohama Municipal University Journal 146 p. 29], containing only Lyramula).

Mesocenaceae T.V. Desikachary & P. Prema, 1996 was circumscribed to include Mesocena, Bukryella, Paramesocena, and Septamesocena but (aside from Dictyochaceae Haeckel, 1887), I have chosen not to retain any of these familial names for the classification of morphotaxa given below.

Vallacertaceae Deflandre, 1950 (‘Vallacertidae’) — circumscribed by Locker (1974 p. 636) to include Octactis — is also synonymised with the morphofamily Dictyochaceae Haeckel, 1887. The nominal family ‘Distephanaceae fam. nov.’ introduced by Locker (1974 p. 637) was illegitimately named and it is taxonomically redundant in any case. As circumscribed by Locker it included four nominal genera: — (1) ‘Distephanus Stöhr, 1880’ (here synonymised with Dictyocha, a nomen palaeomorphosium based of a fossil type), (2) ‘Cannopilus Haeckel, 1887’ (the hypothetical phaeodarian, the name of which I propose to formally reject under ICBN Art. 56) another redundant taxonomic synonym of Dictyocha, and based on Dictyocha hemisphaerica, (3) ‘Deflandryocha Jerkovic, 1963’ (which is retained as a morphogenus within Dictyochaceae Haeckel, 1887), and (4) the little known ‘Nothyocha Deflandre, 1949’ (which is here retained provisionally as another morphogenus but is probably better regarded as a junior synonym of Dictyocha).

The disposition of the nominal genera of silicoflagellate morphotaxa within the sole recognised morphofamily is thus shown in Table I. The fifteen morphogenera retained are Dictyocha, Mesocena, Paramesocena, Paradictyocha, Corbisema, Phyllodictyocha, Clathropyxidella, Deflandryocha, Naviculopsis, Nothyocha, Cornua, Vallacerta, Hannaites, Lyramula, and Variramus. Six nominal genera are considered redundant: Bachmannocena is a junior heterotypic synonym of Mesocena and is accordingly rejected. Septamesocena is a superfluous renaming of Mesocena and is accordingly rejected as an illegitimate name. Distephanopsis is regarded as a junior heterotypic synonym of Dictyocha. Neonaviculopsis is regarded as a junior heterotypic synonym of Naviculopsis, as is Crassicorbisema. Caryocha is considered a heterotypic synonym of Clathropyxidella.

As for the higher classification, the placement of the silicoflagellates within the unranked and informal taxon of ‘axodines’ (which may probably end up at ordinal rank if and when a new classification of Eukaryotes is settled upon) is quite sufficient for the present. Although a typified class name for the silicoflagellates was adopted by Silva in 1980 (Dictyochophyceae) it now seems unnecessary to retain it. The question has been raised as to whether Art. 11.7 precludes the application of names of morphotaxa to living taxa. While it is felt to be possible to use a name of a morphotaxon for a living taxon (it is not illegal to do so), the practice is conducive to taxonomic confusion and should be avoided. The use of the name Dictyochophyceae is therefore rejected. We simply do not need a class name for silicoflagellate morphotaxa. The alternative descriptive class names Silicoflagellatae Lemmermann, 1901, Silicophyceae Rothmaler, 1951, and Silicoflagellatophyceae Glezer, 1966 are all rejected as unnecessary at present. One named morphofamily of silicoflagellates (not three or four) is quite sufficient at that rank.

The ordinal name of the silicoflagellate morphotaxon Dictyochales Haeckel, 1894 (‘Dictyocheae’ see Silva (1980 p. 56) is unnecessary, as is the ordinal name Cornuales T.V. Desikachary & P. Prema 1996, with which it is here synonymised. The ordinal name Vallacertales (Vallacertales Glezer, 1970 p. 234) is also unnecessary for the purposes of the classification adopted here. The adoption of an ordinal name for a presumptive order of non-fossil silicoflagellate neotaxa might be regarded as premature, given that only one putative species has been formally described and named. However a similar situation arises with the familial name for the neotaxon and it will be best if all the names at family, order and class rank are published simultaneously for the avoidance of future error. We are currently without any validly published family name for the silicoflagellate neotaxa and such a name is certainly necessary. That name will be provided separately (in view of ICBN 2000 Art 29.1): Octactinaceae Parkinson, 2002 (Type: Octactis Schiller, 1925 p. 66).

