Upper lower Cambrian ( provisional Cambrian Series 2 ) trilobites from northwestern Gansu Province , China

Upper lower Cambrian (provisional Cambrian Series 2) trilobites are described from three sections through the Shuangyingshan Formation in the Beishan area, northwestern Gansu Province, China. The trilobite fauna is dominated by eodiscoid and corynexochid trilobites, together representing at least ten genera: Serrodiscus, Tannudiscus, Calodiscus, Pagetides, Kootenia, Edelsteinaspis, Ptarmiganoides?, Politinella, Dinesus and Subeia. Eleven species are described, of which seven are identified with previously described taxa and four described under open nomenclature. The composition of the fauna suggests biogeographic affinity with Siberian rather than Gondwanan trilobite faunas, and the Cambrian Series 2 faunas described herein and from elsewhere in northwestern China seem to be indicative of the marginal areas of the Siberian palaeocontinent. This suggests that the Middle TianshanBeishan Terrane may have been located fairly close to Siberia during middlelate Cambrian Epoch 2.


INTRODUCTION
China is a complex collage of continental blocks and accretionary belts, as well as several smaller blocks and terranes, including amalgamation of the North China plates and the southern marginal areas of Siberia.Fossils are of fundamental importance for interpretation of complex palaeogeographical situations, providing age and biogeographic constraints in formulation of tectonic models.This paper focuses on upper lower Cambrian (provisional Cambrian Series 2) trilobites from the Beishan area, northwestern Gansu Province of China.The trilobite fauna includes several palaeobiogeographically and biostratigraphically important taxa that are valuable for constraining early Cambrian plate configurations.
The base of the Xishuangyingshan Formation, overlying the Shuangyingshan Formation, bears the trilobite Galahetes opimus Zhang in Zhou et al. (1982, p. 231, pl. 59, figs 7-9) at a locality 13.5 km southwest (266°) of Section 1.The genus Galahetes Öpik, 1975a, known from Australia and Hainan Island of China, is indicative of the lower part of the traditional middle Cambrian (provisional Cambrian Stage 5).
Fossiliferous Cambrian Series 2 strata in the Beishan area were first discovered by Liu Guangxia and Chen Lijun (unpublished report from 1964).According to their report, lower Cambrian (Cambrian Series 2) strata that crop out around the Dahuoluo Well consist of greyish-white and white biosparites, occasionally with black chert interbeds.The succession has a total thickness of 21 m and has yielded trilobites of the genera Bergeroniellus, Calodiscus, Dinesus (= Erbia), Kootenia and Serrodiscus.Subsequently, Zhao Xiangshen and Ye Yu (unpublished stratigraphic report from 1966) restudied the Cambrian strata in this area and recovered some Cambrian trilobites.In an unpublished report (1967) they referred to lower Cambrian (Cambrian Series 2) strata as 'the Shuangyingshan Formation'.The Cambrian Series 2 trilobites collected by Zhao and Ye were described by Zhou Zhiqiang and Li Jinseng in the book Paleontological Atlas of Northwest China, Shaanxi-Gansu-Ningxia Volume, Part 1 published in 1982.The trilobites include Serrodiscus areolosus, Tannudiscus conicus, Dinesus spinellosus and Subeia beishanensis.Further collections, which provided the bulk of the material for the present paper, were made by Zhao Xiangshen and Zhou Zhiqiang in 1982.A trilobite list and plate including the above-mentioned four species, Kootenia sp. and Edelsteinaspis sp. were published by Zhao et al. (1984).Xiang & Zhang (1985) and Zhang (1987) described an early Cambrian trilobite assemblage from the western part of the Xinjiang Uygur Autonomous Region, China.The assemblage comes from the Huocheng Formation at Guozigou and Akqat, Huocheng County (Fig. 2).The 2-m-thick Huocheng Formation consists mainly of light grey, thick-bedded micritic limestone with shell fragments.It is disconformably underlain by black silty, siliciclastic mudstones intercalated with limestone concretions and phosphate-bearing layers (Linkuanggou Formation), and disconformably overlain by phosphatic, argillaceous siliciclastic rocks intercalated with thinbedded or lenticular limestones with trilobites of the genus Xystridura Whitehouse, 1936 (the lower part of the Kensay Formation).The trilobite assemblage from the Huocheng Formation includes six genera: Calodiscus, Dinesus (= Erbia), Edelsteinaspis, Kootenia, Pagetides and Tannudiscus (Fig. 2).These are shared with the fauna in the Shuangyingshan Formation of Beishan.

