Type species.—Kenocharixa kashimaensis sp. nov., by original designation.

Additional species.—We here transfer to Kenocharixa the species Charixa goshouraensis Dick, Komatsu, Takashima and Ostrovsky, 2013 and Conopeum stamenocelloides Gordon and Taylor, 2015.

Etymology.—The name combines Charixa, referring to similarity to that genus, with “keno”. referring to the characteristic kenozooidal extensions filling interzooidal grooves.

Diagnosis.—Colony encrusting, unilaminar, sheet-like or erect, bilaminar. Gymnocyst moderate to negligible; long cauda lacking. Opesia occupying most of frontal surface, oval or elliptical in outline. Cryptocyst sloping, granulated. Zooids with closure plates scattered throughout colony, often grouped; variable in frequency. Spines lacking. Medium-sized kenozooids (here termed ‘interzooidal polymorphs’ to distinguish them from the smaller proximolateral kenozooids) present or absent; scattered throughout colony. Small kenozooids arise from proximolateral corners of zooids on one or both sides, sending out long extensions that fill interzooidal grooves. Zooids have distal and distolateral buttressed recesses leading to multiporous septula. Ancestrula initially buds three zooids, distally and distolaterally; ancestrula and several generations of daughter zooids can have closure plates. Avicularia, ovicells lacking.

Remarks.—In the type species, most zooids bear paired, initially triangular kenozooids covering the proximal gymnocyst, a feature that along with the absence of avicularia and ovicells suggested placement in Conopeum Gray, 1848, the type species of which, Millepora reticulum Linnaeus, 1767, has similar kenozooids. Dick et al. (2013) discussed at length the taxonomic problems associated with Conopeum, noting that few Recent or fossil species currently placed in that genus show the same suite of characters as the type species. In particular, it is controversial whether paired proximal kenozooids are diagnostic for the genus. Dick et al. (2013) described from southern Japan another malacostegan species, Charixa goshouraensis, of late Albian–early Cenomanian age. The genus Charixa Lang, 1915 provided a reasonable but not entirely satisfactory fit for that species, which has a multiserial colony with contiguous zooids and commonly has zooids with closure plates. In Charixa, colonies tend to be pluriserial with irregularly arranged, partly contiguous zooids, and closure plates are generally uncommon. With regard to generic placement, we realized that both Charixa goshouraensis and the new species from Shimokoshikijima resembled species in the Cretaceous genera Charixa and Spinicharixa Taylor, 1986a, but shared a unique character not present in the latter two genera—proximolaterally derived kenozooids that ramify into the surrounding interzooidal grooves, eventually connecting with rami from other kenozooids and filling the grooves. We thus erect the new genus Kenocharixa for these two species from the Cretaceous of southern Japan.

We also transfer to Kenocharixa the species Conopeum stamenocelloides Gordon and Taylor, 2015, described from the Eocene of New Zealand; this species resembles Kenocharixa kashimaensis in having paired, triangular proximolateral kenozooids that extend into and fill the interzooidal grooves, and autozooids with closure plates. The cryptocyst in C. stamenocelloides is much wider than in K. kashimaensis, especially proximally, whereas that in K. goshouraensis is intermediate between the two, and generally wider proximally than laterally. Kenocharixa goshouraensis lacks the obligate paired, triangular kenozooids seen in the other two species, but they are often paired and triangular (e.g. Dick et al., 2013: figs. 7B, D, E; 8B). Conopeum stamenocelloides differs from K. kashimaensis and K. goshouraensis in forming an erect, bilaminar colony and in having interzooidal connections in the form of small mural septula (Gordon and Taylor, 2015). The difference in colony form is not an impediment to placement in the same genus, as various cheilostome genera contain both encrusting unilaminar and erect species (e.g. Thalamoporella, Porella). The nature of the interzooidal connections need also not be an impediment; the septula in C. stamenocelloides appear to be set in mural depressions (Gordon and Taylor, 2015: fig. 21A, B) that may be the vestiges of the buttressed recesses seen in K. kashimaensis and K. goshouraensis.

