New Late Carboniferous Heritschioidinae (Rugosa) from the Kuiu Island area and Brooks Range, Alaska

Late Carboniferous. Colonial corals. Coral ontogeny. Kuiu Island. Brooks Range. KEYWORDS


INTRODUCTION
This paper is the first of two studies of the middle Late Carboniferous colonial coral faunas of the northeastern Kuiu Island area, part of the Alexander terrane in southern Alaska, and the Brooks Range, northern Alaska (Fig. 1). A more detailed discussion of the Late Carboniferous paleogeography and relationships of various coral faunas will be addressed in that second paper. However, here we note that the Alexander terrane and many other allochthonous terranes accreted to the northern and western margins of Pangaea in the Mesozoic have been interpreted to have been derived from a region between Baltica, Laurentia, and Siberia, in proximity of the northern Caledonides in the Early Paleozoic (Colpron and Nelson, 2011). In the mid Paleozoic time these terranes were dispersed westward and according to Colpron and Nelson (2011) the Alexander and Wrangellia terranes, which were hypothesized to have been amalgamated in the Late Carboniferous (Gardner et al., 1988), were somewhat isolated from Laurasia out in the Panthalassa Ocean throughout Late Carboniferous to Early Permian times. During these times the Brooks Range is interpreted to have been in place in northern Alaska (Colpron and Nelson, 2011). In contrast, Katvala and Henderson (2007) placed part of the Wrangellia terrane (Vancouver Island) relatively close to North America, at least in the Early Permian, closer to Pangaea than proposed by Colpron and Work in progress by Erik Katvala suggests that all of the specimens from the Kuiu Island area described here are Moscovian (Late Carboniferous) in age. The fauna from the Brooks Range was dated by Armstrong (1972) as Atokan (Bashkirian, Late Carboniferous) in age. The coral faunas from these two regions are quite different from all other known coral faunas of middle Late Carboniferous age except for the presence of the somewhat similar type species of Heritschioides in the Quesnel terrane in southern British Columbia.

STRATIGRAPHY
In the first work on the stratigraphy of the Kuiu Island area, Buddington and Chapin (1929) divided the upper Paleozoic rocks into the "lower division of the Permian" and the "upper division of the Permian". According to Muffler (1967) his newly named Halleck Formation, corresponds to the "lower division" whereas the Pybus Formation of Loney (1964) was considered by Muffler (1967) to correspond to the "upper division of the Permian" of Buddington and Chapin (1929). Muffler (1967) recognized and named the older Carboniferous rocks in the area: the "crinoidal limestone" and the Saginaw Bay Formation. Katvala and Henderson (2007) and Katvala et al. (2009) have determined that the local Carboniferous and Permian stratigraphic column consists of the Cannery Formation, an unnamed "crinoidal limestone", the Saginaw Bay Formation, the Halleck Formation, and the Pybus Formation (Fig. 2). The Saginaw Bay Formation was divided by Muffler (1967) into four members. According to him, from older to younger, they were: the Volcanic Member, the Black Chert Member, the Chert and Limestone Member, and the Silty Limestone Member. According to Katvala (personal communication, 2013), however, the Chert and Limestone Member should be included in the Silty Limestone Member, and the Black Chert Member is Devonian in age, based on conodont faunas, and therefore does not belong in the Saginaw Bay Formation. Here the two members of the Saginaw Bay Formation shown on Figure 1 are the Silty Limestone Member, which includes the Chert and Limestone Member of Muffler (1967), and the Volcanic Member.
The corals reported upon here are from the Saginaw Bay Formation. An early Middle Pennsylvanian (Atokan) fauna, probably from the Chert and Limestone Member (sensu Muffler, 1967) of the Saginaw Bay Formation, was first identified by Dutro and Douglass (1961). On the basis of recent conodont studies (Katvala, personal communication, 2013), however, the Saginaw Bay Formation is interpreted to be mostly if not entirely of Moscovian age. This unit is underlain by the "crinoidal limestone" of Bashkirian age (Katvala, personal communication, 2013), and overlain by the Halleck Formation, considered Early Permian in age, based on conodont faunas (Katvala and Henderson, 2007;Fig. 2).

MATERIAL AND METHODS
Samples from seven localities in the Kuiu Island area, six of which were utilized in this study, were available for study. United States Geological Survey localities are indicated by numbers followed by the letters PC (e.g. USGS 27727-PC). Localities of Erik Katvala are indicated by a combination of letters and numbers. Where several samples of the same species were collected, individual colonies in both collections are indicated by an additional number (e.g. #1). The geographic position of localities is shown on Figure 1, and detailed descriptions of localities and their stratigraphic positions are indicated in Appendix I. Thin sections of all figured specimens have been assigned the letter C followed by the number of that slide. Slides of all holotypes are housed in the collections of the National Museum of Natural History and bear a USNM number as well as a C number. Paratypes, other figured specimens, and other samples collected by Katvala, many containing solitary corals, have been placed in the collections of the University of California Museum of Paleontology at Berkeley.