In the classification adopted here there is no need for infraspecific taxa at any rank (subspecies, variety, forma etc.). Consequently their nomenclature is rejected as redundant. The use of vernacular terminology for the morphs of silicoflagellate taxa is recommended, and the redundant names can be left to sink into well-deserved obscurity.

No attempt is made here to provide recombinations for all the nominal species under the generic names which are accepted as ‘validly published’ ‘legitimate’ and ‘correct’ in a taxonomic sense. Such an attempt was made by Alfred A. Loeblich et al. in their Annotated index of fossil and recent silicoflagellates and ebridians with descriptions and illustrations of validly proposed taxa (1968) but there seems little point in adding to the heavy burden of the names of taxa. The vast majority of these names and the associated taxonomic concepts are probably worthless, but are likely to continue to frustrate and puzzle palaeontologists, geologists and deep-sea drillers.


Neotaxa of silicoflagellates

Unranked ‘axodines’

Note: This unranked taxon includes four nominal families: (‘zoologically’ they are the Actinophryidae, Pedinellidae, Rhizochromulinidae and Dictyochidae of Patterson). The last named, however, is essentially a family of morphotaxa, with a name misapplied to neotaxa. In botanical nomenclature the four nominal families of axodines are equivalent to the Actinophryaceae Claus, 1874 (a name which does not yet exist in botanical nomenclature), Pedinellaceae Pascher, 1910, Rhizochromulinaceae D.J. Hibberd & M.J. Chretiennot, 1970, and Dictyochaceae Haeckel, 1887 (this last being a morphotaxon for which it is intended to substitute the name Octactinaceae, as above). Patterson's classification of what he called ‘Silicoflagellata’ (see is rejected because it does not recognise the distinction between neotaxa and morphotaxa.

Family Octactinaceae Parkinson, (to be published separately)
T: Octactis Schiller, 1925 p. 66. The name of the family is formed from the genitive octactinos and should spelled Octactinaceae rather than ‘Octactidaceae’. Note: the living family corresponding to the morphofamily Dictyochaceae Haeckel, 1887.)
Genus: Octactis Schiller, 1925
T: Octactis pulchra Schiller, 1925 p. 66. Note: the living genus corresponding to the morphogenus Paradictyocha Frenguelli, 1940.
Species: Octactis pulchra Schiller, 1925 p. 66. Note: the living genus corresponding to the morphospecies Paradictyocha polyactis (Ehrenberg) Frenguelli, 1940


Morphotaxa of silicoflagellates

Note: This ranked taxon includes only one morphofamily, Dictyochaceae Haeckel, 1887.

Morphofamily Dictyochaceae Haeckel, 1887
T: Dictyocha Ehrenberg, 1839. Note: morphotaxon corresponding to the living Octactinaceae (to be published)
Included nominal genera:
Dictyocha Ehrenberg, 1837
T: Dictyocha fibula Ehrenberg, 1839
Paradictyocha Frenguelli, 1940
T: Paradictyocha polyactis (Ehrenberg) Frenguelli, 1940. Note: the fossil state of Octactis pulchra Schiller, 1925.
Naviculopsis Frenguelli, 1940
T: Naviculopsis biapiculata (Lemmermann) Frenguelli, 1940
Corbisema Hanna, 1928
T: Corbisema geometrica Hanna, 1928
Lyramula Hanna, 1928
T: Lyramula furcula Hanna, 1928
Cornua Schulz, 1928
T: Cornua trifurcata Schulz, 1928
Vallacerta Hanna, 1928
T: Vallacerta hortonii Hanna, 1928
Hannaites Mandra, 1969
T: Hannaites quadria Mandra, 1969
Variramus McCartney et al., 1990
T: Cornua aculeifera Deflandre, 1950 [combination intended ‘Variramus aculeiferus’ has not been validly published under ICBN]
Paramesocena Locker & Martini, 1985
T: Paramesocena apiculata (Lemmermann) Locker & Martini, 1985 [genus should probably be eliminated as a junior synonym of Mesocena]
Mesocena Ehrenberg, 1843
T: Mesocena elliptica (Ehrenberg) Ehrenberg, 1843
Clathropyxidella Deflandre, 1938
T. Clathropyxidella similis Deflandre, 1938
Phyllodictyocha Deflandre, 1947
T: Phyllodictyocha recta (Schulz) Deflandre, 1947
Deflandryocha Jerkovic, 1963
T: Deflandryocha cymbiformis Jerkovic, 1963
Nothyocha Deflandre, 1949
T: Nothyocha insolita Deflandre, 1949
Arctyocha D. Bukry, 1985
T: Arctyocha quadralta (Hanna) Bukry, 1985
Eunaviculopsis H. Y. Ling, 1977
T: Eunaviculopsis navicula (Ehrenberg) H. Y. Ling (name not validly published)