Characters in Kenocharixa patchily overlap with those in Spinicharixa and Charixa. The primary difference between the latter two genera is that zooids in Spinicharixa have numerous spine bases around the margin of the cryptocyst, whereas those in Charixa lack spine bases or have only a small distal pair (Taylor, 1986a). The three species we place in Kenocharixa appear to lack spine bases altogether, a feature shared with some species in Charixa but distinguishing it from Spinicharixa. Taylor (1986a) also considered Spinicharixa and Charixa to differ in the degree of colony coherence, with species in Spinicharixa having pluriserial or multiserial colonies with regularly arranged zooids, and those in Charixa having pluriserial colonies with irregularly arranged zooids. This difference is not as clear as with the presence or absence of marginal spines; of the two species currently placed in Spinicharixa, S. pitti Taylor, 1986a has a multiserial colony with zooids in regular quincuncial arrangement, whereas S. dimorpha Taylor, 1986a has a pluriserial colony with zooids less tightly packed, as is more typical of Charixa. The three species we place in Kenocharixa have multiserial, sheet-like colonies with tightly packed zooids, although the arrangement of zooids is less regular in K. goshouraensis than in the other two. The frequency of closure plates may also provide a useful diagnostic character; zooids with closure plates appear to be variably present and generally uncommon in species of Spinicharixa and Charixa, but more common in two of the species we place in Kenocharixa; Gordon and Taylor (2015) did not observe them in K. stamenocelloides, although they examined only two small specimens. Interestingly, the key character separating Kenocharixa from Spinicharixa and Charixa—the ramifying kenozooids—occurs in rudimentary form in Charixa; one of Taylor's (1986a: fig. 7c, p. 206) figures shows a kenozooid arising from the fusion of proximal buds from several autozooids and sending short extensions into interzooidal grooves.

Diagnosis.—Colony encrusting, unilaminar, sheetlike. Zooids elongate-oval, gymnocyst negligible; opesia occupying most of frontal area; mural rim narrow, coarsely granulose. Spines, ovicells, and avicularia lacking. Paired triangular kenozooids at proximal corners of zooids. Zooids with closure plates uncommon, same size as autozooids, usually grouped. Small interzooidal polymorphs occur sparsely among autozooids. Zooids interconnect via buttressed recesses leading to multiporous septula. Ancestrula and periancestrular zooids produce closure plates.

Etymology.—The specific name is an adjective referring to Kashima, the municipality in which the type locality is located.

Material examined.—Thirty-seven specimens from Locs. 1, 3, and 4. Holotype, NMNS PA18399 (Loc. 3). Paratypes, NMNS PA18400–18410. See Table 1.

Measurements.—Measurements from autozooids and closure plate zooids are in Table 2. Interzooidal polymorphs: IpL = 0.23–0.39 mm (0.292±0.048 mm); IpW = 0.12–0.23 mm (0.173±0.030 mm); IpOpL = 0.05–0.17 mm (0.103±0.028 mm); IpOpW = 0.04–0.09 mm (0.066±0.013 mm) [n = 15 from five colonies]. Ancestrula, NMNS PA19409: L = 0.241 mm; W = 0.150 mm; OpL = 0.086 mm; OpW = 0.076 mm [n = 1].

Description.—Colony encrusting, unilaminar, sheetlike (Figure 7A). Zooids in large parts of colony arranged in long, unbranched columns offset from one another one-half zooid length, giving rise to quincuncial pattern (Figure 7A). Large sectors of colony contain strictly parallel columns with no bifurcations; new sectors arise by distal-distolateral bifurcations on zooids along one side of single column. Autozooids (Figure 7B) elongate oval, delineated by narrow groove; gymnocyst negligible, covered by kenozooids. Cryptocyst moderately wide proximally and laterally around opesia, sloping at steep angle (Figure 8B), coarsely granulose; opesia occupying most of frontal surface. Most zooids produce paired triangular kenozooids (Figures 7B; 8B, C) proximolaterally; kenozooidal surface coarsely granulose, cryptocystal; kenozooids sometimes merging, forming single, broad proximal kenozooid (Figure 7B). In some cases, central opening of kenozooid seen to lead to one or two pores in proximal gymnocyst of underlying autozooid. Kenozooids extend lobes of calcification into adjacent interzooidal grooves (Figure 7B, D); with age, lobes completely fill zooidal grooves (Figures 7C; 8A, B), eventually rising above level of mural rim (Figure 8C) and forming coarse, nodular thickening covering proximal end of zooids (Figure 8D, E). Interzooidal polymorphs occur uncommonly among autozooids; usually much smaller than autozooids (arrowheads, Figure 7C) but occasionally nearly as large (upper two arrowheads, Figure 7D), with elliptical opesial opening and rounded mural rim; rim coarsely granulose in zone around opesia but smooth around margin. Zooids with closure plates occur uncommonly, singly or in groups of two to seven (Figure 7D); opesia reduced, oval or elliptical in outline, surrounded by tumid, rounded, coarsely granulose cryptocystal rim, with narrow zone of smooth gymnocyst around margin; opercular impression smooth, surrounded by horseshoe-shaped groove corresponding to opercular margin. Interzooidal polymorphs can occur among zooids with closure plates (Figure 7D). Zooids interconnect via distal and paired distolateral buttressed recesses, each leading to multiporous septulum (Figure 8A). Ancestrula (Figure 8F) budding triplet of zooids distally and distolaterally; ancestrula, triplet of zooids, and some but not all zooids in second and third generations from ancestrula forming closure plates; ancestrula and first two generations of daughter zooids with tapering proximal gymnocyst lacking kenozooids, which appear in some zooids in third generation from ancestrula. Intramural budding not observed.