All specimens have undergone various diagenetic alterations: compression, replacement by dolomite and/or silica, recrystallization, and dissolution resulting in small geode-like structures within skeletons. Those alterations are irregularly distributed within colonies, leaving some corallites or parts of them preserved well enough for photography of the microstructure of septa and for preparation of series of peels elucidating the blastogeny. Some of those series were photographed and drawn using a computer method.
In the description of all species the smallest diameter (d), as seen in transverse section, was measured (in millimeters), and the number (n) of major septa counted. These measurements are recorded as n:d values.    Lin et al. (1995), a position accepted by Fedorowski et al. (2007) who restricted clinotabulae to the Prae-and Paleo-Tethys taxa, thus informally eliminating that character from the diagnosis of Heritschioides established by Sando (1985, p. 979). Thus, all taxa bearing that important feature are unrelated to the Heritschioidinae and are eliminated from the consideration. Also, the occurrence of two orders of septa -major and minor -considered diagnostic by Sando (1985), are observed in many unrelated taxa and must be eliminated from consideration of subfamily and generic placements.

SYSTEMATIC PALEONTOLOGY
In this paper we have added a cerioid-aphroid genus to the list of genera included in the subfamily. Thus, the subfamily now consists of Heritschioides Yabe, 1950, Copia Vassilyuk and Kozyreva, 1974and Kekuphyllum gen. nov. GENUS Heritschioides YABE, 1950 Type species: Waagenophyllum columbicum SMITH, 1935 Heritschioides kuiuensis FEDOROWSKI and STEVENS, sp. nov. Figures 3-7 Etymology. Named for the Island of Kuiu from which the corals were collected Holotype. USNM 545244 from sample FH-C-F2 #4. Locality and age of the holotype. FH-C-F2; Moscovian Material. Holotype listed above with nine transverse and three oblique longitudinal thin sections, and nineteen peels. Paratype 1, sample FH-C-F2 #3 with four transverse and six longitudinal thin sections, and 24 peels; paratype 2, sample USGS 21334-PC #1, with two transverse and one longitudinal thin section; paratype 3, sample USGS 5134-PC with one transverse thin section (not illustrated). Sample USGS 4305-PC #2 with one transverse and one longitudinal thin section is questionably assigned to this species. See Appendix I for locality information.
Description. Corallum fasciculate. Corallites closely packed. External wall 0.1-0.2mm thick. Corallite diameters generally 6-16mm; n:d = 20:5, 20-22:6, 23:7, 27:11, 28-30:13-14. One corallite 32:23x19mm ( Fig. 3F) with axial structure either lacking or diagenetically destroyed. Major septa long, moderately thin in dissepimentarium, slightly dilated at margin of tabularium, most lack taper, some penetrate axial column to become confluent with septal lamellae. All those characters similar but variable within and between colonies beginning in late neanic growth stage (compare small corallites in Figs. 3C, G; 7A, B, D, H, I), may be differently accentuated at different growth levels of same corallite. Major septa vary during corallite growth from almost completely confluent through partly isolated to almost completely free from axial structure (e.g. Fig. 7C-A, respectively). Cardinal septum slightly shortened in some largest corallites, hardly distinguishable or elongated towards axial structure in others ( Fig. 7A-C). Axial structure may extend into cardinal fossula when cardinal septum slightly shortened (Figs. 4G, right; 7D, right corallite). Minor septa generally short, some extend as prongs into tabularium. Axial column one third to onehalf corallite diameter; in transverse view composed of many straight or wavy, very thin septal lamellae including extra lamellae. Length and arrangement of both kinds of septal lamellae highly variable within a single colony; best exemplified by holotype ( Fig. 3A-C, G), also present in all paratypes. Up to eight or more rows of steeply dipping axial tabellae and few septal lamellae when longitudinal section well oriented (Figs. 3D, E; 7E, G, J); irregular, shorter, more numerous, and less steeply arranged when section slightly eccentric, i.e. when periphery of axial structure is sectioned (Fig. 7F, upper). Linked axial tabellae at axial column periphery different in shape within same corallite; longer when axial lamellae rare, occasionally interrupted by inner tabellae of tabularium. In transverse section median lamella commonly obscure, thin, absent from parts of mature growth of several corallites, elongated towards and connected to cardinal septum in immature and some mature corallites. Axial structure either absent from some mature corallite growth stages or diagenetically damaged (Fig. 3F). Shallow cardinal fossula marked by slight depression of dissepimentarium in rare corallites, hardly distinguishable in most. Dissepimentarium occupies 1/5-1/4 corallite radius, generally composed of two to four ranks of small, mostly globose dissepiments as seen in longitudinal section (Figs. 3D, E; 7E, F, G, J).
