Discussion.—Stein and Seiden (2011) provided an emended diagnosis for Artiopoda. We here present evidence that Molaria spinifera Walcott, 1912, belongs to Artiopoda, based on exopod structure, which is one of the diagnostic characters, but it does not share the diagnostic character of filiform antennulae. Characters like a flagelliform telson and an elongate caudal tergite without tergopleurae, as well as articulating ridges parallel with the anterior margin of the tergites suggest affinity of M. spinifera with Emeraldella brocki Walcott, 1912, a basal artiopod (Stein & Selden, 2011). Unless those characters can be shown to be symplesiomorphies present in basal artiopods, the antennular morphology of M. spinifera has to be considered autapomorphic for that taxon.

The type genus, Emeraldella Walcott, 1912, is the only included genus.

Type species.—Emeraldella brocki Walcott, 1912, p. 203–205.

Other included species.—Emeraldella brutoni n. sp.

Emended diagnosis.—Artiopod with head containing one antennular and three limb-bearing postantennular segments (plesiomorphy); antennulae long (about same length as trunk), containing more than 80 articles; first pair of postantennular limbs having reduced number of podomeres and no exopod. Trunk including 10 to 11 segments, and 1 elongate caudal segment with reduced tergopleurae and pair of ventral caudal flaps. Proportions of endopod podomeres strongly differentiated along body. Endopod curving outward proximally, downward at short, knee-like, fifth podomere; sixth podomere long, distinctly stenopodous. Exopod tripartite, proximal and middle parts articulating with basipod and first podomere, respectively. Lamellae on proximal part wide, with fine setules distally. Telson flagelliform, jointed, about same length as trunk, lateroventrally flanked by caudal flaps.

Discussion.—Stein and Selden (2011) emended the diagnosis of Emeraldella as given by Bruton and Whittington (1983). The inclusion of another species with different trunk segment count, as well as recognition of a flagelliform telson necessitates further emendation, given above.

Emeraldella is known with confidence from only its type locality in the Burgess Shale of British Columbia and the Wheeler Formation in the Drum Mountains, Utah, as described here. Briggs and Robison (1984) questionably assigned a specimen from the younger Marjum Formation of the House Range, Utah, to Emeraldella based mainly on the presence of 11 trunk tergites in combination with ill-defined characters of tergites and head shield. Other diagnostic characters are not discernible in that specimen. Stein and Seiden (2011) showed that Emeraldella brocki has 12 rather than 11 tergites, whereas Emeraldella brutoni n. sp. has 11 trunk tergites, diminishing the utility of trunk-segment count as a diagnostic character of Emeraldella. It further appears that the tergite count in the Marjum specimen rests on the inferred boundaries between the head shield and the first thoracic tergite, in combination with a count of questionable limb traces in the trunk. Thus, there is no clear evidence of an eleventh limb pair, and the caudal end of the specimen is ill preserved. Following Stein and Seiden (2011), the Marjum specimen is here regarded to be an indeterminate arthropod.

Diagnosis.—Large species of Emeraldella with ten trunk tergites with well-developed tergopleurae.

Description.—Body 70.8 mm long from anterior margin of head shield to base of telson. Maximum measurable length about 100 mm, with telson being incomplete. Maximum body width 43.5 mm between posterolateral tips of second trunk tergite.

Head shield semielliptical, about 14–16 mm long (sag.) and 41.4 mm wide (tr.); shield, including posterior border, not preserved axially, rendering exact length unknown. Anterior edge of head shield likely extended axially (Fig. 2.1–2.2), being unclear whether it represents part of head shield, a prehypostomal sclerite, or anterior of displaced hypostome.

Trunk, excluding caudal portion, composed of 10 tergites with well-developed tergopleurae, which become more falcate rearward. Tergite arch also increases rearward. Most tergite boundaries only discernible laterally (Fig. 1, Fig. 3), but overlap seems to be substantial, up to one-third of tergite length (sag.). Traces of articulating ridge preserved parallel to anterolateral margin on fourth tergite (Fig. 1.2, Fig. 2.4). Because the axial region of the tergites is not preserved, evidence of possible trilobation is lacking. Tergite count is based on lateral extremities of tergopleurae, which are best preserved on left side of trunk in the part (Fig. 1.1) and right side in the counterpart (Fig. 1.2). Tenth tergopleura, seen on right side of counterpart, is in immediate proximity to lateral edge of caudal portion. Posterior margin of tenth tergite is faintly visible as dark band in counterpart (Fig. 1.2). Evidence of tergopleurae in caudal portion is lacking.

Features in the caudal portion are difficult to discern (Fig. 2.3). The telson is on a layer of sediment that is flaking off proximally, exposing the caudal flaps onto which the outline of the telson is faintly impressed. A transverse dark stain crosses the center, coinciding with a fold traversing most of the caudal portion, which could mark the trunk-telson boundary, but the lateral margins of the caudal flaps appear to be continuous with the rest of the caudal portion. Immediately posterior to the transverse dark stain, a crescentic impression arches backward axially. A number of additional transverse marks, all indistinct, are posterior to that but are expressed as nicks rather than impressions (Fig. 2.3, arrows).