BIOGEOGRAPHICAL ASSESSMENT
Another related early Cambrian trilobite assemblage was discovered by Duan & An (2001) in a drill core recovered at Wuxingzhen Town, Yichun City, Heilongjiang Province, northeastern China (Fig. 2).The Wuxingzhen Formation is represented in the drill core and can be divided into three parts: ( 1 As mentioned above, parts of Northwest China, including at least the Middle Tianshan-Beishan Region of Zhou & Zhen (2008), are regarded as an extension of a geological region extending westwards into Kazakhstan.To the north, Repina (1972, pp. 297-299) recognized a Sayan-Altaj (also spelt Altai or Altay) province, in which limestones hold a trilobite fauna characterized by the genera Kooteniella (?= Kootenia), Edelsteinaspis, Chondragraulos and Erbia (i.e.Dinesus).Still further north, the evaporitic inner basinal part of the Siberian Platform had a 'Pseudoeteraspis-Poliella biofacies' during the early-middle Toyonian (Repina 1972;Pegel 2000).The occurrences in northern China may represent the southern margin of this greater  (Xiang & Zhang 1985;Zhang 1987).A second occurrence is in the Shuangyingshan Formation of Subei County, Beishan area, Gansu Province (Zhou et al. 1982;this paper).The third occurrence is in the Wuxingzhen Formation at Yichun City, Heilongjiang Province, northeastern China (Duan & An 2001).
Asiatic realm, recognized by Kobayashi (1972) as a North Asiatic subprovince (of an Asiatic-Mediterranean province) particularly rich in 'corynexochids' and 'ptychopariids'.However, the Middle Tianshan-Beishan Region has its own lithological and biofacies characters, although some species are shared with the Siberian province.The general aspect is certainly central Asiatic to Siberian rather than, for instance, Gondwanan, and, as noted by Zhou & Zhen (2008) and Álvaro et al. (2013), the Middle Tianshan-Beishan Terrane may have been located fairly close to the Siberian plate during late early Cambrian times (Cambrian Epoch 2).

SYSTEMATIC PALAEONTOLOGY
Because of the fragmentary state of the material, the genus Kootenia with at least five species is not treated herein.The morphological terminology used here follows that of Whittington in Kaesler (1997).Figured and cited specimens are deposited in the Nanjing Institute of Geology and Palaeontology, Chinese Academy of Sciences, Nanjing (NIGP), and Xi'an Institute of Geology and Mineral Resources, Xi'an (XIGM).
Other material.Three cephala (NIGP152225, 152226, 152229) and three pygidia [XIGM Tr017 (Fig. 3C, D;Zhou et al. 1982, pl. 57, fig. 19) The pygidium is subtriangular in outline, rounded posteriorly and 85-91% as long as it is wide anteriorly.The axis is strongly convex (tr.), tapered backwardly at a uniform rate and terminates slightly in front of the border furrow.It consists of 10 rings and a terminal piece, each with a short but prominent median spine or spine-like tubercle directed upwards and slightly rearwards.These spines or turbercles decrease in size and length posteriorly.The ring furrows are well incised.The axial furrows are distinctly incised and rather broad.The pleural field is unfurrowed, sloping down strongly to the border furrow, slightly narrower than the axis and confluent behind the axis in a narrow strip.The border furrow is deeply incised.The border is narrow (tr.), slightly narrowing posteriorly and carries ten pairs of tiny border spines, directed posteroventrally.
The surface of the cephalon is strikingly ornamented with a fine, reticulate sculpture of narrow, low, raised ridges.Well-preserved specimens show that the external exoskeletal surface of the pygidium is covered with the same sculpture as the cephalon.
Remarks.New material permits a detailed supplement of the original description of this species (Zhou in Zhou et al. 1982, p. 220).It is easily distinguished from other Serrodiscus species by its large, nearly erected posterior glabellar spine and by its pygidial axial rings with prominent median spines or spine-like tubercles.Distinct median axial tubercles have also been described in pygidia of S. bellimarginatus (see, e.g., Rushton 1966, pl. 1, fig. 3a, b).
Diagnosis.A species of Tannudiscus with a subconical frontal glabellar lobe, which reaches the anterior border furrow and is 52-61% the length of the posterior glabellar lobe.Cephalic border rapidly widened (sag.and exsag.)anteriorly.Pygidial axis tapered, much wider than pleural field, nearly reaching border furrow, with five weak ring furrows.
Description.The cephalon is subquadrate in outline, 85-110% as long as it is wide and broadly rounded anteriorly.The glabella is large, 74-80% as long as the cephalon, bilobed and reaches the border furrow anteriorly.The anterior lobe is subconical in outline, rounded anteriorly, 52-61% as long as the posterior lobe, strongly convex (tr.), anteriorly sloping down to the border furrow.The transglabellar furrow is narrow, deeply incised and curved slightly forward medially.
The posterior lobe is subquadrate in outline, slightly widened rearwards, rounded posteriorly, rising steeply from cheeks and highest at the posterior 1/3 where it slopes steeply rearwards.An occipital furrow is absent.
The dorsal and preglabellar furrows are narrow and deep, anteriorly joining the border furrow.The fixed cheek is moderately convex (tr.), sloping down to the border furrow.The border furrow is deep and anteriorly nearly transverse.The border is gently convex, wide, 18% of the cephalic length (sag.) and strongly narrowing rearwards.