Barrie, F.R. and Greuter, W. 1999.
XVI International Botanical Congress: preliminary mail vote and report of Congress action on nomenclature proposals. Taxon 49 (1999): 771–784.
Bukry, D. 1985.
Correlation of late Cretaceous arctic silicoflagellates from Alpha Ridge Paper - Geological Survey of Canada. 84(22): 125–135.
Bukry, D. 1987.
Eocene siliceous and calcareous phytoplankton, Deep Sea Drilling Project Leg 95. Initial Reports of the Deep Sea Drilling Project 95: 395–415.
Bukry, D. and Monechi, S. 1985.
Late Cenozoic silicoflagellates from the northwest Pacific Deep Sea Drilling Project Leg 86: Paleotemperature trend and texture classification. Initial Reports of the Deep Sea Drilling Project 86: 367–397.
Daugbjerg, N. 1996.
Mesopedinella arctica (Pedinellales). II. Phylogeny of Mesopedinella, including a cladistic analysis of Dictyochophyceae. Phycologia 35: 563–568.
Deflandre, G. 1940.
L'origine phylogenetique des Lyramula et l'evolution des Silicoflagellidées. Comptes Rendus Hebdomadaires des S&eactte;ances de l'Acad&eactte;mie des Sciences [Paris] 211: 508–510 fig. 1–12.
Deflandre, G. 1947.
Phyllodictyocha gen. nov. Silicoflagellidées et formes affines des Miocene d'Hongrie. Bulletin de la Société de Botanique de France 93: 335–337, figs. 1–5.
Deflandre, G. 1949.
Nothyocha gen. nov. Silicoflagellidé d'un type nouveau et aberrant, d'allure radiolarienne. Comptes Rendus Hebdomadaires des S&eactte;ances de l'Acad&eactte;mie des Sciences [Paris] 229: 673–674, figs. 1–4.
de Queiroz, K. 2000.
The definition of taxon names: a reply to Stuessy. Taxon 49: 533–536.
Desikachary, T.V. and Prema, P. 1996.
Silicoflagellates (Dictyochophyceae). Stuttgart: J. Cramer in der Gebrüder Borntraeger Verlagsbuchhandlung Berlin, 1996. vi, 298, [+ 20] pp. 83 pl. (Bibliotheca Phycologica 100)
Dumitrica, P. 1973.
Palaeocene, late Oligocene and post-Oligocene silicoflagellates in southwestern Pacific sediments cores in on DSDP Leg 21. Initial Reports of the Deep Sea Drilling Project 21: 837–883.
Dumitrica, P. 1978.
Badenian silicoflagellates from central Paratethys. In Brestenska, E. (ed.) Chronostratigraphie und Neostratotypen 6: 207–229, incl. 5 pl.
Ehrenberg, C.G. 1837.
[Written in the 3rd person by the Secretary of the academy] Hierauf legst Hr. Ehrenberg eine briefliche Nachricht des Hrn Agassiz in Neuchatel über den ebenfalls aus mikroskopischen Kiesel-Organismen gebildeten Polirschiefer von Oran in Afrika.12 Berichte ... der Königlichen Akademie der Wissenschaften zu Berlin ... 1837: 59–61.
Ehrenberg, C.G. 1838.
Über das Massenverhältniss in jetzt lebenden Kiesel-Infusorien und über ein neues Infusorien-Conglomerat als Polirschiefer von Jastrabe in Ungarn. Abhandlungen der Königlichen Akademie der Wissenschaften zu Berlin ... 1836: 109–135 (issued 1838).
Ehrenberg, C.G. 1838.
Die Infusionsthierschen als volkommen Organismen. Ein Blick in das tiefere organische leben der Natur. Leipzig: L. Voss, 1838. Xviii, 457 pp. LXIV col. pl.
