acastid trilobites from Gotland and Scania, Sweden - Springer Link

13 downloads 0 Views 5MB Size Report
Acastella werden beschrieben. Einige Merkmale sprechen daffir, dal~ Acaste und verwandte Taxa vorwie- gend aus dem Wenlock zur monophyletischen ...
I Pal~iont. Z.

67

3/4

261-285

8 Abb., 1 Tab.

Stuttgart, N o v e m b e r 1993 I

Silurian (Ludlow) acastid trilobites from Gotland and Scania, Sweden LARS RAMSKC)LD, Stockholm & GREGORY D. EDGECOMBE, E d m o n t o n *

With 8 figures K u r z f a s s u n g : Acastide Trilobiten aus dem Ludlow von Gotland enthalten drei stratigraphisch nicht fiberlappende Arten von Acastella REED: A. rnadidipes n. sp. (Hemse Mergel und Eke Schichten), A. breviceps (ANGELIN 1851) (Burgsvik Schichten) und A. amatrix n. sp. (Hamra Schichten und wahrscheinlich Burgsvik Schichten). Das Typenmaterial sowie weiteres Material von A. breviceps (ANGELm) aus der Klinta Formation in Schonen wird revidiert. Meraspiden- und frfihe Holaspiden-Stadien von Acastella werden beschrieben. Einige Merkmale sprechen daffir, dal~ Acaste und verwandte Taxa vorwiegend aus dem Wenlock zur monophyletischen Unterfamilie Acastinae geh6ren. Acastella ist nach g~ingiger Auffassung eine Entwicklungslinie, deren devonische Arten enger mit den Asteropyginae verwandt sind als mit der Typus-Art aus dem Ludlow. Baltische Arten aus dem Ludlow-Pridoli, die frfiher zu Scotiella DELO gestellt wurden, bilden eine eigene Gruppe, Ewacaste n. gen., zu der auch eine Art aus den obersten Eke Schichten Gotlands geh6rt. A b s t r a c t : Acastid trilobites from Ludlow strata on Gotland include three stratigraphically nonoverlapping species of Acastella REED: Acastella rnadidipes n. sp. (Hemse Marl and Eke Beds), A. breviceps (ANGELm 1851) (Burgsvik Beds), and A. arnatrix n. sp. (Hamra Beds, and probably Burgsvik Beds). Type and other material of A. breviceps (ANGELIN)from the Klinta Formation, Scania, is revised. Meraspid and early holaspid material of Acastella is described. Some characters support Acaste and allied taxa of mostly Wenlock age as a monophyletic subfamily Acastinae. Acastella as widely recognized is a grade with Devonian species more closely related to Asteropyginae than to the Ludlow type species. Baltic LudlowP~idoli species previously placed in Scotiella DELO form a distinct group, Ewacaste n. gen., that includes a species from the uppermost Eke Beds on Gotland.

Introduction The first acastid trilobite to be described from Sweden was Phacops breviceps ANGELIN 1851. This species was based upon material from both Ringsj6n in Scania and from Gotland. A complete, extended specimen was figured, but no locality was given. The figure is a composite reconstruction of material labelled as being from Klinta a Ringsj6n. N o Gotland material of the species was figured. LINDSTROM (1885) regarded Phacops breviceps ANGELIN 1851 as a junior synonym of the British Phacops downingiae MURCmSON 1839. L~NDSTR6M gave a list of Gotland localities for P. downingiae, form the Eke, Burgsvik, and Hamra Beds. In addition, he listed the species from "Palissaderna, Visby", an occurrence that is discussed separately below. Unfortunately, LmDSTR6M did not illustrate any acastid specimens.

* Adresses of the authors: Dr. LARS RAMSK6LD, Sektionen f6r Palaeozoologi, Naturhistoriska Riksmuseet, Box 50007, S-10405 Stockholm, Sweden; present adress, Department of Historical Geology & Palaeontology, University of Uppsala, Norbyv~igen 22, S-75236 Uppsala, Sweden; Dr. GREGORY D. EDOECOMBE, Department of Geology, University of Alberta, Edmonton, T6G 2E3, Canada; present adress: Division of Earth Sciences, Australian Museum, P.O. Box A285, Sydney South, NSW 2000, Australia. 0031-0220/93/0067-0261 $ 6.25 9 1993E. Schweizerbart'scheVerlagsbuchhandlung,D 70176Stuttgart

262

Lars Ramsk61d & Gregory D. Edgecombe

After LINDSTR6M'Sstudy, Gotland acastids have been left virtually untouched for over a hundred years. In connection with mapping and stratigraphic research, "Acaste downingiae" was listed as occurring in the Hemse, Eke, Burgsvik, and Hamra Beds, all of Ludlow age (HEDE 1921: pl. 1; HEDE in Mt3NTHE et al. 1925; MUNTHE et al. 1927, 1929). The species was also used to correlate the upper part of the Gotland sequence with the Klinta Formation (former Oved-Ramsfisa beds) of the Scanian sequence (HEDE 1921). None of the specimens were figured, but from the levels and localities given, the material must have included several or all of the species described in this paper. More recently, ALBERTIet al. (1982) recognized a Proetus signatus level in the Baltic region, and stated "It should be noted that we have P. signatus and Acastella prima TOMCZYIiOWA together in the upper part of the Burgsvik Sandstone and Oolite. Above the oolites [i. e., in the Hamra Beds] we have recorded an Acastella species which may be interpreted as phyletically transitional between "prima" and "dayiana'." The Gotland material of AL~ERTr et al. (1982), collected by H. and L. ALBERTI,has now been transferred to the Swedish Museum of Natural History. Among the several acastids in the collection, we find none that can be assigned to either A prima or A. dayiana; this reference is apparently to Acastella amatrix n. sp. Of particular interest is that an acastid has been reported from the Lower Wenlock H6gklint Beds of Gotland (LINDSTR6M 1885: 43; OXV~NSin BRUTON et al. 1979: 118). If correct, this occurrence would be the earliest known for the subfamily Acastinae, only predated by the reported, though not documented, occurrence (no specimens listed or figured) of Acastocephala macrops in the mid and upper Llandovery of Britain (THOMAS et at. 1984). The alleged occurrence is, however, incorrect. LINDSTR(SMstated (1885: 43) that "Phacops Downingiae" had been collected from "the upper limestone beds at the waterfall by Palissaderna in Visby". In current terminology this means the H6gklint Beds unit c at Vattenfallsprofilen 1. More recently, OWENS in BRI3TON et al. (1979: 118) reported Acastocephala cf. macrops (SALTER)from this horizon. Both LINDSTROM'Sand OwENs reports are based on the same specimens (Ar31128), a cephalon and pygidium preserved on the same small piece of rock. Adjacent to the trilobite parts are also some brachiopods. Two of these can be identified as "Stropbomena impressa" of MISNTHE (1902: figs. 3, 4) (see Sbaleria aft. ornatella in BASSETT~ COCKS 1974: 17). This characteristic species is common in limestones belonging to the Hemse and Eke Beds, but it does not occur below the Hemse Beds. The second brachiopod species in the slab belongs to Streptocbonetes, and appears to be conspecific with S. striatellus DALMAN, another species from the Hemse and Eke Beds. It is thus certain that LXNDSWR6M'Ssample had been mislabeled, and that it is from a horizon much higher than the H6gklint Beds. The acastid specimens themselves are internal moulds, and thus cannot be safely identified to the species level. OWENS' (in BRUTON et al. 1979) assignment to Acastocepbala cannot, however, be upheld because the moderate size of the eyes and the presence of a small pygidial mucronation precludes that referral. There is no reason to doubt that they belong to the only species known from the Hemse and Eke Beds, AcasteUa madidipes n. sp. This study includes all acastids that are presently known from Sweden. The exoskeleton of these trilobites has a marked tendency to part from the internal mould, and a majority of specimens described from other regions are internal moulds. For this paper, particular effort has been spent on preparing material with the exoskeleton intact. The number of internal moulds greatly exceeded the number of testiferous specimens available. We have not, however, figured any internal moulds. Such material often offers limited detail for comparison with species based on exoskeletal material or those for which external moulds are available. It is advocated that trilobite species in general, and acastids are no exception, should not be based on internal moulds; such material is generally better left in open nomenclature. The stratigraphic scheme used here for Gotland follows LAUFELX)(1974b); Scanian units are those of JEvvssor~ ~ LAUFELD(1987). Gotland and Scanian locality names followed by numbers are defined accordin~ to the system of reference localities introduced by LAIZrZLD(1974a) for Gotland, and are describea in that study or in LARSSO~q(1979).

Silurian (Ludlow) acastid trilobites

263

Morphological terminology employed here follows HARRINGTONet al. (1959). Glabellar lobes (L) and furrows (S) are numbered from 0 to 3 anteriorly (JAANUSSON1956). Dorsal view of the cephalon is that of CLARKSON(1966). The specimens are housed in the Swedish Museum of Natural History (prefix Ar), and the Type Collection of the Geological Survey of Sweden (SGU). All photographs are of the external surface of the exoskeleton. Except in scanning electron micrographs, all specimens were painted with matt black opaque and lightly coated with ammonium chloride prior to photography. Specimens were collected by L. R. unless stated otherwise in the figure captions.