Remarks.—The species most similar to Kenocharixa kashimaensis is the congener K. goshouraensis (Dick et al., 2013). Zooid size is similar, and in the zone of astogenetic repetition both species have elongate zooids with the opesia occupying most of the frontal area (compare Figure 7A herein with Dick et al., 2013: fig. 8A). The arrangement of zooids in the colony is less regular in K. goshouraensis, where zooids in some parts of the colony occur in quincunx but in other parts are arranged in non-offset columns, so that zooids are side-by-side in rows. Zooids of K. goshouraensis generally have a more extensive gymnocyst, especially proximally but also laterally, resulting in larger interzooidal grooves; in the zone of astogenetic change, the proximal gymnocyst can be extensive, with the opesia oval and only half as long as the zooid. As in K. kashimaensis, the kenozooids in K. goshouraensis that later elongate in the interzooidal grooves originate proximolaterally, but they are initially smaller, less regularly triangular, and less stereotyped in position (they can be paired, or lacking on one or both sides), and they tend to elongate primarily in the lateral groove between zooidal columns, rather than also filling the transverse groove between zooids (Dick et al., 2013: fig. 7A–E). Kenocharixa goshouraensis lacks the medium-sized interzooidal polymorphs seen in K. kashimaensis. The closure plates are similar between the two species, but those in K. kashimaensis are frontally tumid and convex, whereas those in K. goshouraensis are somewhat sunken inside the raised margin and have a smaller opesia. In K. goshouraensis, some intramurally budded zooids were seen to form closure plates (Dick et al., 2013: fig. 7F), whereas no intramural budding was observed in K. kashimaensis. The ancestrular budding pattem is identical in the two species; in both, the ancestrula gives rise to a triplet of daughter zooids distally and distolaterally, and the ancestrula and at least some zooids in the first three generations from the ancestrula have closure plates.

Zooids like those in Figure 7B might be interpreted as lacking heavy secondary calcification because they are young. However, Figure 8E shows less heavily calcified zooids adjacent and distal to heavily calcified zooids, suggesting that the former represent instead a basal primary layer of calcification that is typically overlain by a heavy secondary layer, with the latter having been diagenetically dissolved in part of the colony. The mechanism by which this could occur is not clear.

Type species.—Charixa vennensis Lang, 1915.

Material examined.—NMNS PA18434 (Loc. 1), one small colony.

Measurements.—ZL = 0.28–0.42 mm (0.349±0.037 mm); ZW = 0.18–0.27 mm (0.225±0.026 mm); ZOpL = 0.22–0.29 mm (0.247±0.024 mm); ZOpW = 0.11–0.18 mm (0.139±0.021 mm); n = 15.

Description.—Colony (Figure 9A) encrusting, unilaminar, irregular, forming pluriserial lobes. Autozooids mostly arranged in quincunx; irregularly hexagonal or oval, delineated by deep, moderately wide groove. Gymnocyst (Figure 9C) evident proximally and proximolaterally, but probably not laterally in zone of astogenetic repetition; long proximal cauda lacking. Mural rim raised; granulated cryptocyst sloping, widest proximally, tapering in width laterally around opesia, lacking distally. Opesia (Figure 9A–C) occupying most of frontal area, elongate-oval or elliptical, widest in center or proximal half. Zooids bear pair of small spine bases at distolateral corners (arrowheads, Figure 9C); ancestrula, periancestrular zooids (arrowheads, Figure 9D), and at least some autozooids (arrowheads, Figure 9B) show a larger spine base on one side in proximolateral part of mural rim. Interzooidal polymorphs occur as intramurally budded kenozooids (arrowheads, Figure 9A); center of frontal wall with small, elliptical or circular opesia. Ancestrula (arrow, Figure 9D) tatiform, with three pairs of spine bases; initially producing distal and paired distolateral daughter zooids; ancestrula possibly with closure plate, with traces of grooves corresponding to lateral opercular margins, although these apparent grooves could be casting artifacts, in which case the opesia is occupied by an intramurally budded kenozooid. Zooid distal to ancestrula with two intramural buds, with intramurally budded kenozooid forming inside those. Other zooids up to third generation from ancestrula with intramurally budded kenozooids. Proximolateral spine base enlarged on one side in ancestrula and two subsequent zooids (arrowheads, Figure 9D). Ovicells and avicularia not observed. Interzooidal connections evident as distal and one or two pairs of distolateral buttressed recesses.

