Bryozoans are colony-forming aquatic invertebrates. Ryland 〚23〛 and McKinney and Jackson 〚11〛 provide good general accounts of the Bryozoa. Recent molecular studies have shown the phylum to be allied to brachiopods, molluscs and annelids within the Lophotrochozoa 〚6〛. Some 6000 species of bryozoans are alive at the present-day, predominantly marine and ranging from the intertidal to the abyss in bathymetric distribution although most inhabit the continental shelf. The secretion of a calcareous skeleton in the overwhelming majority of marine species explains the rich fossil record of the Bryozoa that commences in the Ordovician. Bryozoans are the dominant skeletal epifauna over some parts of the continental shelf where they can be important in structuring benthic communities; for example, the branching colonies of arborescent bryozoans provide habitats for diverse sessile and mobile animals. Similarly dense communities of bryozoans in the geological past are responsible for the bryozoan limestones that are common in the geological record. They include some reef limestones with bryozoans as the principal reef builders.

The individual modules forming a bryozoan colony are known as zooids. Although within a colony the zooids are all genetically identical, they are not necessarily identical in morphology. This is because partitioning of biological functions between zooids often results in zooidal polymorphism. Feeding zooids (autozooids), larval brooding zooids (gonozooids) and defensive zooids (avicularia) are among the variety of polymorphs that occur in bryozoans. Each autozooid has a lophophore consisting of a ring of tentacles surrounding a mouth. Bryozoans are suspension feeders, using cilia on the tentacles to create a current of water that drives food particles (mainly phytoplankton) towards the mouth. The lophophore emerges through an opening, the orifice or aperture, in the frontal surface of the zooidal skeleton. When danger threatens the lophophore can be retracted rapidly into the safety afforded by the tubular or box-shaped zooid. Colonies grow principally by budding new zooids, an asexual process. Sexual reproduction, and also dispersal, is usually accomplished via a short-lived larval stage. After settling on a hard or firm substratum, the larva metamorphoses into the founder zooid (ancestrula) of the colony that then begins budding new zooids. The pattern of zooidal budding is important in determining colony-form, which varies considerably between species and sometimes within a single species. Some bryozoan colonies take the form of thin encrusting sheets or runners or large multilamellar mounds, whereas others grow upright into bushy or mesh-like shapes, and yet others develop cap-shaped, free-living colonies.

There is little overlap taxonomically between the orders of bryozoans that dominated Palaeozoic faunas and those of the post-Palaeozoic. D’Orbigny was mostly concerned with the latter. The great majority of taxa he described belonged to two orders now known as the Cyclostomata and Cheilostomata but termed ‘Centrifuginés’ and ‘Cellulinés’ respectively by d’Orbigny (note that d’Orbigny included some genera belonging to extinct stenolaemate bryozoan orders within the ‘Centrifuginés’). Zooids in the first order are usually tubular zooids whereas those in the second are typically box-shaped. Cyclostomes have an Ordovician to Recent range, cheilostomes Jurassic to Recent although none older than Cretaceous were known until the 1970s. Some of the richest post-Palaeozoic bryozoan localities in the world are to be found in France, and both the Jurassic and Cretaceous of France contain a diversity of bryozoans unrivalled elsewhere for rocks of these ages. D’Orbigny had access to and described fossil bryozoans from such classic French Jurassic localities as Ranville and Luc-sur-Mer (Calvados), and Cretaceous localities, including Le Havre and Fécamp (Seine-Maritime), Le Mans (Sarthe), Rouen and Saintes (Charente Maritime), Villedieu and Tours (Loire), and Meudon (Hauts-de-Seine).

Contemporary bryozoan systematics relies overwhelmingly on skeletal characters. Hence there is no conflict between the taxonomies employed by neontologists and palaeontologists. Both colonial and zooidal characters are utilised, the former tending to be given more emphasis in cyclostomes, the latter in cheilostomes with their greater levels of zooidal complexity and polymorphism. In spite of his grounding in micropalaeontology and observation and description of zooid-level characters in bryozoans, d’Orbigny emphasised colony-level characters when classifying bryozoans. Therefore, his classification of cheilostomes differs from modern schemes more than does his classification of cyclostomes.