Ehrenberg, C.G. 1839.
Über die Bildung der Kreidefelsen und des Kreidemergels durch unsichtbare Organismen.. Abhandlungen der Königlichen Akademie der Wissenschaften zu Berlin ... 1838: 59–147, pl. 1–4, tab. 1–2 (separate 1839, volume issued 1839)
Ehrenberg, C.G. 1840a.
Über noch jetzt zahlreich lebende Thierarten der Kreidebildung und den Organismus der Polythalamien. Abhandlungen der Königlichen Akademie der Wissenschaften zu Berlin ... 1839: 81–174 Pl. 1–4 (separatum issued 1840 with a variant title Über noch zahlreich jetzt lebende Thierarten der Kreidebildung (Berlin, 94 pp. 3 pl.; volume issued 1840).
Ehrenberg, C.G. 1840b.
[Untitled work, referred to in the Namen-Register on p. 261 of the volume as ‘Characteristik v. 274 neuen Areten von Infusorien’. 13] Berichte ... der Königlichen Akademie der Wissenschaften zu Berlin ... 1840: 197–219 .
Ehrenberg, C.G. 1843a.
[Untitled and written in the 3rd person by the Secretary] Hr Ehrenberg las den ersten Thiel seiner Beobachtungen über die Verbreitung des jetzt wirkenden kleinsten organischen Lebens in Asien, Australien und Afrika und über die vorherrschende Bildung auch des Oolithkalkes der Juraformation aus kleinen polythalamischen Thieren. 14 Berichte ... der Königlichen Akademie der Wissenschaften zu Berlin ... 1843: 100–106
Ehrenberg, C.G. 1843b.
Verbreitung und Einfluss des mikroskopischen Lebens in Süd- und Nord-Amerika. Abhandlungen der Königlichen Akademie der Wissenschaften zu Berlin ... 1841: 291–446 Pl. 1–4.
Ehrenberg, C.G. 1844a.
[Untitled and written in the 3rd person by the Secretary] Hr. Ehrenberg las eine Mittheilungen über 2 neue Lager von Gebirgsmassen aus Infusorien als Meeres-Absatz in Nord-Amerika unde eine Vergleichung derselben mit den organischen Kreide-Gebilden in Europa und Africa. 15 Berichte ... der Königlichen Akademie der Wissenschaften zu Berlin ... 1844: 57–97 (Locker dates this work to 1845).
Ehrenberg, C.G. 1847.
[Untitled and written in the 3rd person by the Secretary] Hr. Ehrenberg las über die mikroskopischen kieselschaligen Polycystinen als mächtige Gebirgsmasse von Barbados und über das Verhältniss der aus mehr als 300 neuen Arten bestehenden ganz eigenthümlichen Formengruppe jener Felsmasse zu den jetzt lebenden Thieren und zur Kreidebildung. Eine neue Anregung zur Erforschung des Erdlebens. 16 Berichte ... der Königlichen Akademie der Wissenschaften zu Berlin ... 1847: 40–60.
Ehrenberg, C.G. 1854.
Mikrogeologie, das Erden und Felsen schaffende Wirken des unsichtbar kleinen selbständinges Lebens auf der Erde. Leipzig: Leopold Voss. 374 pp., 40 pl. (Nov. 1854)
Engelmann, W. 1846.
Bibliotheca historico-naturalis: Verzeichniss der Bücher über Naturgeschichte welches in Deutschland, Scandinavien, Holland, England, Frankreich, Italien und Spanien in den Jahren 1700–1846 erschienen sind. Leipzig: W. Engelmann, 1846. 1v.