Remarks.—We assign this species to Charixa on the basis of the generic diagnosis by Taylor (1986a). The colony is irregularly arranged; zooids have a moderately well developed proximal gymnocyst; the cryptocyst is narrow and coarsely granulose; and in most zooids, spine bases appear to be limited to a small distal pair. The occurrence of a single larger, spine base proximolaterally on the mural rim in some zooids might represent an enlarged spine or scutum. Incertae sedis A (next description) shows a raised structure in the same position, which might suggest that the two species are related, perhaps comprising a previously unrecognized genus.

Material examined.—NMNS PA 18435 (Loc. 3), one colony, partly overgrown by Kenocharixa kashimaensis.

Measurements.—ZL = 0.39–0.49 mm (0.425±0.032 mm); ZW = 0.25–0.32 mm (0.275±0.021 mm); OpL = 0.21–0.28 mm (0.242±0.021 mm); OpW = 0.15–0.18 mm (0.161 ±0.010 mm); all measurements, n = 15.

Description.—Colony (Figure 10A) unilaminar, encrusting, sheet-like. Autozooids (Figure 10B, C) oval in outline, delineated by deep groove; arranged in irregular rows, with zooids only occasionally in quincuncial pattern (Figure 10A, B). Gymnocyst evident proximally and proximolaterally. Mural rim raised, rounded, widest proximally, tapering laterally, narrow distally, covered with cryptocystal granulation; cryptocyst steep, lacking distally. Opesia oval to elongate-oval; distal end straight. Some autozooids show coarse pair of short, erect spine bases at distolateral corners (arrowheads, Figure 10C, D); small spine bases possibly present around mural rim but difficult to discern. Most zooids have a raised nodule (arrows, Figure 10C) proximolaterally on mural rim on one or occasionally both sides; nodule of ambiguous structure but in some cases appearing hollow, with opening directed proximomedially (arrow, Figure 10D), possibly representing either small avicularium or base of large spine or scutum. No ovicells, avicularia, or closure plates observed; ancestrula not observed. Nature of interzooidal connections unclear.

Remarks.—This species, with zooids having a proximal gymnocyst and a pair of distolateral spines, might be placed in Charixa, although the extensive, coherent, sheet-like colony is unusual for that genus. The proximolateral nodular structure in Incertae sedis A is in the same position as the large proximolateral spine base in Charixa sp. A (above), raising the possibility that the two species are closely related and represent an undescribed genus. The arrangement of zooids is different between the two; in Charixa sp. A, a mostly quincuncial arrangement of zooids appears early on (Figure 9A). While the ancestrula is not evident in our specimen of Incertae sedis A, the portion of the colony in the lower left of Figure 10A appears to be within the zone of astogenetic change, and there the zooids are arranged more in transverse rows than in quincuncial pattern. Clarifying the identity of Incertae sedis A will require additional, better material than the single specimen we have.

Material examined.—NMNS PA 18436 (Loc. 1). The specimen consists of two fragments of the same colony in rock matrix. One fragment is part of a frontally abraded branch, with the walls partly dissolved; the other comprises the terminal part of a multiramous branch, the original frontal surface mostly intact but with the walls dissolved in one of the branch ends, showing the internal orientation of the zooids. SEM images were taken from the intact specimen rather than from a cast.

Measurements.—Frontal ZL = 0.28–0.37 mm (0.313 ±0.035 nm); Frontal ZW = 0.22–0.36 mm (0.269±0.050mm); OpL = 0.12–0.17mm (0.141 ±0.020 mm); OpW = 0.11–0.18 mm (0.125 ±0.026 mm); n = 6.

Description.—Colony erect, with subcylindrical, cylindrical, or compressed branches; terminal fragment of one branch (Figure 11 A) 7 mm long, 2.7 mm across, and 2 mm thick, with three rami at end; zooids completely surrounding branch. Zooids appear to be long, originating from central axis, overlapping one another for most of their length within branch, turning frontally near end (Figure 11A, upper right), with only most distal part of zooid showing frontally. Frontally exposed area irregular in outline: circular, oval, elliptical, or hexagonal. Opesia small; circular, oval, or elliptical; surrounded by sloping, granulated cryptocyst (Figure 11D). Gymnocyst lacking. Zooids interspersed with infrequent, irregularly occurring kenozooids (arrowheads, Figure 11C). No spine bases or ooecia observed. Neither early astogeny nor basal portions of colony observed.