The earliest evidence of Alcide d’Orbigny’s contact with bryozoans comes from a letter written by his father Charles d’Orbigny in 1819 which is translated in Heron-Allen 〚7, p. 7〛 as follows: “One day he 〚Alcide〛 brought me some Polyzoa 〚Bryozoa〛 which he had just gathered on the rocks at Marsilly at very low tide; we placed them in water with the idea of seeing one of the Polyps develop”. The discovery of foraminifera associated with these bryozoans led father and son to the false conclusion that they were predators of bryozoans.

D’Orbigny’s research on bryozoans began in earnest when describing Recent species collected during his exploration of South America. During the last decade of his life, as his interests became focused on systematic palaeontology and stratigraphy, d’Orbigny’s interests in bryozoans turned towards the rich Cretaceous faunas of France. Bryozoans form an appreciable component of the huge collection of Mesozoic and Cainozoic invertebrates that he assembled.

All of d’Orbigny’s research on bryozoans was published in the 13-year period between 1841 and 1854. However, the exact date of publication of individual works, however, is often difficult to establish. This is particularly unfortunate in view of the issues of taxonomic priority raised by his publications. Sometimes the date given on the title page is incorrect. D’Orbigny’s custom of appending the year in which he wrote a taxonomic description to the taxonomic name adds further to the confusion as these dates invariably antedate the year of publication which is when the taxon becomes ‘available’ in a nomenclatural sense.

D’Orbigny’s schemes of bryozoan classification evolved during the course of his research. The first livraison of the Terrains crétacés 〚17〛 tabulates (p. 22) a tripartite division of bryozoans into three orders: ‘Cellulinés’, ‘Tubulinés’ and ‘Foraminés’. These orders were based on the form of the zooecia, respectively cell-like, tubular and with holes. Colony-form was then used to distinguish 13 families, five belonging to each of the ‘Cellulinés’ and ‘Tubulinés’ and three to the ‘Foraminés’ (Table 1). That this early scheme was soon abandoned is evident from d’Orbigny’s Recherches… paper 〚18〛, where the ‘Classe Mollusques bryozoaires’ is reduced to two orders, ‘Cellulinés’ and ‘Centrifuginés’, the ‘Tubulinés’ and the ‘Foraminés’ becoming divisions of the ‘Centrifuginés’. ‘Cellulinés’ and ‘Centrifuginés’ respectively correspond very closely with Busk’s orders Cheilostomata and Cyclostomata, which are used today in bryozoan classification. Among the slight differences are d’Orbigny’s inclusion of the cheilostome family Myriozoumidae (= Myriaporidae) in the Centrifuginés and the boring ctenostome Terebripora in the Cellulinés.

D’Orbigny 〚17〛 and 〚18〛 distinguished two orders of Cellulinés – ‘Radicellés’ and ‘Empâtés’. The former included erect bushy colonies that were basally rooted, as well as some encrusting runner-like forms that may be stoloniform (e.g., Aetea). The assigned families, though often including a heterogeneous mix of genera by today’s standards, nevertheless generally correspond to modern families except where the name of the type genus has been synonymised. Hence the Acamarchisidae corresponds to the Bugulidae (Acamarchis is a junior objective synonym of Bugula), and the Catenaridae to the Catenicellidae (Catenaria is junior homonym of a cestode genus and the name was modified to Catenicella by de Blainville). D’Orbigny included in the families Flustridae and Cellaridées (i.e. Cellariidae) several unrelated genera that today comprise the types of other families; otherwise these family names are still in use. The Electrinidae (now Electridae) included a mix of lightly-calcified encrusting and erect rooted taxa, the latter now ironically included in his family Candidae. D’Orbigny 〚17〛 named the Candidae in the first livraison (1851) of the Terrains crétacés (Table 1), giving a brief diagnosis, but he did not subsequently use it, submerging it in the Electrinidae (Table 2).