Caudal flaps are poorly preserved, with relation to remaining caudal portion being barely discernible. Fringing spines laterally and posteriorly are vaguely visible. Internal flap margins are visible as longitudinally arched dark stains. Diagonal lines run from anterolateral corner of caudal portion to the telson; it is unclear whether they are joints in the caudal flaps.

The part shows paired dark patches in the axial region. They are more faint than the prominent black stains on the counterpart and to some extent in the part, being best seen under crossed polarizers when the specimen is dry (Fig. 2.4, arrows). They appear to be segmental, although those of trunk segments six and seven are confluent on the left side. Posterior to trunk segment seven, they are no longer discernible. The patches may represent midgut diverticula, which are commonly preserved in fossil arthropods (Butterfield, 2002). Other, darker patches occurring axially on both part and counterpart and abaxially on the part are more difficult to associate with any particular structure (Fig. 1, only the abaxial ones drawn in Fig. 3, diagonal hatching).

The head shield is flaked off axially, revealing cuticular structures abaxially under the head and the head-trunk boundary, best seen in the part (Fig. 1.1, arrows). These are likely remnants of appendages, but the poor preservation does not allow further description. Apart from that, appendage preservation is only present in the proximal part of the left antennule, which is too poorly preserved for further description.

The cuticle carried a reticulate pattern, best seen on the right side of the head shield and first trunk tergite of the counterpart (Fig. 1.2).

Etymology.—After Professor David L. Bruton, University of Oslo, for advancing knowledge of Emeraldella.

Holotype.—KUMIP 321500, part (Fig. 1.1) and counterpart (Fig. 1.2). The tergites overlap from anterior to posterior in the part and the axial portions of at least the head shield are flaked off, exposing ventral structures beneath it. This is taken to indicate that the specimen is preserved in dorsal aspect.

Discussion.—Specimen USNM 136642 of Emeraldella brocki (Bruton & Whittington, 1983, fig. 22–23; Stein & Selden, 2011, fig. 4A) shows a median extension similar to that of KUMIP 321500. Stein and Selden (2011) tentatively interpreted the structure in E. brocki to be a prehypostomal sclerite (using the term rostral plate). Specimen USNM 136440 of E. brocki (Bruton & Whittington, 1983, fig. 11, plate 3; Stein & Selden, 2011, fig. 10A₁) also shows a distinct median extension, but the anterior of that specimen is folded, presenting the posterior of the head shield in cross section. The median projection is therefore rather dorsal and covers the width of the thoracic axis. Trilobation of the head shield may have been expressed posteriorly in E. brocki (Stein & Selden, 2011), but whether it could account for the extent seen in this specimen is questionable. A possible cause for the extent might be stabilization of the axial region against compaction through early permineralization (Bruton & Whittington, 1983; Butterfield, 2002).

Poor preservation does not allow detailed morphological assessment of the caudal portion in Emeraldella brutoni. The caudal flaps of E. brocki insert beneath the anterior part of the caudal tergite, and their structure appears to be complex (Stein & Selden, 2011). Judging from the extent of the lateral margins and diagonal lines traversing them, the caudal flaps of E. brutoni also appear to extend almost to the anterior margin of the caudal portion. Specimen USNM 136640 (Stein & Selden, 2011, fig. 10A₂,) shows features similar to the diagonal lines observed in E. brutoni, and in similar orientation. Unfortunately, preservation does not allow assessment as to whether the features represent joints in the flaps of either species. Compared with E. brocki, presence of a caudal tergite is inferred in E. brutoni, but there is little direct evidence other than the transverse fold and dark stain, which may indicate the posterior border of such a tergite. It is not clear whether the backward arching crescentic impression marks the trunk-telson joint. Stein and Seiden (2011) interpreted a crescentic backward arch in the anterior part of the telson of E. brocki to bound membranous cuticle surrounding the anus. Without preservation of a gut tract in the specimen of E. brutoni, the crescentic impression cannot be positively identified as a similar structure.