Remarks.—Incertae sedis sp. A resembles species in the Cretaceous family Chiplonkarinidae Taylor and Gordon, 2007. Chiplonkarinidae includes malacostegangrade cheilostomes in which the colony is primarily erect, with long, tubular zooids forming an inner endozone and an outer exozone. Hollow spines and closure plates are lacking; the gymnocyst is reduced or lacking; and the cryptocyst is narrow and does not form a proximal shelf (Taylor and Gordon, 2007). The details of the inner and outer zones differ among the four genera (Bock, 2017) presently placed in the family. In Chiplonkarina Taylor and Badve, 1995, autacoids are oriented parallel to the long axis of the branch in the endozone, but turn 90° into the exozone, where they become perpendicular to the branch surface. In Heteroconopeum Voigt, 1983, the proximal parts of autacoids in the endozone are polygonal in cross section, stacked vertically along the center of the branch to form prismatic cylinders, and have pores in both the lateral and transverse walls between zooids; zooids bend sharply into the exozone and open on the colony surface. The endozone occupies a much larger proportion of the cross-sectional area of a colony branch in Heteroconopeum than in Chiplonkarina. In Zimmerella Taylor, 2008, the colony is erect and dendroid, and branches have a scalloped axial lumen surrounded by endozonal zooids, which are large and flask-shaped and produce exozonal zooids by frontal budding. Finally, in Basslerinella Taylor and McKinney, 2006, the colony forms bifoliate fronds rather than cylindrical branches; frontally, kenozooidal outgrowth fills in the spaces between autacoids and leaves a digitate edge surrounding the autozooidal opesia, which is extensive.

The autacoids in our specimen are clearly long, tubular, and turn frontally near their distal end (Figure 11A, upper right), suggesting the presence of an endozone and exozone, and indicating placement in Chiplonkarinidae. However, the frontal aspects of zooids evident in our specimen are not clearly identifiable with any of the four genera in this family. Determination of generic identity will ultimately require a cross-sectional view of a branch, and we were unwilling to risk damaging our only specimen, which in other respects is exceptionally good, to try to obtain this view. Further identification wifi require additional material.

Material examined.—NMNS PA 18437 (Loc. 2). The specimen, consisting of calcified walls embedded in rock matrix, was observed and photographed through a stereomicroscope; no cast was possible. In life, one or more fully developed bryozoan colonies, and at least two young colonies, inhabited the highly concave interiors of two small Crassostrea valves that were probably fused together side by side when alive and remained fused either in situ or among shell rubble after death. In our specimen, the bryozoans occupy the outer surfaces of the interior casts left by the Crassostrea valves, with the mollusk shells having disappeared to expose the basal surfaces of the bryozoans. While many of the zooids in the light photomicrographs in Figures 12 and 13 appear to have the frontal walls exposed, these are actually the basal walls, and what appears to be the orifice (e.g. Figures 12D, E; 13D) in many zooids is actually a circular or oval, non-calcified distal portion of the basal wall. This is clearly evidenced by parts of colonies where the rock matrix has been partly or completely lost from the interiors of zooids, as is evident in Figure 13A; in a few of these zooids, the outline of the opesia (arrowheads, Figure 13A) and the shape of the interior of the zooidal frontal wall can be discerned. Of necessity, the description below is based mostly on the basal surface.

Measurements.—Primary axial zooids: ZL = 0.65– 0.82 mm (0.738±0.057 mm); ZW = 0.10–0.19 mm (0.152±0.022 mm); n = 13. Secondary axial zooids: ZL = 0.41–0.76 mm (0.529±0.105 mm); ZW = 0.15– 0.21 mm (0.183±0.018 mm); n = 12. Tertiary zooids: ZL = 0.33–0.50 mm (0.411±0.062 mm); ZW = 0.017– 0.028 mm (0.237±0.028 mm); n = 11. Orifice, secondary axial zooids: OpL = 0.14 mm; OpW = 0.10–0.11 mm; n = 2.