The pygidium is semi-elliptical in outline, nearly as long as wide, with the anterior margin forwardly arched.The axis is strongly convex (tr.), tapered rearwards, nearly reaching the border furrow, and 79-84% of the pygidial length and 49-52% of the pygidial width anteriorly; its posterior part rapidly slopes down to the postaxial region.In addition to a narrow (sag.)articulating half ring, the axis is divided into five rings and a terminal piece.The ring furrows are shallow abaxially and almost invisible adaxially.The axial furrows are distinct and narrow.The pleural fields are strongly convex, steeply sloping down to the border furrow and rapidly narrowing backwards.The border furrow is deeply incised.The border is nearly flat, 10-11% of the pygidial length (sag.) and anterolaterally rapidly narrowing forwards.
All parts of the exoskeleton lack an apparent sculpture.
Remarks.The species that is most similar to Tannudiscus conicus appears to be Tannudiscus tannuolaicus Pokrovskaya (1959, p. 178, pl. 11, figs 16, 19, 20; see also Rushton 1966, text-fig. 8f).Features that are shared by both species are the glabella reaching the border furrow, a subconical anterior glabellar lobe and a cephalic border that widens anteriorly.Significant characters of T. conicus, which may separate it from T. tannuolaicus, include the shorter anterior lobe, being about 60% the length of the posterior lobe, a wider pygidial axis that does not reach the border furrow and the axial ring furrows of the pygidium that are visible abaxially.The species is like T. altus (Repina et al. 1964, p. 263, pl. 39, figs 1-3;Repina & Romanenko 1978, p. 111, pl. 2, figs 6-15) in having a subconical anterior glabellar lobe, a long and strongly convex posterior glabellar lobe and a pygidial axis not reaching the border furrow.It is distinguished by its glabella that reaches the anterior border furrow, its longer (sag.)cephalic border and its wider pygidial axis with shallow axial ring furrows laterally.Tannudiscus balanus Rushton, 1966 from England and SE Newfoundland is distinctive in having an expanded anterior glabellar lobe that is wider than the posterior lobe.
A Tannudiscus sp. was described from northern Tianshan by Xiang & Zhang (1985).Unfortunately their single illustration does not allow any comparison.Tannudiscus conicus is also known from Mt. Bolhinur in Xinjiang.As far as we can judge, it is indistinguishable from our material.
Family CALODISCIDAE Kobayashi, 1943 Genus Description.The cephalon is semi-elliptical in outline and 82-87% as long as it is wide.The preoccipital glabella is subcylindrical in outline, slightly constricted medially, rounded anteriorly, 59-66% of the length and 26-29% of the width of the cephalon, convex (tr.) and rather strongly sloping down to the border furrow anteriorly.Two pairs of transglabellar furrows are very weak on the external exoskeletal surface, but clearly expressed on internal moulds.The anterior glabellar lobe is nearly as long as the total length of the posterior two glabellar lobes combined.The dorsal, preglabellar and occipital furrows are deeply incised.The occipital ring occupies 17.5% of the cephalic length (sag.) and is strongly narrowing abaxially with a strongly backwardly arched posterior margin, lacking a spine or node.The cheeks are nearly as wide as the preoccipital glabella posteriorly, convex (tr.) and abaxially downsloping to the border furrow.The border furrow is deep and broad; an ill-defined depression occupies the space between the glabella and the anterior border.The anterior border is narrow (sag.) and convex, slightly narrowing rearwards.
The posterior border furrow is distinctly defined, and widened and anterolaterally curved outwards.The posterior border is convex and slightly widened laterally.
The pygidium is semi-elliptical and about 80% as long as it is wide.The axis is strongly convex (tr.), rising steeply from the pleural fields, gently tapered rearwards, rounded posteriorly, almost reaching the border furrow, 88% of the pygidial length and 42% of the pygidial width anteriorly; it is composed of four distinct rings and a very short terminal piece.The anterior three ring furrows are distinctly incised, while the rearmost is weak.The axial furrows are deep.The pleural fields are convex (tr.) abaxially, strongly sloping down to the border furrow and have three pairs of distinct pleural furrows excluding the deep anterior border furrow.The interpleural furrows are very faint.The anterior pleural band has a fulcrum at about 1/2 the distance from the axial furrow to the lateral margin.The border furrow is distinct and of moderate depth.The border is low, gently convex, about 1/10 of the pygidial length, slightly serrated and provided with four or five pairs of very short, blunt spines.
The surface of the test is indistinctly granulated.