In transverse sections interseptal, irregular dissepiments located mainly in middle part of dissepimentarium, few lonsdaleoid. Small dissepiments, oblique to larger ones and to septa, grouped mostly at periphery of largest specimens, constitute first step in formation of pseudonaotic septa. Tabularium with incipient biformity in rare septal loculi, composed of complete and incomplete tabulae rather steeply inclined upward to axial column, eight-10/mm. Boundary of tabularium with dissepimentarium marked by light stereozone thickening.
Microstructure diagenetically altered in most septa ( Fig.  4E, F). Remnants of fine trabeculae apparent in oblique sections of reasonably well preserved septa in holotype and one paratype (Fig. 4C, D, respectively), 30µm wide, and separated from one another by similar distances; diagenetic alterations may have increased trabeculae thickness. Shape and width of "dark middle lines" (= the primary septa) depends on orientation of thin sections; in well oriented transverse sections (Fig. 4E) fans of crystalline fibrils of various sizes form an axial sector of septum irregular in shape and width, probably result of significant diagenetic alteration of trabeculae; similar appearing "middle dark lines" in septa in other specimens in all three species of Heritschioides described in this paper also probably represent fine trabecular microstructure. Longitudinal sections exhibit growth bands within septa, shape of which depends on orientation of septum relative to cut surface; bands appear more distinctive when cut oriented obliquely to septal axis and tend to unite when cut perpendicular to axis ( in showing diverging bunches of microcrystalline fibrils, suggesting its secretion in a narrow endodermal fold, closely comparable to a septal pocket. That microstructure differs from dividing walls, dissepiments (Fig. 4H, arrow with 'D') and tabulae, in which microcrystalline fibrils grow only in one direction (as has previously been documented independently by Wells (1969) and Sorauf (1970). Partition microstructure similar to that of septa suggesting trabecular origin.
Blastogeny investigated in short series of transverse sections of three neighboring holotype offsets ( New late Carboniferous Heritschioidinae (Rugosa) from Alaska 36 C). Two marked 'a', 'b' derived from upper left parent's corallite; offset 'c' derived from upper right parent's corallite. Information from immature corallites in random transverse sections also employed. Hystero-brephic growth stage unknown. In early hystero-neanic growth stage ( Fig. 5A, B, corallite 'a') axial septum intersects offset's lumen; other septa, exclusively major, short and wavy. Cardinal septum, probably an atavoseptum, located at offset's periphery, withdrawn from corallite axis at slightly more advanced growth stage, not shorter than other major septa. Counter septum remains long (Fig. 5C, corallite 'a'). Withdrawal of cardinal septum from corallite axis variable. In offset 'b' level at which cardinal septum withdrew unknown; that septum increases in length during growth ( Fig  Early ontogeny of one protocorallite attached to a mature corallite inside paratype 1 (Fig. 6) has allowed the first complete study of this kind to be published. Diagenetic alterations destroyed some details, including earliest ontogeny; most important characteristics of protocorallite preserved. First two drawings (Fig. 6A, B) correspond to curved part of specimen. Obliqueness and dolomitization precludes deciphering increase and arrangement of septa in brephic growth stage. In ontogenetically earlier stages (Fig. 6A), only vague outlines of probable septa recognizable within sclerenchymal mass; three bodies, septal in microstructure, intersect corallite lumen 0.2mm higher (Fig. 6B). Left body may correspond to cardinal septum (Fig. 6C, marked by an arrow), which with major septum in middle dominate in next 0.8 millimeters of growth ( Fig. 6D-G). Increased number and arrangement of septa irregular. Protosepta do not dominate in early ontogeny (Fig. 6 D-G); become obvious only after axial septum clearly constituted (Fig. 6I). Length and attachment of other major septa to one another also changes, commonly quickly, during neanic corallite growth ( Fig. 6D-H); reason for irregularity unknown, corresponds to that observed in blastogeny (see above). Arrangement of major septa becomes more regular after axial septum formed ( Fig. 6I-L); still number of septa in various quadrants uncertain. Recognizable alar septa marked by black quadrangle (Fig. 6H-L).