Fensome, R.A. et al. 1998.
Typification and the names of fossil plants — a review and discussion. Taxon 47: 695–702.
Frenguelli, J. 1940.
Consideraciones sobre los silicoflagelados fosiles. Revista del Museo de La Plata, ser. 2, 2: 37–112.
Gemeinhardt, K. 1930.
Silicoflagellatae in Rabenhorst, L., Kryptogamenflora von Deutschland, Osterreich und der Schweiz 10: 1–87. Leipzig: Akademische Verlagsgesellschaft.
Glezer, Z.I. 1970.
Silicoflagellatophyceae in M. M. Gollerbakh, M.M., Cryptogamic plants of the USSR, v. 7, pp. 1–363. (translation from the Russian of Z.I. Glezer (1966) Flora Sporovykh Rasteniy SSSR (Flora Plantarum Cryptogammarum URSS, 7) by Israel Program for Scientific Translations Ltd, Jerusalem, 1970.
Gran, H.H. and Braarud, T. 1935.
A quantitative study of the phytoplankton in the Bay of Fundy and the Gulf of Maine (including observations of hydrography, chemistry and turbidity). Journal of the Biological Board of Canada 1: 279–467, figs. 1–69, Tab. 1–60.
Greuter, W. et al. 2000.
International Code of Botanical Nomenclature (Saint Louis Code). Koeltz Scientific Books, Konigstein. 474 p.
Haeckel, E.H.P.A. 1862.
Die Radiolarien (Rhizopoda radiaria). Eine Monographie. Berlin: Georg Reimer, 572 pp. 35 pls.
Haeckel, E.H.P.A. 1879.
Ueber die Phaeodarien, eine neue Gruppe kieselschaliger mariner Rhizopoden. Sitzungsberichte der Jenaischer Gesellschaft fur Medicin und Naturwissenschaft. Sitzung von 12 Dec. 1879.
Haeckel, E.H.P.A. 1882.
Entwurf eines Radiolarien-Systems auf Grund von Studien der Challenger-Radiolarien. Jenaische Zeitschrift für Naturwissenschaft 15(n.f.v.8): 418–472.
Haeckel, E.H.P.A. 1887.
Report on the radiolaria collected by H.M.S. Challenger during the years 1873–1876. Challenger Expedition Reports 18. 1803 pp., 140 pl.
Haeckel, E.H.P.A. 1899.
Die Welträthsel: Gemeinverständliche Studien über monistiche Philosophie. Bonn: E.Strauss, 1899. viii, 473 pp.
Hanna, G.D. 1928.
Silicoflagellates from the Cretaceous of California. Journal of Palaeontology 1: 259–263, pl. 41.
Hay, W.H. 1977.
Calcareous nannofossils. In Ramsay, A.T.S. (ed.) Oceanic Micropalaeontology v. 2. Pp. 1055–1200. London. Academic Press.
Henriksen, P., Knipschildt, F., Moestrup, Ø., and Thomsen, H.A. 1993.
Autecology, life history and toxicology of the silicoflagellate Dictyocha speculum (Silicoflagellata, Dictyochophyceae). Phycologia 32: 29–39.
Hertwig, R. 1879.
Der Organismus der radiolarien. Jenaischer Denkschriften 2: 129–277, Taf. vi-xvi
Hsin Yi Ling 1977.
Late Cenozoic silicoflagellates and ebridians. In: Proceedings of the First International Congress on Pacific Neogene Stratigraphy, Tokyo, 1976. Tokyo: Science Council of Japan, 1977. vi, 433 pp.
Hsin Yi Ling 1981.
Crassicorbisema, a new silicoflagellate genus, from the southern oceans and Palaeogene silicoflagellate zonation. Transactions and Proceedings of the Palaeontological Society of Japan n.s. 121: 1–13.