Of the twelve families included in the ‘Empâtés’ – encrusting, free-living and fixed-erect forms – only one, the ‘Porinidae’, is in current use. The others are based on genera that were subsequently abandoned or are now nomina oblita. There is one potential exception, however, viz. the Steginoporidae. This was based on a cribrimorph (acanthostegan) genus whose type species, Steginopora ornata, was exceptionally well illustrated by d’Orbigny. Unfortunately, there appear to be no known specimens of S. ornata, but pending revision of the many cribrimorph taxa that now exist, and with the assurance that topotypic material would be easily recognisable, the family name Steginoporidae could yet have validity.

The classification of Centrifuginés used by d’Orbigny in the Terrains crétacés similarly distinguished two suborders (Table 3). The more minor of these, the Radicellés, with colonies attached by ‘roots’, contained only two families, Crisidae and Serialaridae, distinguished respectively by the presence and absence of articulations. Whereas d’Orbigny’s understanding of Crisidae was essentially identical to the modern concept of this cyclostome family, his Serialaridae contained genera now placed in the bryozoan order Ctenostomata, including Amathia. D’Orbigny divided his larger suborder – ‘Empâtés’ – into four divisions: ‘Operculés’, ‘Fasciculinés’, ‘Tubulinés’ and ‘Foraminés’. The ‘Operculés’ were, as the name implies, characterised by having zooids with an operculum, a hinged flap closing the orifice. D’Orbigny mixed families we now regard as cyclostomes (Eleidae) and cheilostomes (Myriozoumidae = Myriaporidae) in this division, but he deserves credit for recognising that the Eleidae, despite their substantial convergence with cheilostomes, are in fact cyclostomes. His three other divisions – ‘Fasciculinés’, ‘Tubulinés’ and ‘Foraminés’ – are all non-operculate cyclostomes differing in the arrangement and structure of the zooids. The zooids are bundled into fascicles in the ‘Fasciculinés’, have projecting peristomes in the ‘Tubulinés’, and are non-projecting in the ‘Foraminés’. These divisions correspond respectively to skeletal organisations termed mixed-walled, fixed-walled and free-walled 〚26〛. The phylogenetic significance of these organisational types, and hence their value as taxonomic subdivisions, has yet to be established.

There is one important respect in which d’Orbigny’s classification of cyclostomes differs from schemes used nowadays. The skeletal structures and shapes of brooding zooids (gonozooids) are today regarded as crucial in the definition of higher taxa of cyclostomes 〚26〛. Gonozooids, however, played no part in d’Orbigny’s classification and very few gonozooids were even figured by d’Orbigny: two examples 〚17, pl. 761, figs 3′ and 7〛 are referred to in his text as ‘cellules accessoires’. D’Orbigny’s apparent failure to notice gonozooids is less surprising when the rarity of these polymorphs in most cyclostome species is taken into consideration: it is not unusual to require several hundred branch fragments of erect colonies before a gonozooid is discovered.

D’Orbigny’s generic classification of cyclostomes emphasised colony-form. Sometimes this philosophy was applied to such an extent that encrusting colonies with different shapes as a result of growing on different substrates were placed in different genera. Colonies of Actinopora and Reptoclausa, which grew around perished cylindrical substrates were, for instance, the basis of d’Orbigny’s genera Conotubigera and Semiclausa respecively. Genera within his families tended to be recognised and named according to their colony-form, leading to a very logical system of nomenclature. For example, genera with multilayered colonies had ‘multi’ within their names, those with clavate colonies ‘clavi’, reptant colonies ‘repto’ etc. The root of the family was also incorporated in the generic name, apart from those genera named before he had developed the full classification expounded in the Terrains crétacés. Thus, genera within the Family Cavidae have names ending with ‘-cava’, those within the Clausidae end with ‘-clausa’, Sparsidae with ‘-sparsa’ etc. Both the family affliliation of any given genus and some aspect of its colony-form are consequently evident from its name. For example, Multisparsa is a multilayered genus belonging to the Sparsidae, and Claviclausa is a clavate genus within the Clausidae. An unfortunate consequence of this system is that genera with comparable colony-forms placed within the similarly spelt families Caveidae and Cavidae have names that are very easily confused, e.g., Semimulticavea vs Semimulticava, Clavicavea vs Clavicava.