Systematics Family Acastidae D r L o R e m a r k s : As treated by EDGECOMBE (1993), the family Acastidae includes those taxa conventionally assigned to the subfamilies Acastinae DELO 1935, Acastavinae STRtJVE 1958, and Asteropyginae DELO 1935. Acastidae s.s. is a post-Llandovery clade, to which midLlandovery Llandovacaste EDOECOMBE 1993 is recognized as most closely related. Some previous treatments (STRUVE 1959) have assigned Acastinae and Acastavinae + Asteropyginae to different families, Calmoniidae and Dalmanitidae, respectively. This classification is opposed by synapomorphies that group acastines, acastavines, asteropygines, and SiluroDevonian calmoniids (ELr)REDGE 1979; EDGECOMBE 1993) as a clade distinct from dalmanitids. It is also opposed by the widely held view (ToMcZYKOWA 1962a, b, 1991; TOMCZYKOWA in TOMCZYKOWA & WITWICKA 1974; SHERGOLD 1966) that Acaste (subfamily Acastinae) is ancestral to Acastella (placed in Acastavinae by STRUVE 1958, 1959). Such a relationship, of direct ancestry and descent, would demand a closer classification of these taxa. The subfamilial classification of Acastidae is fraught with problems, particularly concerning the monophyly of Acastinae and Acastavinae. Asteropyginae is an unambiguous monophyletic group, but Acastavinae is paraphyletic, if not polyphyletic. The ancestral position of Acastavinae relative to the asteropygine clade was explicitly stated by GAm)L (1972), TIMSi (1981), SMEENR (1983), and TOMCZYKOWA (1991). PILLET (1961: 94) summarized character information that Acastinae was defined by primitive states relative to Acastavinae + Asteropyginae, and Acastavinae was based on plesiomorphies for Asteropyginae. This expresses the conventional topology of the acastid tree: Acastinae and Acastavinae are successive transformational grades. Indeed the original concept of Acastavinae (STRUVE 1959: 475) emphasized a "distinct Asteropyginae trend". To this grade have been assigned the following genera: Acastava RICHTER & RICHTER 1954, Acastella REED, 1925, Acastellina RICHTER & RICHTER, 1954, Acastocephala SHERGOLD 1966, Pelitlina HAAS 1968, Centauropyge HAAS 1968, Mirnocryphaeus GAm)L 1972, Chimaerastella GANDL 1972, and Acastopyge TOMCZYKOXVA1974. These genera have been grouped based on a tendency towards greater spinosity (genal spines; pygidial marginal spines) than in Acastinae, but without full expression of the asteropygine states. Discussion below (under Acastella) explores the possibility that the subfamily Acastinae can be defined based on synapomorphies. There is little hope of salvaging Acastavinae as a monophyletic group, and we have not recognized a formal taxon for these 'stem lineage' (sensu Ax 1985) Asteropyginae. TOMCZYKOWA(1991) has recently outlined an alternative classification of the acastaceans, recognizing Calmoniidae, Acastidae, and Asteropygidae as separate families. Acastidae is divided into subfamilies Kloucekiinae, Acastinae, and Acastavinae. This group is unquestionably paraphyletic. TOMCZYKOWAherself (1991: 47) explicitly stated that acastavines are the closest relatives of Asteropyginae, and recent work has amply demonstrated that calmoniids are more closely related to acastines and acastavines than are "kloucekiines" (EDGECOMBE 1993). The effect of excluding calmoniids and asteropygines from Acastidae sensu TOMCZY-

264

Lars Ramsk61d & Gregory D. Edgecombe

KOWA is thus a loss of informative monophyletic groups. The classification of acastids with calmoniids was rejected by TOMCZYKOWA (1991: 34) based on the subjective criterion that they "differ enough". The separate familial rank of Asteropygidae was justified by ToMczYliowA (1991: 47) by a list of differences from Dalmanitidae. The comparison is of little relevance, since asteropygines are more closely related to Acastidae sensu ToMczYxowA (or, more precisely, to ingroup acastids) than to Dalmanitidae. TOMCZYXOWA's concept of Kloucekiinae groups all acastaceans showing effaced $2 and $3. However, total character evidence clearly indicates that this grouping is paraphyletic. Ordovician "kloucekiines" are a paraphyletic stem lineage to Acastacea s.s. (EDCECOIVIBE 1993), and Siluro-Devonian taxa that "answer the diagnosis" of Kloucekiinae (ToMcZYKOWA 1991: 35) are ingroup Acastacea s.s. (see discussion below on Scotiella). O u r phylogenetic approach rejects reliance on a single character that conflicts with a greater b o d y of evidence.

Aca~tc._~ A

T"

B

/

ta

Fig. 1 Alternative hypotheses for relationship of Scotiella and Acastella to Acaste, and affinities of Ewacaste n. gen. - A: Phylogenetic tree wim Scotiella and Acastella separately derived from Acaste (SrIeRCOLD 1967b: 2, 18); species of Ewacaste classified as Scotiella. Characters are: 1 - slender pygidial posteromedian spine; 2 - very short L1; 3 - weakly divergent branches of $3; 4 - genal angle non-sl~inose; 5 - $2 and $3 very lightly impressed. Characters with 'R' are reversals. - B: More parsimonious claaogram with Scotiella logani and Acastella sharing common ancestry more recent than Acaste. Characters 1-5 are as in Fig. 1A. - C: Relationships of Acastidae as discussed in text, with Ewacaste n. gen. in "Wenlock group .

Silurian (Ludlow) acastid trilobites

265

Genus Acastella REED 1925 Type species: Pbacops (Acaste)Downingiae MURCmSON,var? 8 spinosusSALTER1864; by original designation (REED 1925: 75); revised SrtERGOLD(1967a). R e m a r k s : Problems with Acaste GOLDFtrSS 1843, and AcasteUa REED 1925, are legion, and their revision as definable clades involves reassessment of most Wenlock to Early Devonian acastids. Some broad morphological trends are obvious for acastines through the Wenlock-Gedinnian interval. Wentock species now classified as Acaste, Acastoides DELO 1935, and Acastocephala SHERGOLD1966, share a number of morphological features that distinguish them from most Ludlow and younger Acastinae. We informally refer to these taxa as the "Wenlock group" (as noted below, a few species persist into the late Silurian). In these stratigraphically early taxa, the genal angle is rounded or bears a tiny node (RICHTER & RICHTER 1954: pl. 3, fig. 39a), the glabellar anterior margin is evenly rounded, and a deep, long median impression is developed in the posterior part of the frontal glabellar lobe. The pygidial posteromedian margin is rounded or bears, at most, a weak point (e. g., Acaste downingiae). Species of this "Wenlock group" have a relatively wide glabella; in most the branches of $3 diverge at a large angle, $2 is distinctly convex forward, and the preglabellar furrow is deep. In contrast to the "Wenlock group", most Ludlow-P~idoli species (usually assigned to Acastella REED) possess a slender genal spine, the cranidial anteromedian margin is subangular, the frontal lobe posteromedian impression is usually shorter and shallower than in the "Wenlock group", and the pygidial posteromedian margin bears a node or spine. Early Devonian species referred to Acastetla typically have larger genal spines than the slender or thorn-like spines of LudlowP~dolf species, and L1 is often relatively longer (exsag.). In addition, the pygidial terminal spine is stout, and some species display marginal denticulations in mature growth stages. While the Wenlock acastines represent a morphologically coherent group, evidence for monophyly is ambiguous. The convential notion that Acastella evolved from Acaste (via intermediates such as Acaste dayiana RICHTER & RICHTER 1954, and Acastella prima TOMCZYKOWA 1962a; see TOMCZYKOWA, 1962a, 1991, and TOMCZYXOWAin TOMCZYKOWA & WITWiCKA 1974: 69) is in accord with stratigraphic distribution; most species of Acaste are Wenlockian, whereas Acastella is first described from the Ludlow and has been assigned numerous Devonian species. Certainly most of the typical characters of the "Wenlock group" are symplesiomorphic: the wide glabella, markedly convex $2, and deep preglabellar furrow are retained from primitive acastid outgroups such as Llandovacaste EDGECOMBE 1993. However, using alternative methods (to stratigraphy) to infer character polarity, some typical characters of the "Wenlock group" can be interpreted as apomorphic. Ontogeny (e.g., Acastella amatrix n. sp.) indicates that a rounded, Acaste-like genal angle is a transformation from a slender-spined (Acastella-like) form in juvenile holaspides. As such, ontogeny polarization predicts that the "Wenlock group" is united by a synapomorphy, rounded genal angles. This polarity is in accord with outgroup polarization, accepting that Llandovacaste is the closest relative (sister group) of post-Llandovery Acastinae s.s. (Fig. 1C). This outgroup has a slender genal spine resembling that of Acastella. Outgroup comparison with Llandovacaste and more distal acastacean taxa (i. e., primitive Calmoniidae such as Andinacaste ELDREDGE& BRANIgA 1980; EDGECOMBE1993) interprets absence of a longitudinal glabellar posteromedian impression as the basal state for Acastidae s.s. The long, deep furrow possessed by the "Wenlock group" Acaste, Acastoides? constricta (SALTER),and Acastocephala (see SHERGOED 1966) is thus regarded as apomorphic relative to the short (Acastella spinosa; A breviceps) or very shallow to obscure furrow seen in most post-Wenlock acastids. Ontogeny appears to provide the same polarization for this character; the smallest known cranidium of Acastella breviceps (Fig. 6A) displays only the faintest trace of a longitudinal median impression,

266

Lars Ramsk61d & Gregory D. Edgecombe

whereas large holaspides have a well-impressed furrow or paired pits (i. e., absence of furrow is more general or plesiomorphic). Thus, some typical characters of the ~Wenlock group" of Acastinae may be synapomorphies. This interpretation conflicts with the conventional notion that acastid phylogeny can be viewed as an early grade (Acastinae) that transforms to Asteropyginae through an acastavine grade based in Acastella. This is not to say that the conventional tale of acastines as ancestors does not more or less correctly describe a general pattern in morphology within Acastidae, as described above. A monophyletic Acastinae (i.e., Acaste, Acastoides? constricta, Acastocephala, Ewacaste n. gen.; Fig. 1C) explains the persistence of the "Wenlock group" into the Late Silurian, with P~fdoll species like Acaste podolica BALASHOVA 1968 and A. zerinae EDGECOMBE 1993. They would be stratigraphically late species within a clade. A strict gradistic interpretation of Acastinae (as an early, Wenlock ancestral stock) has a more difficult task of interpreting these species, which should have undergone transformational pseudoextinction by the P~idolf. The above discussion has focused on the problem of paraphyly in distinguishing Acaste from Acastella, i.e., where the basal node for Acastella can be identified. However, the 'upper' limit of Acastella is also plagued by paraphyly. Acastella has been assigned a large number of species, most of Early Devonian age (see HOLLARD 1963; GANDL 1972; TOMCZYKOWA1991). The Ludlow (Ludfordian) type species, A. spinosa (SALTER), is actually among the stratigraphically earliest occurrences. The role of Acastella as an ancestral taxon has been most explicitly conveyed in the phylogenetic tree of TIMM (1981). A grade of Silurian-Devonian Acastella species were depicted as giving rise to the subfamily Asteropyginae, with Paracryphaeus GANDL 1972, as the basal asteropygine. SMEENK (1983: 417) and TOMCZYKOWA (1991: 42) accepted this interpretation of "Acastavinae" (specifically, Acastella) being ancestral to (monophyletic) Asteropyginae. If this interpretation is correct, AcasteUa is paraphyletic. Thus, the descent of Asteropyginae is a first-order problem for the monophyly of AcasteIIa as that genus has been conceived. Another problem is posed by Scotiella DELo 1935, which appears to be more closely related to Acastella spinosa (i. e., a Late Silurian core group of Acastella) than are some Devonian species referred to Acastella. SHERGOLD (1967b) reviewed the status of Scotiella, considering its relationship to Acastella to derive only from a similar parental source in Acaste (Fig. IA). This view interprets Scotiella and Acastella as coeval independent lineages of predominantly North American and European-North African diversity, respectively. We suggest, however, a more parsimonious explanation for similarities shared by Scotiella and Acastella that are not shared by their alleged ancestor, Acaste (Fig. 1B). Characters such as a slender pygidial posteromedian spine, very short L1, and relatively weakly divergent branches of $3 (and, doubtfully, genal spines) need not he interpreted as convergently acquired in Acastella and Scotiella logani (from an absence or primitive homologue in Acaste). Instead, they imply a closer relationship between AcasteUa and Scotiella s.s. than either shares with Acaste. TOMCZYKOWA (1962b: 202) made the same proposal, that