International Commission on Zoological Nomenclature 1926
International rules of Zoological nomenclature. [Reprint by the Biological Society of Washington] in Proceedings of the Biological Society of Washington 39: 75–104.
International Congress of Zoology 1964.
International Code of Zoological Nomenclature adopted by the XV International Congress of Zoology. London: International Trust for Zoological Nomenclature, 1964. xx, 176 pp.
Jørgensen, P. 2000.
Names are defined, but not as taxa. Taxon 49: 779.
Krieger, W. 1954.
VII. Abteilung: Chrysophyta. In Melchior, H. and Werdemann, E. (eds.) A. Engler's Syllabus der Pflanzenfamilien . . . 12. Aufl. I Band. Berlin: Borntraeger, 1954. Pp. 73–85.
Lee, J.J., Leedale, G.F., and Bradbury, P. 2002 ('2000').
An illustrated guide to the protozoa: organisms traditionally referred to as protozoa, or newly discovered groups. 2nd ed. Lawrence, Kan.: Society of Protozoologists. 2v.
Lemmermann, E. 1901.
Beiträge zur Kenntniss der Planktonalgen. CXII Notizen über einige Schewbealgen. XIII. Das Phytoplankton des Ryck und des Greifswalder Bodens. Bericht der Deutschen Botanischen Gesellschaft 19: 85–95.
Lipps, J.H. 1970.
Ecology and evolution of silicoflagellates. Proceedings of the First North American Paleontological Convention. Pp. 965–993.
Lipps, J.H. and McCartney, K. 1993.
Chrysophytes. in Lipps, J.H. (ed.) Fossil Prokaryotes and Protists. Boston: Blackwell Scientific Publications, 1993. ix, 342 pp.
Locker, S. 1974.
Revision der Silicoflagellaten aus der Mikrogeologischen Sammlung von C.G. Ehrenberg. Eclogae Geologicae Helveticae 67(3): 631–646, Tab. 1-IV
Locker, S. 1975.
Dictyocha varia sp. n., eine miozane Silicoflagellaten-Art mit kompliziertem Variationsmodus. Zeitschrift für Geologische Wissenschaften 3(1975): 99–103.
Locker, S. 1995.
Silicoflagellates, ebridians and actiniscidians from Pliocene and Quaternary sediments off southern Chile, ODP Leg 141. Proceedings of the Ocean Drilling Program, Scientific Results 141: 223–233.
Locker, S. 1996.
Cenozoic siliceous flagellates from the Fram Strait and the East Greenland margin; biostratigraphic and palaeoceanographic results. Proceedings of the Ocean Drilling Program, Scientific Results 51: 101–124.
Locker, S. and Martini, E. 1985.
Silicoflagellates and some sponge spicules from the southwest Pacific, Deep Sea Drilling Project, Leg 90. Deep Sea Drilling Project, Leg 90, Scientific Results 90: 909.
Locker, S. and Martini, E. 1986.
Ebridians and actiniscidians from the southwest Pacific. Initial Reports of the Deep Sea Drilling Project 90: 939–951.
Locker, S. and Martini, E. 1987.
Silicoflagellaten aus einigen russischen Paläologen-Vorkommen. Senkenbergiana Lethaea 68: 21–67.
Loeblich III, A.R. et al. 1968.
Annotated Index of Fossil and Recent Silicoflagellates and Ebridians with descriptions and illustration of validly published Taxa. Boulder: Geological Society of America. 319p. (Memoir 106)