Description.—Colony encrusting; zooids in uniserial columns arranged in bipinnate branching pattern (Figures 12, 13), although capacity to become pluriserial exists (Figure 13D). Young colonies (Figures 12A, 13D) with branches well separated and distinct; older, larger colonies with branches irregularly and closely spaced, jumbled (Figure 12B). Zooids in primary axis of branch (Figures 12D, E; 13B, C) each giving rise distally to next axial zooid in series and distolaterally to zooid on each side, initiating paired, opposite secondary axial branches. Zooids in secondary axis show same budding pattern, producing next axial zooid distally and pair of tertiary zooids distolaterally. Tertiary zooids typically bud subsequent tertiary zooids only distally, without giving rise to distolateral daughter zooids. Zooids in primary axis long and narrow, with long cauda (Figures 12E; 13B, C); noncalcified portion of basal wall typically lacking. Zooids in secondary axis wider and shorter on average (Figures 12D, E; 13C), with cauda of variable length; non-calcified portion of basal wall sometimes present. Tertiary zooids relatively short and wide (Figures 12D, E; 13B, C), oval or spindle shaped, often lacking cauda altogether, with relatively large, oval or circular, non-calcified zone in distal part of basal wall. Frontal wall (inferred from two zooids in Figure 13A) transversely highly convex; opesia oval, longer than broad, occupying roughly half the noncaudate portion of frontal wall, apparently oriented nearly parallel to basal surface of colony. Lateral walls with pair of basal tubular, intramural pore chambers distolaterally, one to three smaller tubular chambers on each side more proximally (Figure 12D, arrowheads), and simple pore distally (Figure 12D, asterisked arrowhead). Zooids in some parts of colony may form pluriserial patches in which zooids interconnect via all pore chambers, as seen in young colony (Figure 13D, E). Basal walls of some primary and secondary axial zooids convex in transverse section (evident in Figures 12E; 13B, C). Ancestrula (Figure 13D, E) oval, budding one daughter zooid distolaterally and one proximally. Frontal zooidal morphology not directly observed. No evidence of ovicells or avicularia.

Remarks.—Although we initially interpreted this as possibly a branched, erect species, with some branches preserved in such a way that the frontal surfaces of zooids were visible, this is not the case. Flexible, noncalcified joints between zooids (as would be expected in a branched, erect species) are lacking, and what appeared to be opesiae proved to be non-calcified zones in the basal walls, as evidenced by parts of the colony (Figure 13A) in which the matrix has been lost from the interior of zooids. Finally, the colonies in our specimen are essentially twodimensional; no branches were preserved as tangential cross sections, as might be expected in the preservation of a three-dimensional colony. We interpret cases where zooids or branches overlap one another (Figures 12D, E; 13B) to represent intra-colony overgrowth. A puzzling aspect of preservation is that some primary and secondary axial zooids are not flat on the basal surface, but rather are convex in transverse section (e.g. Figures 12E; 13B, C), as might be expected in the preservation of weakly calcified, cylindrical, erect branches. However, some zooids with a convex cauda have lost the matrix inside the expanded portion of the zooid, confirming that the convex surface is the basal surface.

We saw no indications of ovicells, which should be evident even in basal view if present, and so we consider this species to be of malacostegan rather than neocheilostome grade. In growth form, it resembles the electrid species Spinicharixa dimorpha Taylor, 1986a in which long, narrow, caudate axial autozooids bud the next axial zooid distally and paired, non-caudate zooids distolaterally. The resulting colony is pluriserial, with zooids in large parts of the colony packed closely together. In Incertae sedis C, primary axial zooids are similarly long and narrow, but give rise distolaterally to similarly narrow secondary axial zooids, which in turn give rise distolaterally to noncaudate zooids that produce further non-caudate zooids distally. The resulting colony is more a jumble of closely packed uniserial branches than a pluriserial sheet. The autozooids in S. dimorpha are frontally quite different than those in Incertae sedis C. In S. dimorpha, the opesia in both axial and non-caudate zooids occupies the entire frontal surface and the gymnocyst is negligible (Taylor, 1986a: fig. 25 A), whereas in Incertae sedis C, the inferred opesia is oval in outline and occupies one-third or less of the total zooid length. In S. dimorpha, the opesia is surrounded by a narrow, granulated cryptocyst and the mural rim bears up to eight pairs of spine bases.

Other genera having uniserial budding and reported from the Cretaceous include Rhammatopora Lang, 1915; Herpetopora Lang, 1914: Pyripora d' Orbigny. 1849; and Pyriporopsis Pohowsky, 1973. As in axial zooids in Incertae sedis C, zooids in Rhammatopora have a long cauda and the opesia is similar in size and shape to that inferred for Incertae sedis C. In Rhammatopora the caudae are much longer and narrower relative to the expanded portion, lateral zooids are budded in a cruciate pattern, and the opesial rim bears the bases of many small spines.