Late neanic growth stage of specimen arbitrarily placed at insertion of first dissepiments and appearance of minor septa in corallite lumen, both appearing in right cardinal and right counter quadrant first (Fig. 6H, I). Axial structure formed at approximately same level, consists of axial part of axial septum, inner margins of major septa attached to it, and several sections of axial tabellae; remains united with inner margins of major septa for at least 1.2mm of growth ( Fig. 6I-L). All axial skeletal elements, including axial tabellae, strongly thickened by sclerenchyme; this and diagenetic alterations make axial structure -perhaps an axial column -appear solid. Individual skeletal elements scarcely recognizable in peels; some small details in drawings subjective interpretations of senior author.
In early ontogeny and hystero-ontogeny (blastogeny) protosepta connected and in variable positions; major septa variously elongated. In contrast to hystero-ontogeny, major septa, including protosepta not shortened during early ontogeny, and both protosepta equally strong when axial structure begins to develop; strong and regular in young protocorallite, weak and irregular in offsets. Absence in literature of any detailed studies of ontogeny and blastogeny of any other species of Rugosa, excludes possibility of making comparisons.
Discussion. H. kuiuensis differs from the type species of the genus, H. columbicum (Smith, 1935), from the late Serpukhovian to early Bashkirian from Keremeos in southern British Columbia, Canada, in possessing a much wider and more complex axial column composed of long regular and extra septal lamellae, in having a much narrower and simpler dissepimentarium with very few lonsdaleoid dissepiments, and showing the beginning of pseudonaotic septa. H. kuiuensis differs from the Atokan species, H. separatus sp. nov., from the eastern Brooks Range, Alaska, in the development of a much more complex axial column clearly isolated from the tabularium and consisting of numerous regular and extra septal lamellae, and in a narrower, simpler dissepimentarium. That colony was described and figured by Armstrong (1972, pl. 4 Heritschioides splendidus FEDOROWSKI and STEVENS, sp. nov. Figures 8, 9 Etymology. Lat. splendidus-impressive. After large size and elegant morphology of corallites. Holotype. USNM 545245 from sample SOB F1. Locality and age of the holotype. One corallum from locality SOB F1, Kuiu Island, Alaska; Moscovian in age.
Material. Holotype listed above with 10 transverse and five longitudinal thin sections, and fourteen peels. See Appendix I for locality information.
Description. Corallum fasciculate; fragment available for study originally greater than 18cm diameter. Corallites seldom offsetting. Offsets appear at similar levels of astogeny (Fig. 8D). Dimensions measured at different   New late Carboniferous Heritschioidinae (Rugosa) from Alaska 42 confluent with septal lamellae, thin in dissepimentarium, highly dilated in tabularium, tapering very slightly if at all toward axial column when free-ended, rapidly and strongly tapering when extending towards axial column (Fig. 8A-G). Cardinal septum in mature specimens slightly shortened, located in inconspicuous, parallel-walled cardinal fossula. Minor septa most often extend through dissepimentarium, rarely penetrate into tabularium as prongs. Axial column continuous, one-forth to one-third corallite diameter, composed of many septal lamellae, including extra lamellae and numerous, densely packed axial tabellae. Two arrangements of septal lamellae shown by transverse sections of two adjacent mature corallites: centripetal with densely packed septal lamellae at periphery with few approaching and/or reaching corallite axis (Fig.  8A, D, G, corallite 'a'; Fig. 9A), and irregularly twisted ( Fig. 8B-D, F, corallite 'b'). In mature corallites median lamella indistinguishable irrespective of arrangement of septal lamellae. Eccentric longitudinal sections cut at dense, peripheral part of axial column (Fig. 8I) show septal lamellae closely crowded, in more centric sections (Fig. 8H, J) fewer septal lamellae exposed. Axial tabellae hemispherical with lateral tabellae well developed, clearly separating axial column from tabularium. In immature growth stages median lamella thin and wavy, connected with ( Fig. 9C) or slightly elongated towards cardinal septum. At more advanced immature growth stage, median lamella, as in corallite shown in Figure 9B, occurs as thin plate or is absent, cardinal septum slightly shortened; presence of cardinal fossula documented by arrangement of sections of tabulae (Fig. 9B, arrow). Remnants of inner margin of cardinal septum or peripheral margin of median lamella represented by thin, wavy strip in cardinal fossula (Fig. 9A). Dissepimentarium composed of up to six or eight ranks of small, globose, irregular and concentric, interseptal dissepiments in major part of dissepimentarium; at maturity short bunches of small dissepiments, incipient pseudonaotic structures, attached obliquely to peripheral limits of septa. Lonsdaleoid dissepiments absent. Tabularium with biform morphology when minor septa extend into tabularium (Fig. 9D, arrows), normal in loculi where those septa restricted to dissepimentarium (Fig.  9E), both types may be developed next to one another (Fig. 9F, biform indicated by arrows). Tabularium (Fig.  8H-J) composed mostly of incomplete, upwardly arched tabulae inclined gently upward to axial column, 10-12/ cm. Boundary between dissepimentarium and tabularium marked by light stereozone thickening; boundary between tabularium and axial column sharp.