McCartney, K. and Harwood, D.M. 1992.
Silicoflagellates from Leg 120 on the Kerguelen Plateau, southeast Indian Ocean. Proceedings of the Ocean Drilling Program, Scientific Results 120: 811–831.
McCartney, K. and Loper, D.E. 1989.
Optimized skeletal morphologies of silicoflagellate genera Dictyocha and Distephanus. Palaeobiology 15(3): 283–298.
McCartney, K. and Loper, D.E. 1992.
Optimal models of skeletal morphology for the silicoflagellate genus Corbisema. Micropalaeontology 38(1): 87–93.
McCartney, K. et al. 1990.
Enigmatic lower Albian silicoflagellates from ODP Site 693: progenitors of the order Silicoflagellata? Proceedings of the Ocean Drilling Program, Scientific Results 113: 427–442.
Mandra, Y.T. 1969.
A new genus of Silicoflagellata from an Eocene South Atlantic Deep-sea Core (Protozoa: Mastigophora). Occasional Papers of the California Academy of Sciences 77. 7 pp.
Moestrup, Ø. and Thomsen, H.A. 1990.
Dictyocha speculum (Silicoflagellata, Dictyochophyceae), studies on armoured and unarmoured stages. Kongelige Danske Videnskabernes Selskab, Biologiske Skrifter. 37: 1–57.
Orchard, A.E. et al., 1996.
Harmonised bionomenclature — a recipe for disharmony. Taxon 45: 287–290.
Patterson, D.J. 1994.
Protozoa: evolution and systematics. Pp. 1–14, in Hausmann, K. and Hulsmann, N. (eds.) Progress in Protozoology. Berlin, G. Fischer Verlag.
Patterson, D.J. 1999.
The Diversity of eukaryotes. American Naturalist 154: pp. 96–124.
Perch-Nielsen, K. 1975.
Late Cretaceous to Pleistocene silicoflagellates from the southern Southwest Pacific, DSDP Leg 29. Initial Reports of the Deep Sea Drilling Project 29: 677–721.
Ride, W.D.L. et al. 1985.
International Code of Zoological nomenclature. 3rd ed. London.
Schiller, J. 1925.
Die planktonischen Vegetation des adriatischen Meers. B: Chrysomonadina, Heterokontae, Cryptomonadina, Eugleninae, Volvocales 1. Systematischer Teil. Archiv für Protistenkunde 53: 59–123, pls. 3–6.
Schulz, P. 1928.
Beiträge zur kenntnis fossiler und rezenter Silicoflagellaten. Botanisches Archiv 21: 225–292.
Silva, P.C. 1980.
Names of classes and families of living algae. Regnum Vegetabile 103. 156 pp.
Stöhr, E 1880.
Die Radiolarienfauna der tripol von Grotte, Provinz Girgenti in Sicilien. Palaeontographica 26: 69–124 Pl. 1–7.
Stuessy, T.F. 2000.
Taxon names are not defined. Taxon 49: 231–233.
Van Valkenberg, S.D. 1971a.
Observations on the fine structure of Dictyocha fibula Ehrenberg. I. The skeleton. Journal of Phycology 7: 113–118.
Van Valkenberg, S.D. 1971b.
Observations on the fine structure of Dictyocha fibula Ehrenberg. II. The protoplast. Journal of Phycology 7: 118–132.
Van Valkenberg, S.D. 1980.
Silicoflagellates. In Cox, E.R. (ed.) Phytoflagellates. Pp. 335–350. Elsevier North-Holland.
Van Valkenberg S.D. and Norris, R.E. 1970.
The growth and morphology of the silicoflagellate Dictyocha fibula Ehrenberg in culture. Journal of Phycology 6: 48–54.