Zooids in Herpetopora species (Taylor, 1988b) also have a narrower and often longer cauda relative to the expanded portion than in Incertae sedis C. Lateral budding sites in Herpetopora involve straight, tubular intramural pores, usually one but sometimes more than one on each side. In contrast, the distolateral pores in Incertae sedis C are intramural but are expanded within the wall, and zooids often have an additional two lateral intramural pores more proximally on each side, although no budding was observed from these lateral pores. Zooids in some Herpetopora species have a relatively small, oval opesia as in Incertae sedis C, with the cryptocyst narrow and non-granulated.

The distinctions between Pyripora and Pyriporopsis are not entirely clear; Taylor (1994) noted that Pyripora species have a better developed proximal gymnocyst (presumably, longer cauda and/or smaller opesia; see also Thomas and Larwood, 1960) than Pyriporopsis; the cryptocyst is narrow and granulose (pustulose), whereas that in Pyriporopsis is striated or lacking. Taylor (1987) noted the following features characteristic of Pyriporopsis but did not contrast these with the character states in Pyripora: zooids predominantly oval in crowded areas of colony, but pyriform (with caudae) in uniserial series in less crowded areas; spines lacking; calcified closure plates occur; the pore chambers are dilated intramural cavities; and the ancestrula is small, oval, budding daughter zooids distally and proximally (the pattern is different in Pyripora catemdaria; Taylor, 1986b).

Available evidence suggests placement in Pyriporopsis. Our specimen is similar to Pyriporopsis portlandensis, the type species of Pyriporopsis, in the branching pattern and distribution of caudate and non-caudate zooids (Pohowsky, 1973: fig. 1), and in having intramural pore chambers distolaterally, with one or two pores present more proximally on each side. While we infer the opesia to be smaller in our specimen than in P. portlandensis, this was based on only two axial zooids, and tertiary zooids could in fact have a larger opesia. Finally, the early budding pattern (Figure 13D, E) in our specimen seems more similar to that in P. portlandensis than in Pyripora catenularia, although which zooid in our specimen is the ancestrula is somewhat ambiguous. We interpret zooid “a” in Figure 13E to be the ancestrula, giving rise to one daughter zooid (arrow) distolaterally and another (asterisk) proximally. We infer the polarity of the ancestrula from the dark, non-calcified basal area, which is presumably distal as in other zooids. The distolaterally budded daughter zooid in turn appears to give rise to a distal zooid and one distolateral zooid, though these zooids are not well preserved. The zooid proximally budded from the ancestrula buds a triplet of zooids distally and distolaterally, each of which in turn buds distally, as well as distolaterally on one or both sides. Pyriporopsis portlandensis shows a similar pattern (Taylor, 1986b: fig. 6) in that the ancestrula buds from the distal and proximal ends, but not laterally. We note, however, that correct identification of the ancestrula is crucial; if the asterisked zooid in Fig. 13E is actually the ancestrula, and if the polarity is in the opposite direction (distal to the right), then the budding pattern is more like that in Pyripora catenularia than in Pyriporopsis portlandensis.

We refrain from assigning our specimen to a genus, even tentatively, because we were unable to observe key frontal features necessary for generic determination. These include variation in the extent of the opesia; the nature of the cryptocyst, if present; and the presence or absence of opesial spine bases, zooidal closure plates, and kenozooids (the narrowness of some the axial zooids raises the possibility that they are kenozooids).

Type species.—Cellepora elliptica von Hagenow, 1839.

Diagnosis.—Colony encrusting, unilaminar, sheet-like. Zooids in zone of astogenetic change oval or hexagonal; those in zone of repetition elongate, and rectangular or nearly so; opesia oval or elongate. Zooids closely set; proximal gymnocyst up to one-quarter length of some zooids, lateral gymnocyst negligible. Cryptocyst coarsely granulose. Interzooidal avicularia common proximodistally or laterally between zooids. Ooecium globose, prominent, lying on gymnocyst of next-distal zooid; subimmersed in older zooids. Ancestrula budding triplet of daughter zooids distally; subsequent periancestrular zooids include large proximolateral pair, with smaller presumed kenozooid between them. Single or paired distal and one or two pairs of distolateral buttressed recesses leading to multiporous septula. Mural spines and closure plates lacking.

Etymology.— The specific name is an adjective from the Latin prolixus (long, drawn out), referring to the elongate zooids.

Material examined.—Seven specimens from Locs. 3 and 4. Holotype: NMNS PA18438 (Loc. 3). Paratypes: NMNS PA18439-18441 (Loc. 3). See Table 1.

Measurements.—See Table 3.