Microstructure variously altered diagenetically. Based on comparison with similar structures in septa of H. kuiuensis sp. nov., bunches of crystalline fibrils in several septa (Fig. 9G), interpreted as trabeculae enlarged during recrystallization process at expanse of sclerenchymal sheets of septa.
Discussion. H. splendidus sp. nov. resembles H. kuiuensis sp. nov, but differs in having a narrower column, much larger corallites, having greater dilation of major septa in the tabularium, and lacking a recognizable median lamella in mature corallites. From H. columbicum (Smith, 1935) it differs primarily in the morphology of the axial column, which comprises both well developed normal and numerous extra septal lamellae, a shortened cardinal septum, an easily recognizable cardinal fossula in mature corallites, the lack of lonsdaleoid dissepiments and pseudonaotic septa, greater dilation of the major septa in the tabularium, and a hemispherical arrangement of tabellae in the axial column. The latter character also distinguishes it from H. kuiuensis.
Occurrence. As for the holotype. Material. Holotype specimen listed above originally greater than 8cm in diameter with four transverse and three longitudinal thin sections, and 20 peels; paratype 1, sample SMI 2004 #3 with two transverse and two longitudinal thin sections; paratype 2, sample SMI 2004 #5, with two transverse and one longitudinal thin section. Sample SMI 2004 F1 with six transverse thin sections and one longitudinal thin section questionably belongs to this species. See Appendix I for locality information.
Description. External wall about 0.1-0.2mm thick. Maximum corallite diameter 13mm; n:d =19:6.0-7.0, 22:8.9, 23:10.0, 24:9.0, 26:9.0-12.0. In transverse sections ( Fig. 10A-C, F) major septa penetrate very slightly into external wall or appear to be attached to it (Fig. 10H), commonly approach axial column, rarely continuous with septal lamellae, thin in dissepimentarium, slightly dilated in tabularium, distal ends tapering to blunt. Cardinal septum rarely and indistinctly shortened; more commonly equal in length to adjacent major septa, may join median lamella when one present. Cardinal fossula lacking. Minor septa commonly penetrate short distance into tabularium where inner margins dilated. Axial column continuous, one third to one-fourth corallite diameter, isolated by vertical lateral tabellae and up to five or more ranks of steeply dipping, elongate axial tabellae. Septal lamellae variable, representing most major septa in some corallites, few in others. Extra septal lamellae extremely rare, recognizable in some mature corallites. Median lamella commonly lacking, obscure when present. Dissepimentarium commonly one-third corallite radius, generally composed of three or four ranks of small, globose dissepiments as seen in longitudinal section (Fig. 10D, E, G); in transverse section dissepiments variable in size and shape: concentric, irregular and pseudoherringbone; lonsdaleoid dissepiments absent. Tabularium composed of complete and incomplete, slightly convex upward tabulae that rise moderately steeply upward to axial column. Biform morphology recognizable mainly in transverse sections, indicated by different arrangement of peripheral parts of tabulae (Fig. 10H). Very light stereozone developed at margin of tabularium. Boundary between tabularium and axial column sharp.
In transverse sections of septa (Fig. 10J) septal microstructure diagenetically altered into irregular bunches of crystalline fibrils; in longitudinal sections ultrastructure totally destroyed, but altered intersecting growth lines and rods interpreted as remnants of fine trabeculae (Fig. 10I, arrows with letters 'b' and 'a', respectively) are preserved. Blastogeny (Fig. 11A-M) investigated in two series of acetate peels from two offsetting corallites in holotype New late Carboniferous Heritschioidinae (Rugosa) from Alaska 44 narrow space for offset within dissepimentarium of parent ( Fig. 11A-C). Minor septa and dissepiments disappear from loculi between three major septa becoming separated into unequal segments: long inner segments with strongly thickened peripheral margins retained in parent; short, thin peripheral segments forming initial skeleton of offset attached to external wall in offset. Neosepta appear early at external wall in offset, suggesting peripheral segment of parent's middle septum of offset becomes cardinal septum in offset. Its thin, tabula-like inner part extends far beyond offset's axis (Fig. 11A-H, arrow). Cardinal septum in second offset intersects its lumen early in hystero-ontogeny (Fig. 11J, arrow), soon reduced to length of adjacent major septa (Fig. 11K, arrow), elongates again to reach elongated counter septum (Fig. 11L, arrow).
Development of counter septum differs in two offsets studied; hardly identifiable in first offset ( Fig. 11G-I), long in second (Fig. 11K-M); in both counter septa inherited from same middle major septum of parent corallite. Thus, both protosepta are atavosepta.