Fig. 2 A-G, L-N: Acastella madidipes n. sp. from Gotland. Hemse Beds, uppermost part (B-G), and Eke Beds, upper part (A, L-N). A: Dorsal view of cephalon, Lau Backar 1, x 3.5; Ar54949. B-D: Anterior, dorsal, and lateral views of cephalon, Gannor 1, x 3.5; Ar54918. E-G: Dorsal, oblique anterolateral, and lateral views of cephalon with smaller, incomplete cephalon covering anterior part, note bryozoan overgrowth on both cephala, Burs parish, Wanges canal, old collection, x 3; Ar31083. L: Dorsal view of partial cephalon, Lau Backar 1, x 6; Ar54951. M-N: Dorsal pygidial and cephalic views of small, enrolled, partly crushed exoskeleton, Lau Backar 1, x 5; Ar54950. H-K: AcasteUa cf. madidipes n. sp. from Gotland, top of Hemse Beds or base of Eke Beds, Olsvenne 4. Anterior, posterior, dorsal, and lateral views of cephalon, • 4; Ar54932.

Silurian (Ludlow) acastid trilobites

267

268

Lars Ramsk61d & Gregory D. Edgecombe

ScotieUa is "probably an offshoot of Acastella'. However, we do not accept that European species assigned to Scotiella (ToMczYKOWA 1962b, 1991; SHERGOLD 1967b; SCHV~mi 1970, 1972) are the closest relatives of S. logani (HALL), the type species. As discussed below, the European species are recognized as a separate lineage, Ewacaste n. gen. The interpretation here of Acastocephala as part of the "Wenlock group" (most closely related to Acaste) differs somewhat from SHERGOLD'S original interpretation that Acastocephala was a possible link to Devonian Acastava RICHTER & RICHTER, SHERGOLD (1966) attributed a diagnostic significance to the presence of pygidial marginal denticles in juvenile stages of Acastocephala. This character is, however, present in juvenile holaspides of Acaste (ScHRANK 1972: pl. 20, figs. 2, 2a), Acastella (Fig. 3F, 5B-G, 7M-O), and Scotiella (SHERGOLD 1967b: pl. 3, fig. 5), and simply represents an intermediate stage in the reduction of the marginal spines of meraspides (ToMczYIIOWA 1991: pl. 10, figs. 4, 5, pl. 11, figs. 1-5). Juvenile marginal denticles are certainly a plesiomorphic condition for Acastidae. SHERGOLI) (1966: 199) cited the straight, transverse form of $2 as evidence uniting Acastocephala macrops (SALTER) with Acastava, but this state is not shared with the second species assigned to Acastocephala, A. cludleyensis (SHERGOLD 1966). The latter species has $2 with prominent anterior convexity, as typifies Acaste. The only significant similarities shared by Acastocephala macrops and A. dudleyensis are the large eye and short, weakly furrowed pygidium. Swedish Ludlow species described in this work resemble Silurian Acastella (e. g., A. prima TOMCZYKOXVA) in the development of a pygidial posteromedian node. None possess the slender posteromedian spine of Acastella spinosa and Scotiella. Expression of genal spines in the Swedish species ranges from an absence in AcasteUa amatrix n. sp. to short spines in Acastella breviceps and A. madipides n. sp.

Acastella rnadidipes n. sp. Fig. 2A-G, L-N; Fig. 3A-F, H; Fig. 4A, B; Fig. 5A, D - G D i a g n o s i s : $2 straight or convex forward, shallow or obsolete abaxially, not confluent with axial furrow; eye fairly large, length (exsag.) 39-45% of cephalic length; visual surface taller anteriorly than posteriorly; genal angle with thorn-like point; exoskeletal sculpture of fine granules; first rachial ring 34.5-36.5% of pygidal width. Holotype: Articulated exoskeleton Ar31227 (Fig. 3E), from the Hemse Beds, top of Hemse Marl, Gannor I (defined in LAUFELD1974a), Lau parish, Gotland. Other material: Hemse Marl, top: Gannor 1, Ar54918-54929; Botvide 1, Ar54933; Wanges kanal, Burs, Ar31083; Burs, Ar54934; Hallsarve 1, Ar54967; Hajslunds, Havdhem, SGU unnumbered. Eke Beds, lower part: Gannor 1, Ar54917; Sigdes kanal (including Sigdes 1), Rone, Ar54938-54948. Eke Beds, upper part: Lau Backar 1, Ar31079-31080, Ar54089-54093, Ar54949-54966. Etymology: Latin madidus, sodden, and pes, feet, referring to collecting conditions at the type locality, which is accessible at low water level from a bar in the Lau Canal.

Fig. 3. A-F, H: Acastella rnadidipes n. sp. from Gotland. Hemse Beds, uppermost part (D-F, H); Eke Beds, lower part (A); Eke Beds, upper part (B-C). A: Dorsal view of partly flattened and disarticulated exoskeleton, Gannor 1, probably 0.7 m above Hemse/Eke boundary, collected by S. LmDBOM1977, X 3; Ar54917. B-C: Dorsal and lateral views of pygidium, Lau Backar 1, • 6; Ar54952. D: Dorsal view of largest pygidium, Botvide 1, x 5; Ar54933. E: Dorsal view of holotype articulated exoskeleton, Gannor 1, oldcollection, • 3; Ar31227. F: Dorsal view of small pygidium, Gannor 1, x 6, note weakly denticulate margin; Ar54920. H: Dorsal view ofpygidium, Gannor 1, x 5.5; Ar54919. G: Acastella cf. rnadidipes n. sp. from Gotland, top of Hemse Beds or base of Eke Beds, Olsvenne 3. Dorsal view of pygidium, • 6; Ar54930. I-L: Ewacaste cf. quaesita ( S c ~ N I i 1972) from Gotland, uppermost Eke Beds, immediately below Eke/Burgsvik boundary, Ronehamn, old collection. Lateral, oblique anterolateral, dorsal, and anterior views of partly exfoliated cephalon, • 3.5; Ar31402.

Silurian (Ludlow) acastid trilobites

269

270

Lars Ramsk61d& Gregory D. Edgecombe

Description: Cephalic length 50-55% of width. Rachial furrow narrow, moderately impressed against L1, gently widening forward and faintly to distinctly convex outward between S1 and $3, with moderate anterior divergence, gently flexed inward at $3, shallowing abruptly against frontal lobe. Glabellar length (sag., including L0) 94-103% of width across frontal lobe. Glabella of faint to gentle convexity (sag.) between SO and posteromedian part of frontal lobe, gently elevating anteriorly, sloping downward on anteromedian part of frontal lobe. Anterolateral margin of glabella faintly convex outward; anteromedian margin rounded. Preglabellar furrow narrow, shallow to moderately deep. Cranidial anterior border short (exsag.), weakly angular medially, with nearly straight anterolateral margin. Anterior branch of facial suture with straight anteromedial course (in dorsal orientation) between gamma and outside anterolateral edge of glabella, then abruptly flexed inward. Librigenal component of cephalic anterior border short (exsag.) abaxially in dorsal orientation, nearly or actually overhung by cranidial border medially. Frontal lobe of glabella gently arched (tr.) across most of width. Longitudinal posteromedian furrow on frontal lobe very short, shallow, occasionally indistinct. $3 weakly sinuous, confluent with rachial furrow, shallow to moderately deep. L3 gently convex (tr.), L2 and L1 progressively more strongly arched. $2 moderately deep, shallowing distally, faintly sinuous or gently convex forward, transverse or gently directed anteromedially. $1 deep, gently concave forward, directed anteromedially. SO deeply impressed abaxially as faintly curved apodermal grooves, oriented slightly anteromedially. SO short (sag.), moderately deep medially, anterior margin gently convex forward. Width of L0 about 70% of glabella width across frontal lobe. L0 longest medially, length (sag.) 26-29% of width, gently convex (sag.), weakly to distinctly raised above L1, moderately arched transversely. Anterior margin of eye opposite $3; posterior margin of eye between midlength (exsag.) and posterolateral corner of L1. Palpebral furrow narrow, shallow to moderately incised, nearly effaced in some large specimens (Figs. 2C, 4A). Palpebral lobe weakly inflated. Visual field Of 24 dorsoventral lens files with maximum of eight lenses per file, typical lens formula from anterior (Fig. 2B - D): 145 667 787 878 788 777 765 432. Subocular groove narrow, shallow. Palpebral area with shallow pitting, gently sloping to rachial furrow. Librigenat field faintly convex, moderately declined outward to border furrow. Lateral border furrow variably shallow, faint to distinctly impressed, with decreased incision anteriorly and posterolaterally. Lateral border of weak convexity, widest just behind midlength (exsag.), gently narrowing forward and backward. Posterior branch of facial suture with gently sinuous course, arched backward then broadly forward between epsilon and outer part of lateral border, sharply flexed backward to margin; omega well forward of epsilon. Postocular fixigena sloping downward to border furrow. Posterior border furrow narrow, moderately deep, with nearly straight, gently anteriorly-directed course (tr.) between rachial furrow and behind about midwidth of palpebral lobe, inflected slightly more anteriorly abaxial to this, narrowing and effacing near juncture with lateral border furrow. Posterior border short (exsag.), gently lengthening abaxially proximal to rachial furrow, then more strongly lengthening to maximum inward of genal angle, with posterior margin convex backward. Thornlike genal spine gently inflected outward, may be reduced to a small point in large specimens (Fig. 4). Thorax of 11 segments. Rachial furrow shallow, narrow. Rachis about one-third width of thorax. Rachial ring of about equal length across most of width, gently lengthened abaxially, arched forward medially. Ring evenly rounded in transverse profile, weakly convex in sagittal profile, flat (exsag.) near rachial furrow, more convex immediately adjacent to rachial furrow, with tendency towards weak inflation. Ring furrow short (sag.), moderately deep. Pleural furrow deep, narrow, with approximately straight course, becoming effaced on articulating facet; facet shortening (exsag.) on posterior segments. Anterior pleural tips with short spine at posterior edge just behind articulating facet, terminae of posterior ribs more bluntly truncated.