Remarks.The relation between Pagetides and Neopagetina has been discussed for a long time (see Blaker & Peel 1997 for a review; see also Fletcher & Rushton 2007).According to the original definitions, the pygidial pleura should be smooth in the former genus but furrowed in the latter.Blaker & Peel (1997, pp. 34-37) summarized the discussion and illustrated specimens from North Greenland supposed to represent the type species of both genera.They concluded that there are in fact pleural furrows also in Pagetides and that there is no character that can be used for differentiating the two genera.However, the lectotype cranidium of the type species for Neopagetina, N. rjonsnitzkii (Lermontova, 1940), differs particularly in lacking the stout cephalic spine that is characteristic of Pagetides (Jell in Kaesler 1997, fig. 244:2a).We follow, however, Blaker & Peel (1997)  Remarks.As with other species in this fauna, the material is meagre and species identification far from obvious.Pagetides huochengensis was described from three poorly preserved cranidia recovered from the Huocheng Formation in the Xinjiang Uygur Autonomous Region.Their overall morphology is similar to the cranidia illustrated here, but the poor quality of the material and illustration and the lack of illustrations of pygidia make the identification of our material uncertain.It also makes our associations of cranidia and pygidia uncertain.A possible Pagetides species described from the Xinjiang Uygur Autonomous Region by Xiang & Zhang (1985) has a wider and smoother cephalic anterior border and does not represent any of the species from the Beishan area.
The general appearance contrasts with most Pagetides species, which as a rule have smooth or weakly furrowed pleural fields and lack median axial nodes or spines.It is difficult to find any similar species even in faunas that are comparatively rich in species of Pagetides.The overall morphology of our species is similar, however, to Discomesites fragum Öpik, 1975b, from the Cymbric Vale Formation of western New South Wales, Australia, particularly with respect to the pygidium.Discomesites fragum is the type species for the subgenus Pagetides (Discomesites) Öpik, 1975b, a subgenus that may be taken to include also our species.
A cranidium associated with a pygidium here referred to as Pagetides?sp. 1 is incomplete, lacking its posterior part and much of the glabella (Fig. 6N).The ornament is granular and the anterior border is extended backwards medially.It is here questionably assigned to Pagetides (D.) huochengensis.Pagetides?sp. 1 Figure 6M Material.One pygidium (NIGP152252) from level BT4, Section 2.
Remarks.The pygidium has a slightly tapering axis with four axial rings separated by narrow inter-ring furrows.The axial rings lack a medial elevation.Four pairs of distinct pleural furrows are present in addition to a pair of nearly effaced furrows behind the axis.Most species of Pagetides have unfurrowed or weakly furrowed pleural fields and a higher number of segments in the pygidium than in the pygidium at hand.

'CORYNEXOCHIDS'
The key character of the 'Corynexochida' is fusion between the hypostome and the rostral plate, but the condition is often unknown and occasionally known to  Palmer & Halley (1979, pp.77-78) described species of Ptarmigania and Poliella, two genera that are almost indistinguishable from each other although they are considered as representatives of the Zacanthoididae Swinnerton, 1915 and the Dolichometopidae Walcott, 1916, respectively.There is apparently a gradation between the two species associations, although in general one is more spiny than the other.There is thus good reason to regard Dolichometopidae as a junior synonym of Zacanthoididae or, at the most, as a subfamily of the latter.
The Edelsteinaspididae Hupé, 1953a similarly may be considered as a junior synonym of Dorypygidae Kobayashi, 1935.Edelsteinaspis differs from 'typical' dorypygids essentially only in having a more elevated frontal glabellar lobe, more deeply impressed glabellar furrows (which appear strongly arched because of the tilt imposed by the anteriorly elevated glabella) and wider pleural fields, features generally not useful in classification above species or perhaps genus level.However, we refrain from introducing any classificatory modifications herein.
As mentioned above, the fauna includes at least about five species of Kootenia (or Kooteniella) that are not dealt with here.
Family DORYPYGIDAE Kobayashi, 1935 As noted by, for instance, Blaker & Peel (1997), it is difficult to distinguish between the genera Kootenia and Ogygopsis.We follow here a practical approach in considering species with a fairly short pygidium as members of Kootenia.Hundreds of species of these two genera have been described, often based on disarticulated sclerites and insufficient characters (cf.Robison & Babcock 2011).For instance, information on the profile of the cranidium in lateral view is often missing.Pygidia are necessary for reliable identification (e.g.Geyer 1994Geyer , p. 1307)).The material from Beishan includes three forms with comparatively long pygidia (with five segments plus a terminal piece) and three forms with short pygidia (with three segments plus a terminal piece).Species identification is in several cases hampered by deformation and poor preservation of the specimens.It seems likely that the material may comprise even more species, but additional findings are needed to confirm this suspicion.
Description.The glabella is nearly parallel-sided, with an evenly rounded front and three pairs of chevronshaped glabellar furrows that are connected medially.The most anterior glabellar furrows have a characteristic course, almost V-shaped, giving the posterior part of the frontal lobe a strikingly triangular, almost pointed appearance.The angle between the eye ridge and the axial furrow is comparatively high and the palpebral lobe is strongly curved, which results in a wide, subtriangular palpebral area.The entire exoskeletal surface is covered with granules of different sizes.The occipital furrow is nearly transverse and the occipital ring has a faint median node.No free cheeks are known.The surface of the pygidium is covered with small tubercles, somewhat larger on the axial rings than on the pleural field, and with variation in tubercle size.The pleural and interpleural furrows in this species are different.The former are wide and shallow, the latter narrow but distinct.They fade some distance from the lateral margin.A border furrow is absent.The axis has four distinct segmental rings plus the terminal piece.The anteriormost ring has a median node.