Alar septa of both offsets inherited from two of parent's major septa, one on each side of major septum that gave rise to offset's protosepta located in offset's peripheral wall from beginning of offsetting process (Fig. 11A, B, dots), clearly marked by being longest septa in cardinal quadrants and by short septa in counter quadrants adjacent to them ( Fig. 11D-H, K, dots).
Increase in number of offset's septa in part common with parent camouflaged by both long-lasting thickening of skeletal elements and formation of partition composed of short thickenings corresponding to major and minor septa. Shortest major septa adjacent to alar septa indicate septal insertion typical for Rugosa; details unknown.
Insertion of minor septa (Fig. 11D, E) appears cyclic because they arise in offset's lumen at approximately same level of growth, but apparently inserted in series having earlier been hidden in thick external wall. Dissepimentarium follows appearance of minor septa; best developed in cardinal quadrants in both offsets.
Development of axial structure forms main difference between two offsets. In last preserved growth stage of first offset, only a very long cardinal septum, slightly thickened in axial part and extending beyond offset's axis; counter septum attached to it by a tabula (Fig. 11I). In second offset counter septum elongated beyond corallite axis, joining major septum right of cardinal septum; later joining cardinal septum for short period (Fig. 11L). Later counter septum free, still extends to offset's axis (Fig. 11M). Long, curved, axially thickened median lamella, free from both protosepta (Fig. 11M), preceded by temporary presence of axial septum.
Blastogeny of two offsets differs in development of counter septum; similar in all other details. Cardinal septum participated in formation of median lamella in both instances, being dominant in first offset. Inconsistent length of protosepta is typical for genus.
Discussion. This species resembles H. columbicum, the type species of the genus, in the similar arrangement of axial tabellae in the axial column, in that the minor septa penetrate the tabularium, and in having rare extra septal lamellae appearing late in the hystero-ontogeny. The type species, however, has much greater n:d values, a wider, more complex dissepimentarium with lonsdaleoid dissepiments common, and pseudo-naotic structures well developed. H. splendidus sp. nov. differs primarily in having much larger n:d values, much denser axial column with many extra septal tabellae and a shortened cardinal septum located in a tabular cardinal fossula. The present species differs from H. kuiuensis sp. nov. in the smaller size of the corallites, a narrower, less complex axial column with extremely rare extra septal lamellae if any, and with longer minor septa. Corallites of this species differ from H. separatus sp. nov. in the smaller diameters of the corallites, the minor septa penetrating the tabularium, and in the major septa closely approaching the axial column in most corallites (see below).
Occurrence. As for the holotype.
Material. Only the holotype colony, considered by Armstrong (1972) as a paratype of his species Corwenia jagoensis (see discussion below). Three transverse thin sections made and illustrated by Armstrong (1972, pl. 4, fig. 3 and pl. 5, figs 3-4) and two well oriented longitudinal thin sections made by the second author were available for the study. All of Armstrong's transverse thin sections and the newly made longitudinal thin sections are illustrated here. Neither the blastogeny nor the microstructure of septa were studied. Although figure 2 on pl. 5 in Armstrong (1972) is labeled as belonging to the same colony as the New late Carboniferous Heritschioidinae (Rugosa) from Alaska 46 other figures, the section is so different from those made by one of the writers (Calvin Stevens), we suspect that it actually was derived from another colony.
Diagnosis. Heritschioides with extreme n:d values 34:20 and 31:24x20mm; thin major septa do not penetrate axial structure; minor septa restricted to dissepimentarium which occupies up to 1/3 corallite radius; cardinal septum may reach simple axial structure with median lamella obscure or absent; septal lamellae much smaller in number than major septa; extra septal lamellae absent or very rare; row of lateral tabellae in axial column incomplete; small lonsdaleoid dissepiments may occur.