The legacy left by d’Orbigny to bryozoology is almost entirely taxonomic – he erected a plethora of species, genera and families of bryozoans, many of which remain in use today. With regard to cyclostome bryozoans (Centrifuginés), a total of about 370 nominal genera can be found in the literature. Of these, Alcide d’Orbigny is the author of approximately 30% (113; omitting an objective junior synonym and a genus now known to be a cheilostome), by a considerable margin the greatest number named by any single taxonomist. A further 13 d’Orbigny species have since been designated as the type species of genera by later authors. While some of d’Orbigny’s cyclostome genera are little known or seldom used (e.g., Pyricavea), others are employed routinely (e.g., Fasciculipora) in contemporary bryozoan taxonomy.

D’Orbigny introduced 75 new cheilostome genera. Following our study, and discarding those for which type species are lost and which may never be properly understood, we regard 40 of these as having current availability. This number represents only 4% of the 1013 genera and subgenera currently compiled for the Treatise revision of cheilostomes. We shall reinstate many that were previously overlooked or ignored by Bassler 〚1〛. These comprise Filiflustrina, Latereschara, Lateroflustrella, Multescharellina, Multescharipora, Pyriflustrina, Reptescharella, Reptoflustrella, Reptolunulites, Reptoporella, Reptoporina, Reptescharinella, Semiescharipora, Semiflustrella, Semiflustrina, and Reptescharipora. On the other hand, a number of d’Orbigny’s genera have long been in use, including Conescharellina, Discoporella, Flabellopora, Fusicellaria, Poricellaria, Porina, and Pyripora, among others.

While it is reasonable to portray d’Orbigny as a taxonomic splitter who over-emphasised colony-form and paid too much regard to stratigraphical differences, Heron-Allen’s contention 〚7, p. 60〛 that for these reasons his work “was to all intents and purposes ignored by MacCoy, Hagenow and Haime in their later and more authoritative works on the group” surely exaggerates the truth.

D’Orbigny’s taxonomic descriptions were often of a comparable standard to those published by other authors at the time. Furthermore, he developed thoughtful and logical systems of classification that, in some cases, have stood the test of time. His work has, however, caused a great many difficulties for subsequent students of the Bryozoa. Problems arise from: (1) dates of publication which do not coincide with those given on the title pages of the works and in some instances (Terrains crétacés plates) are unknown (publication delays were not d’Orbigny’s fault but were associated with political unrest at the time of the abdication of King Louis-Philippe and declaration of the Second Republic); (2) d’Orbigny’s habit of using the same species name (e.g., cenomana) over and over again within different genera; (3) his tendency to create confusingly similar generic and familial names (e.g., Escharipora (Fig. 4A and C) and Escharifora (Fig. 3B and E); Escharinellidae and Escharellinidae); (4) the use of stylised and/or composite figures, as in the Terrains crétacés; (5) taxonomic oversplitting, owing, for example, to slight differences in colony form or stratigraphical differences (no species were ‘allowed’ to occur in more than one stratigraphical stage because of d’Orbigny’s belief that total extinction of all life occurred at the end of each stage, followed by a fresh creation at the beginning of the next); (6) his confusing habit of attributing generically reassigned species to himself (e.g., Pustulopora pseudospiralis Michelin was renamed ‘Peripora pseudospiralis d’Orbigny’ when he transferred the species to the genus Peripora); (7) appending his new taxa with the date that he penned the taxonomic description rather than the actual date of publication and hence availability of the taxon (e.g., ‘Trochopora d’Orbigny, 1847’ was actually published in 1849 〚13〛); (8) in many instances, listing new taxa without figures or adequate descriptions.