Remarks.Of the two known species of Edelsteinaspis with four axial rings in the pygidium, the pygidial border is comparatively narrow in E. gracilis Lermontova, 1940, whereas it is distinctly wider in E. plana.This condition has effects on the position of the last axial ring, which is distinctly anterior to the mid-length of the pygidium in E. plana, but midways in E. gracilis.In these respects, our specimens correspond to E. plana.Co-occurring cranidia from BT5 are of two types.One has a gently curved palpebral lobe placed comparatively close to the glabella.The other has a more strongly bent palpebral lobe placed further away from the glabella.The latter configuration is identical to that in a specimen of E. plana illustrated by Suvorova (1964, pl. 6, fig. 6).
Description.A single cranidium was found in the section and level together with a pygidium referred to this species.They may belong to the same species.The surface sculpture is evenly granular.Unlike the cranidia referred to E. plana, the glabella expands slightly towards the front and is widest across the frontal lobe.The three pairs of glabellar furrows are simple, straight and regularly spaced, and the lateral lobes are of similar shape throughout.The frontal lobe is more subcircular in outline than in the cranidia assigned to E. plana.The eye ridge is straight and the palpebral lobe fairly strongly but evenly curved.The fixigena is fairly flat, but has a weak convexity inside the palpebral lobe.The free cheek is not known.The thorax is not preserved, but the posterior morphology of the cephalon indicates that the hinge is probably quite short.
The pygidium has a fairly broad axis with six short segmental rings and a terminal piece without signs of segmentation.The pleural field has a broad, notably inclined border field which probably corresponds to the doublure.Pleural and interpleural furrows are almost equally developed and notably closely spaced, corres-ponding to the shortness of the axial rings.They almost extend to the margin, which lacks spines.Much of the pleural surface is smooth in detail, but the border has a pattern of faint wavy lines.
Remarks.The known species with six axial rings in the pygidium fall in two groups with regard to the pygidial morphology.Edelsteinaspis ornata Lermontova, 1940 andE. compta Suvorova, 1964 have a comparatively narrow pygidial border without apparent flattening.E. biramjensis Suvorova, 1964 andE. altaica Romanenko (in Repina &Romanenko 1978) have a distinctly wider pygidial border, which is apparently flattened.The drawing of Suvorova (1964, fig. 17b) is greatly misleading, showing a narrow border and the terminal piece of the axis too close to the posterior margin.Our sixsegmented pygidium has a narrow border, as have the two former species.However, it looks more similar to the two latter species in having both interpleural and pleural furrows extending almost to the lateral margin, and possibly in having a greater topography.The glabella is distinctive in its forward expansion and the shape of the frontal lobe.C, dorsal view of pygidium, NIGP152286, × 0.9.D, F-M, Edelsteinaspis plana Chernysheva, 1961 from level BT5.D, H, dorsal and posterodorsal view of cranidium, NIGP152288, × 1.5; F, cranidium, NIGP152292, × 1.9; G, cranidium, NIGP152290, × 3.4; I, J, dorsal and left lateral view of pygidium, NIGP152289, × 2; K, cranidium, NIGP152293, × 2; L, cranidium, NIGP152291, × 4.9; M, pygidium, NIGP152287, × 4.4.
The pygidium most closely resembles those of the two latter species (E.biramjensis and E. altaica), but may also belong to still another, yet undescribed species.It is clear that the cranidium does not belong to any of the species mentioned above.
Family ZACANTHOIDIDAE Swinnerton, 1915 Genus Ptarmiganoides Rasetti, 1951 Type species.Ptarmiganoides bowensis Rasetti, 1951, from the Naomi Peak Limestone Member of the Langston Formation of the Canadian Rocky Mountains; by original designation.
Description.The glabella is long and almost parallelsided, expanding only slightly and gradually anteriorly, and broadly rounded in front.Its lateral profile is nearly evenly curved (Fig. 8B).In front of it there is a frontal furrow and a strongly raised marginal ridge.The most posterior glabellar furrow (S1) is forked, with the main branch extending inwards and strongly backwards adaxially.The occipital furrow is deeply incised and curved forwards abaxially.The occipital ring has a prominent axial spine extending backwards from the posterior edge.The fixigena has an almond-shaped rise from which the surface descends both towards the axial and the palpebral furrows.The palpebral lobe is long, arcuate and narrow, defined by a prominent palpebral furrow.The 2nd and 3rd glabellar furrows from behind are almost transverse and of similar length, whereas the 4th is shorter and directed inwards and slightly forwards adaxially.The entire surface is covered with tiny pits.