The transverse marks on the telson of KUMIP 321500, although they might be interpreted as taphonomic artifacts (e.g., creases due to physical impact, cuticular features, or irregularities), are most likely joints. A jointed, flagellum-like telson is known from Molaria spinifera Walcott, 1912 (Fig. 4.1, arrows; see also Whittington, 1981). The telson of E. brocki was described as styliform (Bruton & Whittington, 1983; Stein & Selden, 2011), but further investigation shows evidence of a jointed telson, even in that species. The telson is more or less fully preserved in only 9 of the 24 specimens (USNM 57702, 136439,136441, 144917, 144919, 144923, 144928, 144929, 250230), and two additional specimens (USNM 144918 and 250227) have the proximal part preserved. Although overlooked in previous studies (Bruton & Whittington, 1983; Stein & Seiden, 2011), the lectotype (USNM 57702) shows articulations in the telson (Fig. 4.3, arrows), which slightly bends at each articulation. USNM 136641, 144918, 144919, 250227, and 250230 show good evidence for similar articulations. The thirteenth segment identified by Bruton and Whittington (1983) is the most proximal article of the telson. Stein and Seiden (2011) identified the anus at the base of this article and considered the article to be part of the telson, but failed to recognize articulations along the entire telson. An articulated, flagelliform telson is present also in Retifacies abnormalis Hou, Chen, & Lu, 1989, and has more recently been reported from Pygmaclypeatus daziensis Zhang, Han, & Shu, 2000 (Xu, 2004). In these taxa, the telson does not articulate with a caudal tergite, but a well-developed pygidium-like shield. Lin (2009) described flexibility of the telson in Burgessia bella Walcott, 1912, which was initially considered to be flagelliform and later to be styliform (see Lin, 2009, and references therein). Lin (2009) assumed a flexible styliform telson that could be stiffened by hydrostatic pressure. Given that telson articulation can be elusive in E. brocki, the material of B. bella should be restudied.

Assignment of KUMIP 321500 to Emeraldella is based upon the presence of a caudal portion without tergopleurae and caudal flaps anteroventrally to a flagelliform telson. A further similarity is the presence of articulating ridges parallel to the anterior margin of the tergites. The specimen has only 10 trunk tergites with developed tergopleurae, rather than 11 plus the caudal tergite found in Emeraldella brocki. As the Wheeler specimen is larger than known specimens of E. brocki, it is not an early instar of that species, but represents a separate species. E. brutoni n. sp. further differs from E. brocki by having tergites that overlap by about one-third of their length (sag.) rather than by one-quarter (cf. Stein & Selden, 2011).

Several authors suggested a close relationship between Molaria spinifera and Emeraldella brocki (see Whittington, 1981, for references). M. spinifera is similar to species of Emeraldella in tergite morphology, presence of a caudal tergite without tergopleurae, and a flagelliform telson. It has only eight trunk segments, and despite the large number of known specimens, caudal flaps are not known from the species. The caudal tergite carries posterolateral marginal spines, which are not known from Emeraldella. The antennula of M. spinifera is minute, slender, and composed of only a few articles (Whittington, 1981), which is unlike the filiform antennulae of other artiopods (Stein & Selden, 2011), including E. brocki. Evidence indicates that the exopods of M. spinifera are of the distinct artiopod type, with a proximal shaftlike, lamella-bearing part and a distal lobe fringed with setae. The distal lobe was figured by Whittington (1981, e.g., fig. 34–35) and is the most commonly preserved part of the exopod. Although Whittington (1981) did not observe lamellae and considered their absence as an argument against close affinities with Emeraldella, they are well exposed in USNM 268930 (Fig. 4.2), but there is also evidence in USNM 268923. A close relationship between Emeraldella and Molaria seems likely, but a comprehensive phylogenetic analysis is needed to further corroborate this.

Occurrence.—The holotype of Emeraldella brutoni n. sp. was collected from an outcrop of brown calcareous shale about 40 meters below the top of the Wheeler Formation, as identified by Halgedahl and others (2009). The locality is about 100 meters east of a broad, dry stream gully in the center S1/2NE1/4 sec. 20, T. 15 S., R. 10 W., Drum Mountains Well 7.5′ topographic quadrangle map (United States Geological Survey, 1971). A diverse associated biota includes trilobites [Asaphiscus wheeleri Meek, 1873, Brachyaspidion sulcatum Robison, 1964, Elrathia kingii (Meek, 1870), Olenoides nevadensis Meek, 1870], other arthropods (Branchiocaris ?pretiosa Resser, 1929, Canadaspis cf. perfecta Walcott, 1912, Mollisonia symmetrica Walcott, 1912), vetulicolians (Skeemella clavula Briggs & others, 2005), worms (Selkirkia willoughbyi Conway Morris & Robison, 1986, Selkirkia sp. Conway Morris & Robison, 1986, and undetermined taxa), undetermined hyoliths, undetermined lingulide brachiopods, sponges (Choia carteri Walcott, 1920, Choia ridleyi Walcott, 1920, Diagoniella hindei Walcott, 1920, Hamptonia bowerbanki Walcott, 1920, Hamptonia parva Rigby, Church, & Anderson, 2010, Hintzespongia bilamina Rigby & Gutschick, 1976, Ratcliffespongia wheeleri Rigby & Church, 1990, Sentinelle? draco Walcott, 1920, Vauxia bellula Walcott, 1920), chancelloriids (Chancelloria sp.), and algae (Margaretia dorus Walcott, 1931, Marpolia spissa Walcott, 1919, Yuknessia simplex Walcott, 1919).

The prominent dry stream gully to the west provides limited vehicle access to the Global Standard Stratotype-section and Point of the Drumian Stage, which is 62 m above the base of the Wheeler Formation on a ridge crest about a kilometer north of the emeraldellid locality.