Description. Corallites densely packed, but crowding increased by compaction as demonstrated by broken peripheral skeletal elements in some neighboring corallites (Fig. 12C, arrow). N:d values of early mature and mature corallites as in diagnosis and 30:21, 29:16.3, 29:15.6, 28:16; juvenile growth stages not observed. Major septa almost equal in length, rarely approach axial column in early mature corallites, slightly and equally thickened in tabularium, thin in dissepimentarium. Cardinal septum varies from slightly shortened to penetrating axial structure. Cardinal fossula barely recognizable or absent. Minor septa thin, restricted to dissepimentarium, some interrupted by small lonsdaleoid dissepiments. Axial structure in transverse section weak, occupies less than 1/3 corallite diameter, less than 1/4 in some. Thin median lamella hardly recognizable or absent. Septal lamellae thin and irregular, far fewer than number of major septa. Extra septal lamellae absent from all thin sections studied. Axial column in longitudinal section well developed, incompletely separated from tabularium; lateral tabellae absent from some parts of axial column. Tabulae mostly incomplete, some enter column. Tabellae steeply elevated adaxially near axial column, commonly anastomosing at periphery. Unquestioned biform tabularium not observed. Dissepimentarium 1/4 corallite radius in small specimens to slightly more than 1/3 in largest. Dissepiments small, most rectangular and irregular, rarely pseudoherringbone and/or lonsdaleoid; incipient pseudonaotic structures common at periphery of largest corallites. Armstrong, 1972 comprises colonies belonging to different genera. Its holotype and most paratypes do not belong in Corwenia, but they most also differ from Heritschioides. The paratype USNM 161040, the only exception, is here re-investigated. This coral was recognized as somewhat similar to Heritschioides by Armstrong (1972, p. 13), who characterized it as follows: "Thus, of all the specimens, it shows the closest relationship to the genus Heritschioides Yabe", which Armstrong accepted as a Permian taxon. We accept that relationship, but it is now known that the type species for the genus is late Serpukhovian-early Bashkirian in age and all other species of Heritschioides are middle Carboniferous rather than Permian (Fedorowski et al., 2007).

Discussion. Corwenia jagoensis
We do not make an attempt here to revise "Corwenia" jagoensis formally, but later we do plan to revise those specimens included in C. jagoensis, which are comparable to colonies from Kuiu Island. The Atokan (Bashkirian) "paratype" USNM 161040, re-described by us here as Heritschioides separatus sp. nov., possesses some of the largest corallites with the simplest axial structure among the species assigned to this genus. It also possesses fewer septal lamellae than the number of major septa, and apparently lacks extra septal lamellae. Both the large size and the simple axial structure are adequate for distinguishing this species from all remaining members of the genus.
Occurrence. As for the holotype. Discussion. Heritschioides is the only genus presently included in the Subfamily Heritschioidinae, although Copia Vassilyuk and Kosyreva, 1974 is here conditionally included in that subfamily. The fasciculate growth form of those genera is diagnostically important enough for the distinction between them and the cerioidaphroid Kekuphyllum gen. nov. Thus, we concentrate on similarities supporting the inclusion of Kekuphyllum in the Heritschioidinae rather than on differences. The early hystero-ontogeny is crucial in that respect. The following characters are most important for the subfamily identification of Kekuphyllum (Figs. 13-15): 1. either the axial septum or the elongated cardinal septum occurs at an early growth stage; 2 the median lamella of the axial structure is derived from those septa, 3; lamella extending towards the cardinal septum, commonly intersects all or most of the axial structure. 4. rare extra septal lamellae appear in axial structures of some corallites; and 5. the continuous axial column closely resembles that in Heritschioides. All those characters are common to Kekuphyllum is the only massive colony included in the Subfamily Heritschioidinae so far, but two species of Gorsky (1938), Lonsdaleia (Wentzelella) diversa and L. (W.) multivesiculosa from the so-called Barents Series of Novaya Zemlya, may belong to Kekuphyllum. We do not suggest that relationship in a formal way because several important characters of Gorsky's specimens, such as the microstructure of the intercorallite walls, the early blastogeny, and the occurrence of extra septal lamellae are unknown from his specimens and are not recognizable in his illustrations. Only the latter character may be suspected to occur in L. (W.) multivesiculosa judging from Gorsky's (1938, p. 120) expression "large number of tabulae intersections", which may refer to septal lamellae. Besides, Gorsky (1938) suggested a Late Carboniferous the word age (in the Russian three partite scale). Thus, his species may be Gzhelian or even Permian in age. Revision of those species and establishment of their stratigraphic occurrence is needed prior to making a final decision of their true relationship to K. sandoense on a generic level. The continuous axial column, lack or very weak development of a cardinal fossula with the cardinal septum slightly shortened in some corallites (Fig. 13F) and the cerioid-aphroid growth form of K. sandoense are important characteristics to be considered for comparison with the Novaya Zemlya species after they are revised.

Kekuphyllum
sandoense FEDOROWSKI and STEVENS, sp. nov. Figures 13-15 Etymology. Named for the late well known coral expert William J. Sando who provided the specimen for study.  Material. Type specimen listed above consisting of several large fragments apparently from one corallum originally at least 18cm in diameter with eleven transverse and seven longitudinal thin sections and 73 acetate peels. See Appendix I for locality information.