Remarks.Palmer & Halley (1979, pp. 94-95) stated that the genera Paralbertella, Ptarmigania and Ptarmiganoides, and even some other genera, differ from each other only in the pygidium.Similarly, Rasetti (1951, p. 179) noted that 'It is a common occurrence among dolichometopid trilobites to find almost identical cranidia associated with a great variety of pygidia ...'.In the absence of pygidia, our generic determination can therefore only be tentative.The combination of a strongly inclined S1 and a narrow preglabellar furrow separating the glabella from a 'border', which is merely a low, thin ridge, is shared with Ptarmiganoides crassaxis Palmer & Halley (1979, pl. 11, figs 12, 18).The two species also have a very similar occipital spine and similar fixigenae.This suggests that our species may belong to Ptarmiganoides.
Information about the pygidium is, however, needed before a confident generic assignment can be made.So far, species of this genus have been restricted to the traditional middle Cambrian (Albertella Zone; provisional Cambrian Stage 5) of North America (Resser 1939;Rasetti 1951;Palmer & Halley 1979, including a thorough revision of Resser's determinations).In terms of the Siberian chronostratigraphy, the Albertella Zone can be correlated with the lower Amgan Stage (Blaker & Peel 1997;Geyer & Peel 2011).Species assigned to Zacanthoides and Zacanthopsis differ in having a considerably longer (sag.)preglabellar area (e.g.Sundberg 1994 for Zacanthoides; Blaker & Peel 1997).Wenkchemnia (Dolichometopidae) and Parkaspis (Zacanthoididae) have a more expanding glabella (Sundberg 1994).
Prozacanthoides sp.undet. of Rasetti (1948, pl. 2, figs 20-24) differs distinctly from our material: the glabella expands steadily towards the anterior, and the preglabellar furrow is wide, with the anterior slope of it extended, although less than in Zacanthoides.The pygidium of Rasetti's species is somewhat Kootenia-like, with marginal spines.Shaw (1955) illustrated a cranidium of Prozacanthoides sp.indet.As in Rasetti's specimens, the glabella is expanding, and there is a distinct preglabellar area.
Our species is also reminiscent of species of the genus Poliella Walcott, 1916, placed by Palmer & Halley (1979) in the family Dolichometopidae.They noted that the main difference between this family and Zacanthoididae is that there are intergenal and pygidial spines only in the latter.The intergenal corner is not visible in our material.
In the absence of a pygidium, we abstain from giving 'our' species a formal name, and the generic attribution is provisional.
Discussion.Jell & Adrain (2003, p. 428) considered the subgenus to be enough separated from Poliellina (Poliellina) for it to warrant genus status.
Politinella sp.Where the exoskeleton is preserved (patches in the glabella and the occipital ring with spine) the surface has fine granules.The pygidial axis is parallel-sided, wide and of constant width to the terminal piece.There are probably four axial rings in addition to the terminal piece, which is highly elevated above the posterior margin.The pleural field slopes strongly from the axis.The pleural and interpleural furrows are deep.Each pleural segment extends into a pointed spine, whereas the posterior end of the pygidium is rounded.
Remarks.The pygidium associated with the cranidium is unusually small for the family, but comparable in size and gross morphology to that in Poliellina and Parapoliella.The species closely resembles Politinella bajanica Ivshin (1979, p. 52, pl. 1, figs 1-12), from which it differs chiefly in the proportionately wider glabella, transglabellar S2 and S3, shallower S4 and a stout occipital spine.Politinella sp. 1 may represent a new species, but open nomenclature is preferred here pending new discoveries.
Description.The cranidium is subtrapezoidal in outline, with length smaller than width at mid-length.The glabella is parallel-sided, evenly rounded anteriorly.
Transversely, the glabella is highly convex, sagittally it is strongly curved.There are three pairs of glabellar furrows.The anterior pair is very weak and transverse, and the middle pair is short, shallow and transverse.The posterior pair is deeply incised and extends inwards obliquely to the occipital furrow outlining raised subtriangular basal glabellar lobes.The occipital furrow is deeply incised and transverse.The occipital ring is convex (tr.), rapidly widened medially.Posteriorly it has a small axial spine directed rearwards and slightly upwards.The axial furrows and additional furrows, which extend anterolaterally from the axial furrow near the anterior end of the glabella to the anterior border furrow, are deep and wide and join the deep preglabellar furrow.The preglabellar field is narrow or almost absent in the sagittal line, rapidly expanding laterally to form a pair of triangular raised lobes.The preocular field of the fixigena is wide (exsag.)and slopes down anteriorly and laterally.The anterior border is narrow (sag., exsag.),convex (sag., exsag.)and of nearly constant width, well defined by a deep and nearly transverse anterior border furrow.The eye ridge is distinct, slightly oblique rearwards, well defined by a distinct furrow behind the eye ridge on the larger specimens, but is faint on the smaller ones.The palpebral lobe is convex, short and situated anterior to the mid-length of the cranidium.The present material reveals that the palpebral lobe becomes relatively smaller during ontogeny.The anterior branches of the facial suture diverge slightly forwards and the posterior branches diverge strongly rearwards from the palpebral lobe.The palpebral area of the fixigena is convex (tr.), 61-65% of the glabellar width at mid-length.The postocular fixigena is broad and slopes strongly down laterally.The deep posterior border furrow widens and curves gently forwards laterally.The posterior border is convex (exsag.),widens outwards and has a distinct angularity directly behind the palpebral lobe, and is directed slightly forwards lateral to this angularity.The librigena is broad and evenly convex (tr.).The lateral border, which is well defined by a deep and wide border furrow, is wide and gently convex and extends into a short genal spine directed backwards and outwards.