Description.—Colony (Figure 14A, B) encrusting, unilaminar, sheet-like. Zooids arranged in quincunx (Figure 14E, F) or somewhat irregularly (upper right, Figure 14D); zooids small and irregularly hexagonal or oval in zone of astogenetic change (Figures 14A; 15B, D), but longer in zone of astogenetic change (Figure 14D–F); also compare ZL measurements between NMNS PA 18440 and NMNS PA18438 (Table 3). Ovicelled zooids long-oval or rectangular (Figure 14E, F), delineated laterally by narrow groove. Opesia occupying most of zooid length in young zooids, roughly two-thirds in ovicelled zooids. Opesial rim raised, rounded, coarsely and densely granulose; sloping cryptocyst lacking, except as interior side of opesial rim. Mural spine bases not observed. Interzooidal avicularia common, forming at zooidal margin (Figure 15B) distal or distolateral to autozooids; abundant in interzooidal grooves in zone of astogenetic repetition (Figures 14C, D; 15A); sometimes less abundant in zone of astogenetic change (Figures 14A, B; 15D); longer than broad, with short, smooth gymnocystal area proximally and coarsely granulose rostrum (Figure 15C); no hinge structures observed. Some avicularia have elongate opesial opening, truncate at one end, tapering toward other, with narrow end directed proximally (e.g. Figure 14F); others have irregular, circular, or oval opening. Among ovicelled zooids, avicularia often lateral to ovicell (Figure 14E, F). Ovicell (Figure 14E, F) slightly wider than long, nearly as wide as autozooids, lying on proximal gymnocyst of zooid distal to maternal zooid; initially prominent (Figure 14E) but partly immersed with increased calcification (Figure 14F). Form of interzooidal connections ambiguous due to poor preservation or casting limitations; zooids have pair of small, distal buttressed recesses (arrowheads, Figure 14D) and one or two pairs of distolateral buttressed recesses (Figures 14D, 15A) that appear to lead to multiporous septula; autozooids interconnect with interzooidal avicularia as well as neighboring autozooids. Ancestrula (Figure 15B, D) oval or rectangular, surrounded by a triplet of daughter zooids distally, a presumed kenozooidal chamber proximally, and a pair of larger periancestrular zooids proximolaterally.

Remarks.—Most of the approximately 15 known species of Marginaria (Bock, 2017) were described from the Late Cretaceous, although two have been reported from the Paleogene (Berthelsen, 1962; Guha and Gopikrishna, 2007). Species in Marginaria characteristically have numerous small interzooidal avicularia scattered among autozooids; these avicularia are budded at the colony margin, along with developing autozooids (Taylor and McKinney, 2006), and commonly lack calcified hingesupport structures, leading some authors (e.g. Guha and Gopikrishna, 2007) to refer to them as kenozooids rather than avicularia. However, they are distributed as one would expect for avicularia. These putative avicularia appear to lack skeletal hinge support structures in various species of Marginaria, but whatever their function, they constitute a shared character for the genus.

The genus Reptoflustrella d'Orbigny, 1853 is similar to Marginaria but supposedly differs from the latter in having the interzooidal avicularia in a stereotyped position proximolateral to most autozooids; in addition, the generic description of Reptoflustrella includes zooids having at least two pairs of distal spine bases and the proximal gymnocyst comprising a quarter to two-thirds the zooidal length (Gordon and Taylor, 2005). The distinctions between the genera are somewhat unclear, as some of the avicularia in R. cenomana d'Orbigny, 1853, the type species of Reptoflustrella, are arguably not in the proximolateral position (Gordon and Taylor, 2005: fig. 1D), and some species in Marginaria have spines and an extensive proximal gymnocyst, as in Reptoflustrella. Voigt (1989) included R. cenomana in Marginaria in a review of the latter genus, thus reducing Reptoflustrella to a junior synonym of Marginaria. In our holotype specimen of M. prolixa, the interzooidal avicularia are commonly in the proximolateral position in the vicinity of ovicelled zooids (Figure 14E, F), but in non-reproducing parts of the colony, they are often positioned proximodistally between successive autozooids (Figure 14D). Furthermore, zooids in M. prolixa lack spines, as is the case in about half the species in Marginaria. While Gordon and Taylor (2005) believed that it was premature to include Reptoflustrella in Marginaria, our species appears in any case to fit better in Marginaria.

Most species in Marginaria have zooids less than 0.40 mm long, typically with an oval or elliptical opesia, and with gymnocyst evident both proximally and laterally. Another species with larger zooids is M. caminoides (Voigt, 1967) (ZL × ZW = 0.60 × 0.45 mm), which, although encrusting, differs from M. prolixa in forming narrowly pluriserial branches; zooids are short-oval or hexagonal in outline; the opesia is oval to elliptical; the interzooidal avicularia are large; and zooids have mural spines.