Silurian (Ludlow) acastid trilobites

271

Pygidial length 57-65% of width. Seven anterior rachial rings 82.5-83.9% of rachial length. Eight to nine well defined rings, tenth discernible in some specimens (Fig. 3A). Deep, transverse apodemes in anterior three inter-ring furrows; narrow, shallower pair in fourth (and fifth in one specimen; Fig. 3A); posterior ring furrows shallow, straight. Rachial furrow more strongly diverging anteriorly against first four rings than posteriorly. Well developed pseudo-articulating half ring behind first rachial ring, less pronounced behind second ring. Lateral margin of largest specimens evenly convex outward (Fig. 3D), with only faint trace of anterior pleural lappet; minute, blunt marginal swellings defined in small holaspides (Fig. 3F), becoming faint in posterior pleurae. Pleural furrow on first rib post-facetal, effacing slightly behind posterior edge of articulating facet. Pleural furrows nearly straight for most of width, flexed backward at distal end, effacing against border. Seventh pleural furrow very shallow in largest specimens, more distinctly impressed in smaller holaspides. Interpleural furrows very narrow, lightly impressed, first three moderately deep immediately adjacent to rachial furrow. Fourth and fifth interpleural furrows faintly defined or obscure. Border narrow, defined by decrease in slope of pleura and obsolescence of pleural furrows. Fine granulation denser on border than that scattered over rachial rings and inner part of pleura. Posteromedian margin with small node or blunt point, lacking distinct mucro. O n t o g e n y : Washed marl samples from Lau Backar 1 have yielded some early ontogenetic material, including late meraspid stages. In an incomplete early holaspid cranidium 2.1 mm long (sag.) (Fig. 5A) the glabella is relatively longer than in adults. The size and position of the palpebral lobe is as in adults, although the postocular area is a little longer (exsag.). The SO and S1 apodemes are considerably shorter (tr.), and $2 reaches the rachial furrow. The granulation is very coarse. The earliest pygidium (estimated total width 1.6 mm; Fig. 5E) is of the penultimate or less likely the last meraspid degree (M9 or M10). There are nine pleural segments ending in backward directed spines. The anterior two inter-ring furrows are much longer than the following ones and than in undoubted small holaspid pygidia. Coarse granules are scattered over the whole surface. Early (but not earliest) holaspid pygidia (width 1.9 mm; Fig. 5F) still have deep inter-pleural furrows and coarse granulation, while the marginal spines are much reduced. At 2.3 mm width the inter-pleural furrows are becoming obscure (Fig. 5D), and at 3.0 mm width they are as weak as in adults. At this size the pygidial margin still carries three short, blunt pleural spines or angularities, but no posteromedian point is developed. R e m a r k s : SCHRANK(1972: 62--63) discussed the uncertainty with which some of the German Geschiebe specimens [Acastella cf. prima of SCHRANK(1972)] could be distinguished

Fig. 4. Acastellamadidipesn. sp. from Gotland, Hemse Beds, uppermost part, Gannor 1. A, B: Dorsal and dorsolateral views of partly exfoliated cranidium, note lack of genal spine, x 4; Ar54921.

272

Lars Ramsk61d& Gregory D. Edgecombe

from A. prima TOMCZYKOWA1962a. Some of SCHV,ANK'Smaterial is possibly conspecific with Acastella madidipes n. sp. (his. pl. 21, fig. 2), whereas other specimens are A. breviceps (ANG~LIN 1851; see synonymy below). The limited type material of Polish P~idoll (M. ultimus Zone; TOMCZYKOWA1991: 43) A. prima is even more poorly preserved than German specimens similar to A. madidipes. The additional material assigned to A. prima by TOMCZYiOWA ~ WI~VICKA(1974: pl. 1, figs. 11, 12) is immature, but recently well preserved material was figured by TOMCZYKOWA(1991). SCHV,ANK(1972) suggested that Acastella cf. prima might be defined by its angular corners of the frontal glabellar lobe, radiating rows of auxilliary impression tubercles on glabellar internal moulds, and short genal spine (versus "a nodule" in A. prima). The genal spine is variable in A. madidipes, ranging from a swelling like that of the largest specimens of A. prima to a slender spine. In most specimens, however, the genal angle is at least weakly spinose, and a distinct spine is retained by specimens larger than the largest known A. prima (ToMcZYKOWA1991: pl. 7, figs. 12, 13). In all figured cephala of A. prima $2 is confluent with the rachial furrow (ToMcZYKOWA1991: pl. 7, fig. 12, left end of left side $2 appears to be covered by matrix, right side $2 shallow adjacent to rachial furrow). This is of no special significance for the juvenile specimens of TOMCZYIIOWA& WlaXCICr,A (1974), since small specimens of A. madidipes (Fig. 2N) also show strong confluence of these furrows, as do juveniles of A. amatrix n. sp., for which large holaspides have $2 isolated abaxially. Confluence of $2 with the rachial furrow is clearly a juvenile trait, which is commonly (but not always) lost in adults. This applies not only to acastids, but to phacopids and dalmanitids as well (cf., e. g., ontogenetic series of Phacops spedeni CHATTERTON 1971: pl. 21; and Dalmanites myops in RAMSli6LD 1985a: pl. 10), and is due to the adaxial migration of the apodemes during ontogeny. In contrast to the condition retained also in adult A. prima, large specimens of A. madidipes n. sp. show the abaxial impression of $2 ranging from shallow (Fig. 2A, L) to obsolete (Fig. 3A, E). The coarsely granulate sculpture which is developed on one thoracopygidium of A. prima (ToMcZYKO~rA1962a: pl. 1, figs. 6, 7) is lacking in other pygida (ToMcZYIIOWA 1991: pl. 7, figs. 5--9), which are in this respect similar to A. madidipes. Acastella madidipes n. sp. is fairly rare in the uppermost Hemse Beds. In quantitative data assembled at the type locality it constituted less than 1% (4 specimens of a total of 471) of the trilobites (Acastella sp. of RAMSK~%D1985b: fig. 14A; horizon is Hemse beds top, not lower Eke beds). In the upper Eke Beds at Lau Backar 1 the species is equally rare; in quantitative data for 111 specimens from a single marl sample (RAMsK6LD1985b; fig. 14B) there were no specimens of this species, although nearly 20 adult specimens have been collected individually at the locality, and numerous juvenile specimens (Fig. 5A, D-G) occur in some marl samples. In addition to variation in length of the genal spine and width (tr.) of $2, Acastella madidipes n. sp. is variable in several additional cephalic characters. These include curvature of $2 (ranging from nearly straight to prominently convex forward), depth of $3 (shallow to moderately impressed), depth of the palpebral furrow, curvature of SO, and course of the cephalic posterior margin. Some specimens from the Hemse Beds (Fig. 2C, E) display a pronounced anterior flexure of the cephalic posterior margin towards the genal spine; the posterior border is longest (exsag.) well inward of the genal angle. The palpebral furrow is shallow, $2 is convex forward, and $3 is relatively shallow. This morphotype contrasts with specimens from the Eke Beds (Figs. 2A, 3A) in which the cephalic posterior border is less strongly flexed forward adaxial to the genal spine. In addition, in these the palpebral furrow and $3 are more deeply incised, and $2 is sometimes nearly straight (also Fig. 2L). The largest sample from a single locality is from Gannor 1, from which specimens of both general forms are known. The largest cranidium from this locality (Fig. 4) has angular genae that lack a spine. The palpebral furrow and $3 are weakly impressed, and the posterior border has maximum lengthening inward of the genal angle.

Silurian (Ludlow) acastid trilobites

273

There are small but consistent differences in the relative rachial width of pygidia of the three Swedish species of Acastella. In the earliest species, A. madidipes n. sp., the width of the first rachial ring is 34.5-36.5% of the total pygidial width. In the next younger species, A. breviceps (ANGELIN),this width is 36-38%, and in the youngest species, A. arnatrix n. sp., rachial width is 38-39% of the pygidial width. In other characters, the three species have remarkably similar pygidia. There are three pairs of deep apodemal pits and a shallower fourth in all specimens except one (Fig. 3A) which shows an additional, shallow pair of pits in the

Fig. 5. A, D-G. Acastella rnadidipes n. sp. from Gotland, Eke Beds upper part, Lau Backar 1. A: Dorsal view of incomplete cranidium, x 27; Ar54089. D: Dorsal view of incomplete pygidium, x 27; Ar54090. E: Exterior view of incomplete late meraspid transitory pygidium, x 27; Ar54091. F: Exterior view of pygidium, x 27; Ar54092. G: Dorsal view of incomplete pygidium, note pathological anterior part of rachis, x 22; Ar54093. B-C: Acastella amatrix n. sp. from Gotland, Hamra Beds unit b, Majstre 1. B: Exterior view of early holaspid pygidium, x 27; Ar54094. C: Dorsal view of smallest holaspid pygidium, x 27; Ar54095.

274

Lars Ramsk61d & Gregory D. Edgecombe

fifth inter-ring furrow. Otherwise inter-ring furrows posterior to the fourth lack apodemes. There are eight or nine well defined inter-ring furrows and a narrow (tr.) posterior one can usually be seen in well preserved specimens. This configuration of apodemes and rings is established already in young holaspides (Fig. 3F). To assess the relative spacing of rings in the rachis, the ratio of length (sag.) of the anterior seven rings to the total rachial length was measured. The ratio is between 81% and 84% in all three species of Acastella, with no consistent differences. The total variation thus is only three percent, and since the seventh ring itself is longer than this (i. e., approximately five percent of rachial length), most differences in the counts of the number of rachial rings in different specimens are due to preservation. The actual difference cannot be more than a single ring equal to three percent of rachial length, i.e., a very small posterior one. In fact, since orientation (anteroposterior tilt) affects the measurements, three percent may be within the error of measurement (although for each specimen the error is considerably less). The conclusion we draw from this is that, since the width of the rachis varies between the species while the spacing of the rings is constant, there is an underlying constraint. We ascribe this constraint to a fixed number of pygidial appendages (which we assume to be four). Table 1. Pygidial rachial width, length of seven anterior rings relative to total rachial length ('7 rings ratio'), and number of apodemal pit pairs for selected species of Acastidae.