The pygidium as represented by NIGP152309 (Fig. 9J, K) is small, subsemicircular, 62% as long as wide and consists of a single ring only.The articulating half-ring is well developed and separated from the axial ring by a wide furrow.The axial ring is semielliptical, 62% as long as wide and much higher than the pleural field, with the moderately convex (tr.) inner part and steeply sloping peripheral part.Axial and postaxial furrows are very shallow.The pleural field is evenly convex, about 30% as wide as the axis, slopes down to the pygidial margin, without pleural, interpleural furrows and border.The postaxial field is narrow (sag.) and slopes down rearwards.
The pygidium NIGP152310 (Fig. 9L, M) differs from the one described above in having a larger pleural field, a larger articulating half-ring, a longer axis that is sub-conical, deeper axial furrows and more irregular tuberculation.However, there is certain similarity between this pygidium and the one attributed to Dinesus spinellosus in the lack of segmental axial rings, in the small convex pleura and in the ornament of tubercles of varying size.It may belong to a different species.On the other hand, species of Dinesus are fairly variable in their morphological characters.We therefore refer it to D. spinellosus, albeit with some uncertainty.
The surface of the cranidium and pygidium is densely ornamented with small-to medium-sized tubercles.
Remarks.Whitehouse (1936Whitehouse ( , p. 73, see also 1939, pp. 228, 231) , pp. 228, 231) suggested that Dinesus was present in the Cambrian of Siberia, being represented by Solenoparia?sibirica (Schmidt, 1886).Palmer (1968) compared the generotypes Dinesus ida and Erbia sibirica and concluded that they are generically different, the difference being the presence of a pair of isolated triangular elevated fields in front of the glabella only in Dinesus.All Siberian species referred to Erbia except for E. sibirica would probably belong to Dinesus.However, the difference is caused by a slight change in the forward extension of the glabella and we do not regard this as a character that can be used alone to distinguish a genus.We therefore regard Erbia Lermontova, 1940 as a junior synonym of Dinesus Etheridge, 1896(cf. Gozalo et al. 2011).Erbiella pjankovskia Fedyanina, 1962 is indistinguishable on a generic level from Dinesus, and Erbiella is therefore a likely junior synonym of that genus.
Although differing in some respects, the species from Beishan has an overall resemblance to Dinesus ida Etheridge, 1896(Whitehouse 1939, p. 229, pl. 24, figs 1-3), the type species of Dinesus.The outline of the glabella, shape of the anterior border, preglabellar field of the cranidium and librigena and the lack of a segmental pygidium are much similar in the two species.Features of D. spinellosus which distinguish it from D. ida are the shorter palpebral lobes, the presence of an occipital spine and the tubercular external surface on the exoskeleton.
Dinesus granulosus (Lermontova) is similar to our species but has a narrower fixed cheek and indistinct eye ridges (e.g.Lermontova 1951, pl. 16, fig. 3;Chernysheva 1961, pl. 30, figs 7-12).A few species have been described from the Russian part of the Altaj Mountains but the preservation of the specimens and the quality of the illustrations are not good enough for us to make a comparison.A specimen from Altaj, identified as Erbia granulosa Lermontova, is hardly determinable (Repina & Romanenko 1978, pl. 32, fig. 5).

Subeia beishanensis Li in
Description.The glabella decreases slightly in width towards the anterior and is somewhat acutely rounded in front.In one specimen there is a faint parafrontal band.Occasionally, three regular pairs of glabellar furrows are faintly indicated.The occipital furrow is somewhat shallower in the middle.The occipital ring is fairly long (sag.) in the middle, where it ends in a tubercle.The axial furrows are clearly defined.The ocular ridge is confluent with the palpebral lobe and curved gradually forwards to reach the glabella at the anterior glabellar furrow (S3).It is fairly wide, but may be narrower than the palpebral lobe.The latter is fairly long and extends backwards to a transverse line passing the posterior glabellar lobe (L1).
The cephalic border is comparatively wide, evenly curved and separated from the remainder of the cranidium by a narrow distinct furrow.The preglabellar field tends to be slightly narrower than the border.The free cheek extends into a genal spine.