1. See Haeckel's Riddle of the universe at the close of the nineteenth century (1900) a translation by J. McCabe of Welträthsel (1899), an exposition of the monist philosophy.

2. This is formally entitled International Code of Botanical Nomenclature (St. Louis Code) adopted by the Sixteenth International Congress St. Louis, Missouri, July-August 1999This replaced the ‘Tokyo Code’ of 1994 (hereinafter ICBN, 1994) and its predecessors (Montreal, 9th Congress, 1961; Edinburgh, 10th, 1966; Seattle, 11th, 1972; Leningrad, 12th, 1978; Sydney, 13th, 1983; Berlin, 14th, 1988) several of these editions being cited hereunder.

3. ‘Fibulid’ (or ‘fibuloid’ or ‘fibuliform’) is used for skeletons with a bridge on the long axis of the skeleton. ‘Asperid’ (asperoid) is used for skeletons with a bridge on the short axis.

4. The Code used by Locker and Martini in this 1985 paper — and cited in the bibliography on page 911 — is in fact the 1975 Leningrad Code, although that had already been superseded by the Sydney Code, 1983.

5. McCartney also admits (1993 pp. 149–150), more cautiously but with implicit expectation: “Silicoflagellates are not known to reproduce sexually, but the great stability of the fossil and recent forms, given their known variability, indicates that silicoflagellates possess temporary phases of sexual reproduction.”

6. W. Engelmann in Bibliotheca historico-Naturalis (1846) p. 589, lists the paper with the title as in the journal and dated 1838, as is the title page of the journal itself. But the British Museum Catalogue p. 515 states that the paper was published separately in 1837 with a distinct title Die fossilen Infusorien und die lebendige Dammerde. (Berlin, 27 pp., table, 2 plates. No copy of this publication has been traced but it was probably repaginated, not recast. I am indebted to Paul Silva for this information.

7. The title page of the “Physikalische Abhandlungen” section of the journal is dated 1840 but, according to the British Museum Catalogue (p. 515) and to Engelmann (p. 589) the separate is from 1839.

8. Arthrodesmus truncatus Ehrenberg Infus. p. 152 (1838), said by Ralfs, The British Desmidieae, Appendix, to be ‘probably a Xanthidium or Staurastrum’.

9. The date of this third paper has been disputed. Loeblich et al. (1968) state that the Abhandlungen as a whole was issued in 1840 but that Ehrenberg's paper had already appeared as a separatum in 1839, but Locker dates the names used in the paper to 1840 from the date of the issue of the volume, ignoring the separatum. I accept that the prior issue of the separatum effectively published the names under the ICBN and accordingly accept the date of 1839 given by Loeblich et al. as the date of valid publication of these names.

10. ICBN 2000, Art. 36.3: “In order to be validly published, a name of a new taxon of fossil plants published on or after 1 January 1996 must be accompanied by a Latin or English description or diagnosis or by a reference to a previously and effectively published Latin or English description or diagnosis.”

11. Because the editions of the ICZN are obscure it may be helpful to note that the first edition of the ICZN was published in 1902 as ‘Règles de la nomenclature zoologique adopté par le Vème Congrès de Zoologie’ in Verhandlung des V Internationalen Zoologen-Kongresses zu Berlin, [12–16 August 1901] pp. 938–972 (Jena: G. Fischer, 1902) and reprinted in Paris in 1904 as ‘Règles internationales de la nomenclature zoologique adopté par les Congrès Internationaux de Zoologie, 1904’ (Paris: F.R. de Rudeval, 1904). This was reprinted in 1926 in Proceedings of the Biological Society of Washington v. 39 (1926) pp. 75–104. A second edition was published in 1964, a third (by Ride, in 1985) and a fourth in 1999. See bibliography under International Commission on Zoological Nomenclature, and under International Congress of Zoology.

12. The article has no title, nor a table of contents nor index.

13. The Secretary of the academy wrote “Hr. Ehrenberg legte hierauf 274 Blätter von ihn selbst ausgefürter Zeichnungen von ebenso vielen Arten in dem 1838 erschienenen grösseren Infusorienwerke noch nicht abgebildeter Infusorien vor, und sprach über die auffallend rasche Entwicklung dieser Kenntnisse.” In the Sachs-Register (p. 265) the paper is cited as “Verzeichn. Von 274 Arten in d. grossen Werk v. Ehrenberg noch nicht abgebildeter Inf.” The paper was read at the meeting of 9 November 1840 but probably published in 1841, although a retitled separatum dated 1840 has the title Kurze Nachricht über 274 seit dem Abschluss der Tafeln des grosseren Infusorienwerkes neu beobachtete Infusorien-Arten (Berlin, 24 pp.) This work was dated as 1841 by Locker, who gives the title of the article as ‘Charakteristik von 274 neuen Arten von Infusorien’

14. This paper was read on 30 March 1843.

15. Read at the meeting of 22 February 1844 and was published separately in 1844 as Ueber die Lager von Gebirgsmassen u. Infusorien als Meeres-Absatz in Nordamerika u. deren Vergleich. mit den organ. Kreidegebilden in Europa u. Afrika (Engelmann p. 563). Locker (1974) gives the date as 1845. I thank Paul Silva for these clarifications.

16. Read on 11 February 1847 and probably published within three months.