Acaste downingiae Acaste subcaudata Acaste zerinae Acaste heberti elsana Acastella tiro A castella?artesia Acastocephala macrops

Rachial width/ pygidial width

7 rings ratio

Apod.emal pit pairs

36-38% 44% less than 32% 39% 36-38% 36 % 36-37%

80-82% 83% 83% 85% 85% 84% 84%

3+ 1 3 3+ 1 3 (+ 1?) 3? (more?) 3 ? (more ?) 3+ 1

A comparison with various other acastids reveals that the width of the pygidial rachis varies considerably, while the rachial ring spacing (as assessed for 7 rings) is constant (Table 1). The species have three pairs of deep apodemal pits, and a fourth pair of varying depth. This pattern is not substantially modified in the Asteropyginae; in the vast majority of taxa there are four pairs of apodemal pits indicating the presence of four appendage pairs in the pygidium. This is so even in most species with an elongated rachis with more numerous rachial rings [well seen in, e.g., Rhenops lethae of GANDL 1972, 7 rings ratio (cf. Table 1) 67-71%; the many Pseudocryphaeus species figured by MORZADEC 1971]. In some extreme forms the pattern is lost, for example in Psychopyge elegans (see MORZADEC 1988), showing 15-20 pairs of apodemal pits.

Acastetla cf. madidipes n. sp. Figs. 2H-K, 3G Material: Hemse Beds, top, or Eke Beds, base, N~is parish: Olsvenne 3, Ar54930-54931; Olsvenne 4, Ar54932; Bodudd 1, Ar54935-54937. R e m a r k s : Specimens from the Hemse/Eke boundary beds in the N~is peninsula conform to Acastella madidipes n. sp. in several of its diagnostic characters: eye size, posterior lowering of the visual surface, size of the glabellar median impression, and pygidial rachial width. However, they display some differences from typical Hemse/Eke specimens. The cephalic posterior margin has a nearly straight transverse course to a relatively long, slender genal spine

Silurian (Ludlow) acastid trilobites

275

(Fig. 2 H - K ) . Some specimens show $2 to be transverse and confluent with the rachial furrow, whereas in others it is narrow (tr.) and obsolete well inward of the rachial furrow. Pygidia from the N~is peninsula (Fig. 3G) differ from A. madidipes n. sp. in having the margin slightly concave near the posteromedian point, which itself is more strongly developed than in other Hemse/Eke specimens. We regard the N~is peninsula material as showing close affinity to the Hemse/Eke Beds material of A. rnadidipes, and differences may, at least tentatively, be attributed to variation at the population level.

A castella breviceps (ANGELIN 1851) Fig. 6A-M; Fig. 7A-J v* 1851 Phacopsbreviceps n. sp. - ANGELIN: 12, pl. 9, fig. 4a, fig. 4partita (cephalon only; pygidium is a proetid). 1882 PhacopsDowningiae Salt. [sic] - TULLBtRG: 13. 1883 PhacopsDowningiae Salt. [sic] - TI3LLBERG:231, 233. 1885 PhacopsDowningiae Murchison - LINDSTR6M: 42. 1888 PhacopsDowningiae Murch. - EICHSTaDT: 137, 153. 1888 PhacopsDowningiae Murch. - WIGAND:41, pl. 6, fig. 3. 1897 Phacopsbreviceps Ang. (= PhacopsDowningiae Murch.) - GR6NWAtL: 8. 1904 PhacopsDowningiae Murch. - TORNEBOHM & HENNIG: 55. 1909 PhacopsDowningiae Murch. - MOBrRG & GR6NWALL:71. 1910 Phacopsbreviceps ANG. -- WESTrRGkRD: 38. 1972 Acaste (Acastella?) cf. prima (Tomczykowa, 1952) - SCHRANII:62 [partita], pl. 20, fig. 8, pl. 21, figs. 3-5 only. 1987 Phacopsbreviceps -JEPPSSON& LAUFELD:8. D i a g n o s i s : Longitudinal furrow on posteromedian part of frontal glabellar lobe deep, short (sag.); eyes small, length 3 6 - 4 0 % of cephalic length; visual surface with parallel upper and lower margins; genal spine short, thorn-like; first rachial ring 3 6 - 3 8 % of pygidial width. L e c t o t y p e : Selected here; incomplete, water-worn cephalon Ar47657 (Fig. 6L-M), from Klinta at Ringsj6n in Scania, Sweden; Klinta Formation, Ludfordian (Upper Ludlow). S y n t y p e : Worn pygidium Ar47658 (Fig. 8M), from the type locality. The specimen is a proetid. O t h e r Scanian m a t e r i a l : Klinta Formation, "Ringsj6n" (probably Klinta): complete exoskeleton Ar32914 (Fig. 6I); Bj~irsj61agfird:cephalon SGU Type 8471 (Fig. 6J-K), pygidium SGU Type 8472 (Fig. 7D-E), plus numerous unnumbered cephala and pygidia (SGU collections). R e m a r k s : ANGELIN'S(1851) figure of a complete, extended exoskeleton is a composite reconstruction based on at least two specimens. The cephalon here selected as lectotype is distorted in such a way (e. g., the eyes protruding anteriorly in dorsal view) that enables its certain identification as the cephalon of ANGELIN'S drawing. In addition, ANGELIN'Soriginal, handwritten label for the specimen is preserved. Stored together with this cephalon, under the same label, is a syntype pygidium (Fig. 8M). This pygidium, which happens to be that of a proetid, is too large to fit with the cephalon. In articulated acastids the pygidium is relatively much smaller than in ANGELIN'Sdrawing. The large pygidium seen in the drawing makes it certain that ANGrLIN did not have a complete specimen at his disposal, but reconstructed one from the parts available. Already LINDSTR6M (1885: 42) stated that in the Riksmuseum there was no complete specimen such as the one figured by ANGELIN. A complete specimen has, however, been collected sometime later in the nineteenth century (Fig. 6I). G o t l a n d material: From the Eke/Burgsvik boundary bed(s) at Ronehamn, Ar31081, Ar31095-31096, Ar31403, Ar54969-54973. From Burgsvik Beds (Burgsvik Sandstone): Burgsvik, Ar31097, Ar31410, Ar31230, Ar31490, Ar31496, SGU Type 8473-8476; Oja, Ar31098-31119; Gr6tlingbo, Gansviken, Ar55241; Gr6tlingbo, south of Gr6tlingboudd, Ar31085, Gr6tlingbo, quarry NE of Rovalds, Ar52273-52275; Vamlingbo, Sn~ickvik, Ar31126-31127; Hoburg, Ar31084; Husryggen 1, Ar54968. D e s c r i p t i o n : This species is similar in many respects to Acastella rnadidipes n. sp., and only features characteristic for A. breviceps need to be described here: Cephalic length 4 5 - 5 2 % of width. Glabellar length 9 0 - 9 3 % width across frontal lobe. Cephalic rachial furrow usually relatively strongly divergent. $2 and $3 commonly widened. Frontal lobe auxilliary impression system of triangular acastacean arrangement, four linear rows of small impressions radiating anterolateraUy from longitudinal median furrow. Latter

276

Lars Ramsk61d & Gregory D. Edgecombe

composed of pair of elongate pits or short (sag.) groove. Width of LO 67-71% width of frontal lobe. Anterior margin of eye opposite anterolateral corner (exsag.) of L3; posterior margin of eye opposite posterolateral edge of L1. Visual surface of 24-26 dorsoventral lens files of up to eight lenses per file; lens formula of largest eye (Fig. 6B-C) from anterior: 345 678 788 878 778 787 777 654 31. Pygidial length 57-66% of width. Seven anterior rachial rings 80.8-84.0% of the rachial length. Nine to ten defined rachial rings. Lateral margin gently, evenly convex outward, weakly concave near posteromedian point. D i s c u s s i o n : The locality Klinta at Ringsj6n indicates that the Scanian type material is from the Klinta Formation, of Ludfordian (Upper Ludlow) age. ANGEL:N'S locality Klinta is still the main locality for the Klinta Formation at Ringsj6n. In this formation, but nowhere else in the Silurian sequence of Scania, an acastid occurs, locally abundant [e.g., at Bj:irsj61ag~rd 2 of LARSSON (1979)]. The Klinta Formation ranges from the late Leintwardinian Polygnathoides siluricus Zone (i. e., in the Neobeyrichia lauensis range) to the late Whitcliffian or possibly early P~fdolhn. The formation is thus equivalent to the entire part of the Gotland sequence from the uppermost Hemse Beds and up. ANGELIN'S Klinta material may be from the Bj~irsj6 Member (JEH'SSON ~ LAUFELD 1987), equivalent to the middle and upper parts of the Eke Beds on Gotland. The strata at Bj~irsj61agfird 2 are slightly younger, equal in age to parts of the Burgsvik Beds. Acastid specimens from Klinta and Bj~irsj61agfird all appear to belong to a single species. The Gotland material agrees in all morphological aspects with the Scanian specimens. Measurable characters such as relative eye length (Scania: 36-39% of cephalic length; Gotland: 36-40%) and pygidial rachial width (Scania: 37%; Gotland: 36-38%) overlap completely between the two sets of specimens. The deep $2 and $3 and the short, deep longitudinal glabellar furrow, short genal spine, course of the facial suture posterolaterally, and dense, coarse granulation are further characters that are also shared between the Scanian and Gotland samples. The Geschiebe material figured by SC~Ir,ANK (1972) is poor compared to the specimens from Gotland and Scania. Only a few specimens (see synonymy list above) are referred here to Acastella breviceps. The basis for their assignment is that the morphology is compatible with an A. breviceps assignment, supported by the occurrence of the specimens in "Phacitensandstein", equivalent to the Burgsvik sandstone. Acastella breviceps bears closest comparison to A. madidipes n. sp. It is most readily distinguished by its prominent longitudinal glabellar furrow, and parallel upper and lower margins of the visual surface (versus anteriorly taller visual surface). Subtler differentia from A. madidipes include a more strongly divergent rachial furrow, relatively wide glabellar furrows, slightly shorter eye, and wider pygidial rachis. Some specimens of A. breviceps have posterior pygidial inter-ring furrows that are gently convex backward, but in A. madidipes they are straight and transverse.