Description. Corallum cerioid-aphroid. Partitions mostly continuous, about 0.3mm thick between peripheral margins of septa, much thicker when septa laterally contiguous (Fig. 14D, F). Rarely observed interruptions (channels) terminate gently, closed by dissepiments arranged in rows perpendicular to partition and parallel to one another (Fig. 14A, arrow) or filling in intercorallite channels between partitions (Fig. 14C, arrow); sharply ended interruptions of partitions (Fig. 14B (Fig. 13A, B, D, E). Major septa long with peripheral margins thickened up to lateral contiguity with minor septa in some (Fig. 14E, F), some closely approach axial column, others may be continuous with septal lamellae, many free-ended, thin in dissepimentarium, slightly thickened at outer margin of tabularium, then tapering toward axial column. Counter septum indistinguishable from other major septa, rarely attached to slightly thickened median lamella with wavy and strongly tapering inner margin, free from median lamella in most. Counter septum not dominant even when continuous with median lamella. Cardinal septum of variable length: slightly shortened, equal to adjacent major septa, or extending to median lamella; axial structure points towards that septum when shortened (Fig. 13F). Minor septa vary in length, most confined to dissepmentarium, few penetrate outermost tabularium, some discontinuous. Cardinal fossula lacking in most, inconspicuous and does not penetrate dissepimentarium when present. Axial column prominent, continuous (Fig.  13C, G, H), from less than 1/2 to slightly more than 1/3 corallite diameter, composed of simple median lamella, numerous septal lamellae, including rare extra lamellae, and up to six rows of steeply dipping, very elongate axial tabellae; lateral tabellae numerous. Boundary between tabularium and axial column mostly distinct. Tabularium commonly normal. Incipient biformity indicated mainly by sections of tabulae attached to inner margins of some longest minor septa; may be observed in longitudinal sections (Fig. 15E, arrow and above it). Tabulae steeply inclined up to axial column, commonly seven or eight/5mm, commonly incomplete with peripheral and periaxial tabellae, rarely complete. Dissepimentarium about one-third corallite radius, composed of two-to several ranks of small irregular and concentric, globose dissepiments with lonsdaleoid dissepiments present at corners of several corallites; may extend around almost entire corallite in some (Fig. 14E). Pseudonaotic septa occur in several corallites. Sclerenchyme thickening at inner row of dissepiments common.
Blastogeny was investigated in several offsets of holotype. Diagenetic alteration, however, has limited the precision of the descriptions and drawings of series of acetate peels that were made.
Earliest growth stage observed ( Figure 15A) closely comparable to species of Heritschioides. Septa divided into thin external segments inherited by offset and attached to external wall and inner segments retained in parent corallite, peripheral margins of which strongly thickened at parent/ offset boundary. Axial septum of offset directly connected to middle thickened septum of parent. Continuous axial septum retained in hystero-ontogenetically more advanced offsets (Fig. 15B, lower left; C). Arrangement of major septa suggest cardinal septum in offset formed from peripheral segment of axial septum (Fig. 15C, white dot); arrangement of septa in offset and its symmetry irregular; recognition of axial septum relative to position of cardinal septum; uncertain in some (Fig. 15B, upper offset).
Three adjacent immature corallites separated by partitions (Fig. 15D) all have minor septa, which penetrate tabularium and complete dissepimentaria. Hysteroontogenetically youngest corallite (Fig. 15D, corallite 'a' cardinal [left] and counter septa marked by white dots) possesses thin, inconspicuous median lamella connected to curved inner margins of both protosepta; some major septa unite with their lamellae in axial structure. In corallite intermediate in advancement in hystero-ontogeny ( Figure  15D, corallite 'b') cardinal septum (marked by white dot) directly connected to slightly thickened median lamella to which most major septa unite. Most advanced corallite in hystero-ontogeny (Fig. 15, corallite 'c', cardinal septum marked by white dot) possesses axial structure completely separated from inner margins of major septa with free, slightly thickened median lamella which points towards shortened cardinal septum shown by short intercept attached to section of a tabula. Differences develop in corallites of any growth stage as indicated by both illustrated (Fig. 13A, B, D-F) and non-illustrated corallites.
Discussion. Being the only species of the genus known so far, K. sandoense does not require discussion other than that of the genus. The only existing taxa similar to our species are pointed out in that discussion.
Occurrence. As for the holotype.

SUMMARY
Recent studies on the Genus Heritschioides and genetically related and/or morphologically similar taxa (Fedorowski et al., 2007;Kawamura and Stevens 2012;present paper) allow some introductory conclusions. These studies show that the stratigraphic position of all representatives of the genus Heritschioides described so far is much older (Fedorowski et al., 2007;Kawamura and Stevens, 2012) than previously thought. The type species, H. columbicum (Smith, 1935), is not Permian, but late Serpukhovian/early Bashkirian in age, thus being the stratigraphically oldest species of the genus. Other species are younger, upper Bashkirian to Moscovian in age. Thus, the genus can be considered characteristic for these ages.