Discussion.This species is so far not known from outside the Beishan area.The parafrontal band has been taken as a characteristic of the subfamily Myopsoleninae.The value of the character is unknown, but Subeia beishanensis is closely comparable to the type species of Myopsolenus, M. magnus (Hupé 1953b, pl. 10, figs 11, 14, 15, and text-fig. 48:10).Subeia beishanensis differs in having less impressed glabellar furrows and a poorly defined parafrontal band.It is possible that Subeia is a junior synonym of Myopsolenus, but the lack of information on the thorax and pygidium hampers an analysis of their relationships.Pocock (1964) noted similarities between Estaingia and genera such as Bergeroniellus, Bergeroniaspis, Xystridura, and the 'true' protolenids Protolenus and Lusatiops.He distinguished the protolenids on their tapering glabella, but this seems not to be a useful character.In the thorax, the fulcral knobs and deep pleural furrows are a similarity with ellipsocephalids.The deep furrows and marginal spines of the pygidium may just indicate a low degree of tagmosal difference between the thorax and the pygidium.Subeia may be regarded as a junior synonym of Pararaia Kobayashi, 1942, but the morphology of the thorax and the pygidium is not known in the latter genus.
A phyletic analysis by Paterson & Edgecombe (2006) has placed Estaingia close to Ellipsocephalus, and Paterson & Brock (2007) place the Estaingiidae as a family of the Ellipsocephaloidea.
Elicicola Jell, 1990(see Jell 1990, fig. 193) is another ellipsocephaloid genus with similarities to S. beishanensis.It has a stronger relief of the cephalon and large muscle attachment areas rather than glabellar furrows.

Fig. 1 .
Fig. 1.A, B, sketch maps showing the location of the Beishan area, northwestern Gansu Province, China.C, distribution of major outcrop areas of Cambrian rocks and the location of measured sections (1-3) in the Beishan area.D, lithological successions and observed biostratigraphic distribution of trilobites in sections 1-3.
) a lower part consisting of dark grey crystalline limestones and dolomitic limestones with an unknown thickness, (2) a middle part composed of a 100-150-m-thick succession with grey metamorphosed limestones (marble) intercalated occasionally with carbonaceous slates and silty slates and (3) an upper part comprising a 200-m-thick succession of dark grey carbonaceous and silty shales intercalated occasionally with micritic and bioclastic limestones.The bioclastic limestones have yielded trilobites of the genera Inouyina, Jangudaspis, Kootenia, Laminurus, Neocobboldia, Onchocephalina, Pagetia, Proerbia and Pseudozacanthopsis (Duan & An 2001; Fig. 2).The trilobite assemblage was considered by Duan & An to represent a fauna of late early Cambrian age (Lenian or Toyonian; provisional Cambrian Age 4) with Siberian affinity.Its composition differs, however, markedly from that of the fauna from the Shuangyingshan Formation.This distinction may result from a difference in both facies and age.

Fig. 2 .
Fig. 2. Map of China showing the three so far known occurrences of Cambrian Series 2 faunas with north Asiatic affinities, all in northern China.One occurrence is in the Huocheng Formation at Guozigou and Akqat in Huocheng County, western Xinjiang Uygur Autonomous Region(Xiang & Zhang 1985;Zhang 1987).A second occurrence is in the Shuangyingshan Formation of Subei County, Beishan area, Gansu Province(Zhou et al. 1982; this paper).The third occurrence is in the Wuxingzhen Formation at Yichun City, Heilongjiang Province, northeastern China(Duan & An 2001).
be different (for instance in Ogygopsis and dinesids; see alsoHopkins & Webster 2009 andRobison &Babcock 2011).Other structural differences also indicate that the 'corynexochids' is a polyphyletic taxon (cf.Robison & Babcock 2011).Dorypygids and edelsteinaspidids share an inflated glabella with parallel glabellar furrows (when present), whereas in zacanthoidids and dolichometopids the glabellar furrows diverge from each other in their extension from the axial furrow.Dinesids, formerly assigned to the Corynexochida in Western literature(Moore 1959, pp.O230-231), do not resemble other 'corynexochids' and do not have a fused rostralhypostomal plate.In Russian literature they are typically treated as solenopleuroids (e.g.Chernysheva 1961;Repina & Romanenko 1978), which seems more reasonable.The oryctocephalids are strikingly different from all others in their glabellar morphology with glabellar furrows typically consisting of pairs of pits not reaching the axial furrows.
Calodiscus Howell, 1935 (Zhang & Clarkson 2012)ce of pleural furrows in Pagetides and regard Neopagetina as a junior subjective synonym of Pagetides.As here defined, Pagetides is a geographically widely distributed genus that is known from several Cambrian palaeocontinents.It was recently described also from South China(Zhang & Clarkson 2012).Description.The anterior border is evenly rounded and separated from the cheek rolls by a fairly narrow border furrow.It is slightly extended backwards medially.The elevated cheeks are separated by a median furrow in front of the glabella.The glabella is nearly parallelsided, with a cone-shaped anterior end and a long, stout spine that is directed backwards and upwards.The glabellar furrows are distinct, but visible only laterally.The pygidium has a slightly tapering axis with six axial rings separated by narrow and deeply incised inter-ring furrows.At least the five anterior axial rings have a large, rounded median elevation.Since this elevation is truncated in all specimens, it is not clear if it represents a large node or the base of a stout spine.Five prominent pleural furrows are present.Interpleural furrows are obsolete.The external exoskeletal surface is coarsely granular.