Fig. 6. Acastella breviceps (ANGEL:N 1851). Gotland, uppermost Eke Beds (B-C, H), Burgsvik Beds (A, D-G); Scania, Klinta Formation (I-M). A: Dorsal view of partial, smallest cranidium, Burgsvik, old collection, x 8; Ar31110. B-C: Anterolateral and dorsal views of incomplete cephalon, Ronehamn 2, x 3.5; Ar54970. D-G: Anterior, ventral, dorsal, and lateral views of large, incomplete cephalon, Burgsvik, old collection, x 3; Ar31490. H: Dorsal view of incomplete cephalon, Ronehamn 2, x 4; Ar54971. I: Exterior view of complete exoskeleton, "Ringsj6n" (probably Klinta), old collection, x 3.5; Ar32914. J-K: Dorsal and lateral views of crushed cephalon, Bjiirsj61ag~rd, collected by G. VONSCHMALENSEE,X 4; SGU Type 8471. L-M: Dorsal and lateral views of incomplete, water-worn lectotype cephalon, Klinta at Ringsj6n, collected by N. P. ANGELIN, X 3.5; Ar47657.

Silurian (Ludlow) acastid trilobites

277

278

Lars Ramsk61d & Gregory D. Edgecombe

Fig. 7. Acastella breviceps (ANGELIN1851). Gotland, uppermost Eke Beds (H-J), Burgsvik Beds (A-C, F, G); Scania, Klinta Formation (D-E). A: Dorsal view of abraded cephalon, Burgsvik, collected by G. voN SCnMALENSEe1880, X 3.5; SGU Type 8473. B: Dorsal view of cephalon, both genal angles damaged, Burgsvik, old collection, x 3; Ar31410. C: Dorsal view of pygidium, Burgsvik, collected by G. YON SCHMALENSEE1880, X 4.5; SGU Type 8474. D-E: Dorsal and posterior views of slightly distorted pygidium, Bj/irsj61ag~lrd,collected by G. VON SCHMALENSEE, X 6; SGU Type 8472. F - G : Lateral anddorsal views of large pygidium, Burgsvik, collected by G. VONSCHM^LENSEE1880, X 4.5; SGU Type 8475. H-J: Posterior, lateral, and dorsal views of pygidium, note denticulate margin, Ronehamn, old collection, x 6; Ar31403.

Silurian (Ludlow) acastid trilobites

279

Acastella amatrix n. sp. Fig. 5B, C; Fig. 8A-L D i a g n o s i s : Cephalic rachial furrow relatively weakly divergent; $2 and $3 lightly impressed; band of granules along anterior edge of L1; eye large, length 41-45% of cephalic length; posterior margin of palpebral lobe opposite posterolateral edge of SO; palpebral furrow shallow; palpebral area relatively inflated; cephalic lateral border furrow very shallow; genal angle with blunt point, no spine; first rachial ring 38-39% of pygidial width. Holotype: Cephalon Ar54974 (Fig. 8C-E), from the Hamra Beds, unit b, Majstre 1 (defined in LAI3FELD1974a), Sundre parish, Gotland. Other material: From Hamra Beds, unit b; Majstre 1, Ar52352, Ar54094-54095, Ar54975-54990. Hamra Beds, unit c; Sundre parish: Barshageudd, Ar54994-54997; Hoburg, Klev (Juves), Ar31086-31094, Ar54991-54993; Hoburg, Ar31082. Burgsvik Sandstone; Hoburg, Ar31392. Etymology: Latin amatrix, mistress, feminine form of mister, alluding to the type locality Majstre, meaning master. R e m a r k s : Cephalic morphology of Acastella amatrix n. sp. is generally similar to Acaste dayiana RICHTER & RICHTER 1954 (compare Fig. 8F and Sr 1970; pl. 3, fig. 1). We presently accept the synonymy of material from the P~fdolf of Poland (ScHRANK 1970; TOMCZYKOWA 1991; TOMCZXKOWA & WITXCICKA 1974), Podolia (BALASHOVA 1968), and North German erratics (ScHRANK 1972) with the Rhineland types of A. dayiana. RICHTER & RIC~tTE~ (1954) cited the type material of A. dayiana as Ludlow in age, but S~tERGOLO(1968: 28) cited evidence that it is from Lower Gedinnian strata. The blunt genal angle of Acastella amatrix is as developed in Acaste dayiana, and distinguishes it from other species of Acastella. However, the largest specimens of Acastella madidipes n. sp. indicate that a non-spinose genal angle is also attained by at least some individuals of that species late in their ontogeny (Fig. 4). Large holaspides of A. madidipes also approach A. amatrix more closely than do small specimens in the shallowing of $3 and the palpebral furrow. Small holaspides of A. matrix (Fig. 8J) more closely resemble most cephala of A. madidipes (see remarks below). It is apparent that the distinctive, in some respects Acaste-like, cephalic morphology of A. amatrix can be produced by terminal addition to the ontogeny of stratigraphically earlier species of Acastella such as A. madidipes. Furthermore, A. amatrix possesses certain apomorphies of Acastella that suggest affinities to A. madidipes rather than to Acaste. Shared derived characters appearing in holaspid ontogeny include a pointed anterior cephalic margin, steeply inclined $3, and a pygidial posteromedian node. We thus regard A. amatrix as an advanced species of Acastella rather than assigning it to Acaste on the basis of its nonspinose genae. Acastella amatrix n. sp. also differs from Acaste dayiana in its longer eye, that extends back to the posterolateral margin of SO, versus terminating opposite L1 in specimens of A. dayiana figured by SCHRANK(1970, 1972) and TOMCZWOWA (1991). Acastella amatrix has a pointed (versus rounded) cephalic margin, shallower $2 and $3, weaker palpebral furrow, finer granulation, and a distinct pygidial posteromedian node. The hypostome of Acastella amatrix n. sp. (Fig. 8G, K) is the first known for a Silurian species of the genus. It compares closely with that of Acaste downingiae (SHERGOLD 1966: pl. 28, fig. 16) and Devonian Acastella tiro (BALASHOW 1968: pl. 3, fig. 13; GANDL 1972: pl. 3, fig. 4; TOMCZXKOWA 1991: pl. 12, figs. 1, 2) in most respects, e.g., a moderately convex hypostomal suture; transverse posteromedian margin of the border; maculae just behind midlength (exsag.) of the middle body. A small pair of posterolateral denticles is shared with A. tiro (ToMczYKOWA1991: pl. 12, fig. 2) and A. latimarginata (GANDL 1972: pl. 4, fig. 1), but is developed as blunt angulations in Acaste downingiae. The middle furrows are short (exsag.) and extend inward only to the maculae. However, their incision is deeper than in Acaste, being more similar to their distinct impression in Acastella latimarginata. These differences with

280

Lars Ramsk61d & Gregory D. Edgecombe

Acaste indicate that hypostomal characters can be used in the diagnosis of Acastella, which would include A. amatrix. Juvenile cranidia (Fig. 8J, L) provide evidence for the following ontogenetic transformations: $2 becomes discontinuous with the rachial furrow and convex forward (straight in juvenile); change from rounded to pointed anterior cephalic margin; reduction of the genal spine; shallowing of $2 and $3; relative widening of the occipital ring; shallowing of the palpebral furrow; decrease in the relative size of granules composing the sculpture, notably on the palpebral lobe. Juvenile pygidia have well impressed inter-pleural furrows, and a course granulation. Two small specimens (width 1.3 and 1.7 mm; Fig. 5B and 5C) have thorn-like marginal spines decreasing in size posteriorly. At 2.3 mm width the most anterior spine is still well defined and the next two form angular protrusions on the margin, while posterior to this the pygidial margin is only weakly lobulate. The pygidial spines are thus lost at approximately the same absolute size as is the case in Acastella madidipes n. sp. A distorted cephalon (Fig. 8A-B) from the Burgsvik Sandstone is probably a conspecific of Acastella amatrix n. sp., which is otherwise known exclusively from the Hamra Beds. The specimen conforms to A. amatrix in the posterior course of the facial suture, course of the posterior border furrow, the rounded genal angle, eye size, shallow palpebral furrow, similar course and depth of the lateral glabellar furrows, and sculpture (e. g., granules along the anterior edge of L1). Acastella amatrix n. sp. is very common in the Hamra Beds, unit b. In quantitative data assembled at the type locality it constituted 25% (66 specimens of a total of 260) of the trilobites (Acastella sp. of RAMSKOLD 1985b: fig. 15B). .Genus Ewacaste n. gen. Type species: Acaste (Scotiella) quaesita SCHRAmt1972. Assigned species: Scotiella opatowiensis TOMCZYKOWA1962b; Scotiella samsonowiczi TOMCZYKOWA1962b; Ewacaste cf. quaesita. Etymology: Compounding Ewa, in honour of EXaATOMCZVKOWA,and Acaste GOLDFUSS1843. D i a g n o s i s : Eye relatively short (exsag.), postocular fixigena prominent; distal end of $3 distinctly anterior to eye; $3 weakly impressed; cephalic anterior margin rounded in dorsal view; L1 long (exsag.); pygidial posterior margin rounded or with weak posteromedian point, lacking independent node or mucro; border weakly defined. R e m a r k s : Ewacaste n. gen. accommodates Baltic Ludlow and P~idoll species that have formerly been assigned to Scotiella DELO 1935. Scotiella is here restricted to its type species, S. logani (HALL1860) (see SHERGOLD 1967b; ELDREDGE & BRANIgA 1980: fig. 21). As noted above, this species and Dalrnanitina logani var. conservatrix McLEARN 1924; share apomorphic characters with Acastella, notably shortened L1 and a strong pygidial mucro.

Fig. 8. A-L: Acastella amatrix n. sp. from Gotland. Hoburg, Burgsvik Beds (A-B); Majstre 1, Hamra Beds unit b (C-L). A-B: Oblique anterolateral and dorsal views of crushed cephalon, old collection x 4; Ar31392. C-E: Anterior, dorsal, and lateral views of holotype cephalon, x 4; Ar54974. F: Dorsal view of incomplete cranidium, x 6; Ar54975. G: Exterior view of incomplete librigena and hypostome, x 6; Ar54978, Ar54979. H: Dorsal view of pygidium, x 6; Ar54981. I:Dorsal view of pygidium, x 7;Ar54982. J: Dorsal view of incomplete cranidium, x 5; Ar54977. K: Ventral view of hypostome, x 6; Ar54980. L: Dorsal view of small cranidium, x 8; Ar54976. M: Proetid gen. et sp. indet, from Scania, Klinta at Ringsj6n, Klinta Formation. Dorsal view of badly worn pygidium, syntype of Phacops breviceps ANGELm 1851, collected by N. P. ANGELIN,X 4; Ar47658.

Silurian (Ludlow) acastid trilobites

281

282

Lars Ramsk61d & Gregory D. Edgecombe

The assignment of Late Silurian species from Poland (ToMcZYKOWA1962b, 1991; SCHRANK 1970) and North German Geschieben (ScHRANK 1972) tO Scotiella has emphasized the obsolescence of $2 and $3 as a defining character. SrtERGOLD(1967b: 4) shifted emphasis to long glabellar furrows, a reduced preglabellar furrow, and short genal mucronations in order to accommodate certain species with $2-$3 deeply impressed, as typifies Acastella (i.e., conservatrix McLEARN). These differentia from Acastella are not applicable to all European species of "Scotiella'. The preglabellar furrow of "Scotiella" quaesita (ScHe~NK 1972: pl. 21, fig. 8, 8A; TOMCZYIiOWA1991: pl. 7, fig. 22) is strongly incised abaxially. It is not apparent that Scotiella logani has glabellar furrows any longer than Acastella; "long glabellar furrows" appears to accurately describe only Acastella conservatrix (SHERGOLD1967b: pl. 1, fig. 1) and some specimens of "Scotiella" samsonowiczi ( S c r t ~ I ( 1972: pl. 21, fig. 6). European "Scotiella" cannot be characterized by having shorter genal mucronations than Acastella, because "S". sarnsonowiczi has genal spines that are at least as long as those of Acastella spinosa (and longer than all Gotland Acastella) and considerably stouter. In short, assignment of Baltic taxa to Scotiella hinges on whether or not effacement of $2-$3 is synapomorphic with S. logani. Ewacaste n. gen. shares certain character states with Acaste for which Scotiella is more similar to Acastella. Notably, L1 is long in all species of Ewacaste, whereas Scotiella logani (SrIERGOLD 1967: pl. 3, fig. 1; ELDI~EDGE~ BRANI~A 1980: fig. 21A-B) has L1 considerably shortened (exsag.), as in Acastella spinosa and A. conservatrix. Long L1 is plesiomorphic for Acastidae (e. g., Llandovacaste), so the resemblance between Acaste (see RICHTER ~ RICHTER 1954: pl. 3, figs. 37a, 38a for specimens with especially long L1) and Ewacaste is not positive evidence for relationship. It does, however, split Scotiella + Acastella from a (plesiomorphic) group including Acaste and Ewacaste (Fig. 1B). Likewise, the pygidium of Ewacaste n. gen. is distinguished from Scotiella and Acastella by the absence of a posteromedian node or mucro. Ewacaste quaesita (ScrtRANK 1972: pl. 21, fig. 9; TOMCZYI~OXVA1991: pl. 7, fig. 24) has a rounded pygidial margin similar to that of Acaste and allied genera of the "Wenlock group". The pygidial margin is faintly pointed in an articulated specimen of Ewacaste samsonowiczi (ScHRANK 1970: pl. 3, fig. 6). This differs from pygidia used in the original diagnosis of E. sarnsonowiczi (ToMcZYKOWa 1962b: pl. 34, figs. 5, 6), which bears a distinct, albeit short, mucro. These pygidia are of Acastella type, and they were recently reassigned to Acastella spinosa, a species co-occurring with E. sarnsonowiczi, by TOMCZYKOWa(1991: 35). A new pygidium for "S." sarnsonowiczi was figured by ToMczvIioxva (1991: pl. 8, fig. 9). This pygidium has a weak posteromedian point, and is similar to SCHRANII'S(1970) specimen. We accept the assignment. Ewacaste quaesita (and Ewacaste cf. quaesita, described below) further resembles Acaste in the absence of a genal spine; the genal angle is pointed. A strong genal spine is developed only in stratigraphically slightly later species (E. samsonowiczi). The interpretation that Ewacaste is a line of descent separate from Acastella/Scotiella and instead closely related to Acaste (the "Wenlock group"), finds some support from stratigraphic distribution. The most primitive (Acaste-like) species of Ewacaste are Wenlock or Ludlow (the type material of E. quaesita) or Ludfordian (Ewacaste cf. quaesita), and range into the lower P.~idoli (Polish E. quaesita), while more highly derived species are P~idollan. Ewacaste samsonowiczi differs from other species assigned to the genus, including the co-occurring E. opatowiensis, in that the glabella is angular anteromedially, as in Acastella. This implies either a topologically basal position for this species (if Acastella/Scotiella is the closest relative of Ewacaste), or a convergent similarity to Acastella in this single species (if Ewacaste is a separate part of the "Wenlock group", as suggested here).

Silurian (Ludlow) acastid trilobites

283

Ewacaste cf. quaesita (ScrtRANK 1972) Fig. 3 I - L Material: Uppermost Eke Beds (Ludfordian), immediately below the Eke/Burgsvik boundary, Ronehamn, Rone parish, Gotland: cephalon Ar31402. R e m a r k s : A m o n g Gotland specimens, a partially exfoliated cephalon from the Eke Beds at Ronehamn is similar to Acastella breviceps, occurring in the same stratigraphic interval, in such characters as course of the posterior branch of the facial suture and eye length. The latter (36% of cephalic length) is at the lowest range of variation in Acastella breviceps (36-40%). The Ronehamn cephalon is distinguished from A. breviceps by its shallow $3 (seen on the cuticle surface abaxially), long (exsag.) L1, and lack of a genal spine. These three characters, as well as the short eye and weakly developed surface sculpture, suggest an assignment to Ewacaste n. gen. Closest comparison can be made between the Gotland Ewacaste and E. quaesita (Scrtv,AN~). In both the genal angle is pointed, without a spine, and the preglabellar furrow is sharply incised. The genal angle is slightly more pointed in the Geschiebe holotype of E. quaesita, but the difference may be ontogenetic (the Gotland specimen is slightly larger). In the stratigraphically later Polish E. quaesita (ToMczYifowA 1991; pl. 7, fig. 21) the genal angle is similar to in the Gotland specimen. Given the limited material presently available, it is plausible that the three occurrences may be conspecific.

Acknowledgements We thank the journal's referees for helpful comments on the manuscript. L. R. thanks O. H. WALLISEi~and L. ALBERTI, G6ttingen, for permission and help in transferring the late H. ALBERTI'S Gotland trilobite material, including some acastids, to Riksmuseum. This is Contribution no. 4 from the Institute of Taxomorphosis.

References ALBEI~Tt,H.; ALI3ERTI,L.; MKNNIL,R. & TOMCZVKOWA,E. 1982. Correlation of the Proetus signatus level (upper Silurian) in Gotland, Poland and the East Baltic area. - Eesti NSV Teaduste Akadeemia Toimetised, Geoloogia 31: 29-32, Tallinn. ANGEHN, N. P. 1851. Palaeontologia Svecica. I: Inconographia crustaceorum formationis transitionis. Fasc. 1. - 1-24, pls. 1-24, Holmiae. Ax, P. 1985. Stem species and the stem lineage concept. - Cladistics 1: 279-287, London. BALASHOVA,E. A. 1968. Trilobity skal'skogo i borshchovskogo gorizontov Podolii. [In:] BALASrtOV,Z. G. (ed.) Siluriysko-devonskaya fauna Podolii. [The Silurian-Devonian fauna of Podolia]: 95-123, pls. 1-3, Izdatel'stvo Leningradskogo Universiteta, Leningrad. BASSE'rr, M. G. & CocKs, L. R. M. 1974. A review of Silurian brachiopods from Gotland. - Fossils and Strata 3: 1-56, pls. 1-11, Oslo. B~tJTON, D. L.; J~NussoN, V.; OWENS, R. M.; SIVETER, D. J. & TRIvv, R. P. 1979. Trilobites. [In:] J~NussoN, V., LAUFELD,S. & SKOCLtJND,R. (eds.) Lower Wenlock faunal and floral dynamics Vattenfallet section, Gotland. - Sveriges Geologiska Unders6kning Serie C 762: 116-120, Uppsala. CrtAvrERTON, B. D. E. 1971. Taxonomy and ontogeny of Siluro-Devonian trilobites from near Yass, New South Wales. - Palaeontographica Abt. A 137: 1-108, pls. 1-24, Stuttgart. CLA~KSON,E. N. K. 1966. The life attitude of the Silurian trilobite Phacopsmusheni SALTER1864. -- Scottish Journal of Geology 2: 76-83, pl. 1, Edinburgh. DELO, D. M. 1935. A revision of the phacopid trilobites. -Journal of Paleontology 9: 402-420, Menasha, Wisconsin. EDCECOMBE, G. D. 1993. Silurian acastacean trilobites of the Americas. - Journal of Paleontology 67: 535-548, Lawrence, Kansas. EICrtSTKDT, F. 1888. Anteckningar om de yngsta 6fversiluriska aflagringarna i Sk~.ne. - Geologiska F6reningens i Stockholm F6rhandlingar 10: 132-156, Stockholm.

284

Lars Ramsk61d & Gregory D. Edgecombe

ELDREDGE,N. 1979. Cladism and common sense. [In:] CRACRArr,J. & ELDREDGE, N. (eds.) Phylogenetic Analysis and Paleontology: 165-198, Columbia University Press, New York. ELDREDGE, N. & BRANIgA,L. 1980. Calmoniid trilobites of the Lower Devonian Scaphiocoelia zone of Bolivia, with remarks on related species. - Bulletin of the American Museum of Natural History 165: 181-290, New York. GANDL,J. 1972. Die Acastavinae und Asteropyginae (Trilobita) Keltiberiens (NE-Spanien). - Abhandlungen der Senckenbergischen Naturforschenden Gesellschaft 530: 1-184, pls. 1-15, Frankfurt am Main. GOLDFUSS, A. 1843. Systematische Ubersicht der Trilobiten und Beschreibung einiger neuen Arten derselben. - Neues Jahrbuch fiir Mineralogie, Geognosie, Geologie und Petrefaktenkunde. 1843: 537-567, pls. 4-6, Stuttgart. GR6NWALL, K. A. 1897. C)fversikt af Skfines yngre 6fversiluriska bildningar. - Geologiska F6reningens i Stockholm F6rhandlingar 19: 188-244, Stockholm. HAAS,W. 1968. Trilobiten aus dem Silur und Devon yon Bithynien (NW-Tiirkei). - Palaeontographica Abt. A. 130: 60-207, pls. 26-37, Stuttgart. HALL, J. 1860. Descriptions of new species of fossils from the Silurian rocks of Nova Scotia. - Canadian Naturalist and Geologist 5: 144-169, Ottawa. HARRINGTON,H. J.; MOORE, R. C. & STUBBLEFIELD,C.J. 1959. Morphological terms applied to Trilobita. [In:] MOORE, R. C. (ed.) Treatise on Invertebrate Paleontology. Part O, Arthropoda 1: 117-126, Geological Society of America and University of Kansas Press, Lawrence, Kansas. HEDE, J. E. 1921. Gottlands Silurstratigrafi. - Sveriges Geologiska Unders6kning Serie C 305: 1-100, Stockholm. HOLLARD,H. 1963. Les Acastella et quelques autres Dalmanitacea du Maroc pr&aharien. Leur distribution verticale et ses cons6quences pour l'&ude de la limite Silurien-D&onien. - Notes et M6moires du Service G6ologique du Maroc 176: 1-66, pls. 1-4, Rabat. Jt~NussoN, V. 1956. On the trilobite genus Celmus Angelin, 1854. - Bulletin of the Geological Institutions of the University of Uppsala 36: 35-49, pl. 1, U..ppsala. JEPPSSON, L. & LAUFELD,S. 1987. The Late Silurian (Sved-Ramsfisa Group in Skfine, south Sweden. Sveriges Geologiska Unders6kning Serie Ca 58 [dated 1986]: 1-45, Uppsala. LARSSON,K. 1979. Silurian tentaculitids from Gotland and Scania. - Fossils and Strata 11: 1-180, Oslo. LAUFELD,S. 1974a. Reference localities for palaeontology and geology in the Silurian of Gotland. - Sveriges Geologiska Unders6kning Serie C 705: 1-172, Stockholm. 1974b. Silurian Chitinozoa from Gotland. - Fossils and Strata 5: 1-130, Oslo. LINDSTR6M, G. 1885. F6rteckning p$ Gotlands siluriska crustac6er. - Ofversigt af Kongliga VetenskapsAkademiens F6rhandlingar 42 (6): 37-99, pls. 12-16, Stockholm. McLEARN, F. H. 1924. Palaeontology of the Silurian rocks of Arisaig, Nova Scotia. - Memoir of the Geological Survey of Canada 137: 1-180, pls. 1-30, Ottawa. MOBERG, J. c. & GRONWALL,K. A. 1909. Om Fyledalens Gotlandium. - Meddelanden frfin Lunds Geologiska F~iltklubb, B 3: i-x, 1-86, pls. 1-6, Lund. MORZADEC, P. 1971. Quelques Asteropyginae (trilobites, Phacopacea) des schistes et calcaires ~ Atbyris undata (D&onien inf&ieur) du Massif armoricain. - Palaeontographica Abt. A 138: 166-184, pls. 15-19, Stuttgart. MORZADEC,P. 1988. Le genre Psychopyge (Trilobita) dans le D6vonien inf6rieur du Nord de l'Afrique et l'ouest de l'l~urope. - Palaeontographica Abt. A 200: 153-161, pls. 1-2, Stuttgart. MUNTHE, H. 1902. Stratigrafiska studier 6fver Gotlands silurlager. - Geologiska F6reningens i Stockholm Fbrhandlingar 24: 221-237, Stockholm. MUNTHE,'H.; HEDE, J. E. & LUNDQVIST,G. 1929. Beskrivning till kartbladet Katthammarsvik. - Sveriges Geologiska Unders6kning Serie Aa 170: 1-120, Stockholm. MUNTHE, H.; HEDE, J. E. & YON POST, L. 1925. Gotlands geologi. En 6versikt. - Sveriges Geologiska Undersbkning Serie C 331: 1-130, Stockholm. 1927.Beskrivning till kartbladet Hemse. - Sveriges Geologiska Undersbkning Serie Aa 164: 1-155, Stockholm. MURCHISON, R. I. 1839. The Silurian System, founded on geological researches in the counties of Salop, Hereford, Radnor, Montgomery, Caermarthen, Brecon, Pembroke, Monmouth, Gloucester, Worcester, and Stafford; with descriptions of the coal-fields and overlying formations. - xxxii + 768 pp., 37 pls., John Murray, London. PILLET,J. 1962. Contribution a l'&ude des faunes de Trilobites du Zemmour (Mauritanie septentrionale). Ann. Fac. Sci. Univ. Dakar 1961 6: 93-118, 3 pls., Dakar. RAMSKbLD, L. 1985a. Silurian phacopid and dalmanitid trilobites from Gotland. - Stockholm Contributions in Geology 40: 1-62, 12 pls., Stockholm. 1985b. Studies on Silurian trilobites from Gotland, Sweden. - 24 pp., Department of Geology, University of Stockholm and Department of Palaeozoology, Swedish Museum of Natural History, Stockholm. -

Silurian (Ludlow) acastid trilobites

285

RrrD, F. R. C. 1925. Some new Silurian trilobites. - Geological Magazine 62: 67-76, pl. 2, London. RICHTER, R. & RICHTER, E. 1954. Die Trilobiten des Ebbe-Sattels und zu vergleichende Arten (Ordovizium, Gotlandium/Devon). - Abhandlungen der Senckenbergischen Naturforschenden Gesellschaft 488: 1-76, pls. 1-6, Frankfurt am Main. SALTER,J. W. 1864. A monograph of the British trilobites from the Cambrian, Silurian and Devonian formations. (1).- Palaeontographical Society, Monographs: 1-80, pls. 1-6, London. SCHI~NK, E. 1970. Die Trilobiten des Silurs der Bohrung Leba 1 (Ostseekiiste der VR Polen). - Berichte der deutschen Gesellschaft der geologischen Wissenschaften. A. Geologie, Paliiontologie 15: 573-586, pls. 1-3, Berlin. 1972. Proetacea, Encrinuridae und Phacopina (Trilobita) aus silurischen Geschieben. -Geologie, Beiheft 76: 1-117, pls. 1-21, Berlin. SHERGOLD,J. H. 1966. A revision of Acaste downingiae (Murchison) and related trilobites. - Palaeontology 9: 183-207, pls. 28-32, London. 1967a.A reappraisal of the North American species of the Siluro-Devonian trilobite genus Scotiella. Postilla Peabody Museum of Natural History 112: 1-20, pls. 1-3, New Haven, Connecticut. 1967b.A revision of Acastella spinosa (Salter 1864) with notes on related trilobites. - Palaeontology 10: 175-188, pls. 24-25, London. - 1968.•nthe•ccurrence•fthetri••bitegeneraAcasteandAcastellainVict•ria.-Pr•ceedings•fthe Royal Society of Victoria 81: 19-30, Melbourne. SMeENK, Z. 1983. Devonian trilobites of the southern Cantabrian Mountains (northern Spain) with a systematic description of the Asteropyginae. - Leidse Geologische Mededelingen 52: 383-511, pls. 1-35, Leiden. STRUVE, W. 1958. Acastavinae n. subfam. [Beitr~ige zur Kennmis der Phacopacea (Trilobita), 2]. Senckenbergiana lethaea 39: 221-226, Frankfurt am Main. 1959. Suborder Phacopina Struve, nov. [In:] Mool~E, R. C. (ed.) Treatise on Invertebrate Paleontology. Part O, Arthropoda 1: 461-495, Geologial Society of America and University of Kansas Press, Lawrence, Kansas. THOMAS, A. T.; OWENS, R. M. & RUSI-rTON,A. W. A. 1984. Trilobites in British stratigraphy. - Special Report of the Geological Society of London 16: 1-78, London. TIMM, J. 1981. Zur Trilobitenstratigraphie des Silur/Devon-Grenzbereiches im Ebbe-Antiklinorium (Rheinisches Schiefergebirge). - Mitteilungen Geologisch-Pal~iontologisches Institut Universitiit Hamburg 50: 91-108, Hamburg. TOMCZYKOXVA,E. 1962a. O rodzaju Scotiella Delo z warstw rzepinskich G6r Swietokrzyskich. [On the genus Scotiella Delo from the Rzepin Beds of the Holy Cross Mountains]. [In:] PASSENDOV,~ER,E. (ed.) Ksiega pamiatkowa ku czci profesora Jana Samsonowicza: 187-205, pls. 34-35, Polska Akademia Nauk, Komitet Geologiczny, Wydawnictwa Geologiczne, Warszawa. [In Polish with Russian and English summaries.] 1962b. O trylobicie Acastella prima n. sp. [Notes on the trilobite Acastella prima n. sp.]. Kwartalnik Geologiczny 6: 260-266, pl. 1, Warszawa. [In Polish with Russian and English summaries]. 1991. Upper Silurian and Lower Devonian trilobites of Poland.-Prace Panstwowego Institutu Geologicznego 134: 1-62, 16 pls. Warszawa. TOMCZYKOWA,E. ~ WITWICr,A, E. 1974. Stratigraphic correlation of the Podlasian deposits on the basis of ostracodes and trilobites in the Peri-Baltic area of Poland (Upper Silurian). - Biuletyn Institut Geologizno 276: 55-84, pls. 1-3, Warszawa. TULLBEaG, S. A. 1882. Sk~nes graptoliter. 1. Allm~in 6fversigt 6fver de siluriska bildningarna i Sk~me och jemf6relse reed 6friga k~inda samtidiga aflagringar. - Sveriges Geologiska Unders6kning Serie C 50: 1-44, Stockholm. 1883. Ueber die Schichtenfolge des Silurs in Schonen, nebst einem Vergleiche mit anderen gleichalterigen Bildungen. - Zeitschrift der deutschen geologischen Gesellschaft 35: 223-269, pl. 10, Berlin. T6RNEBOttM, A. E. ~ HENNIG, A. 1904. Beskrifning till Blad 1 & 2 omfattande de topografiska kartbladen Landskrona, Lund, Kristianstad, Maim6, Ystad, Simrishamn. - Sveriges Geologiska Unders6kning Serie A la I & 2: 1-198, Stockholm. WrSTrRGARD, A. H. 1910. Index to N. P. Angelin's Palaeontologia Scandinavica, with notes. Meddelanden frfin Lunds Geologiska F~iltklubb, B 5: 1-48, Lund. WIGANt), C. 1888. Uber die Trilobiten der silurischen Geschiebe in Mecklenburg. - Zeitschrift der deutschen geologischen Gesellschaft 40: 39-101, Berlin. -

-

-

-

-

Eingang des Manuskriptes am 5.6. 1992; Annahme der revidierten Fassung durch die Schriftleitung am 7. 11. 1992.