LIBRARY OF THE UNIVERSITY OF ILLINOIS AT URBANA-CHAMPAIGN 550.5 FI GEOLOG^r UNIVEF-lSiTV OF ILLINOIS LIBRARY AT URBANA-CHAMPAIGN GEaOGY i^^'pr FIELDIANA Geology Published by Field Museum of Natural History VOLUME 24 REVISION OF THE HOLOCrSTIlES FAUNA (DIPLOPORITA) OF NORTH AMERICA CHRISTOPHER R. C. PAUL r— en C3* SEPTEMBER 29, 1971 MAY 1 5 1972 GEOLOnv (./BR/(Ry Urn vers »t l.irb" FIELDIANA: GEOLOGY A continuation of the GEOLOGICAL SERIES of FIELD MUSEUM OF NATURAL HISTORY VOLUME 24 AS NATURAL ^\ ^*r HISTORY >\ FIELD MUSEUM OF NATURAL HISTORY CHICAGO, U.S.A. FIELDIANA Geology Published by Field Museum of Natural History VOLUME 24 REVISION OF THE HOLOCrSTITES FAUNA (DIPLOPORITA) OF NORTH AMERICA CHRISTOPHER R. C. PAUL Assistant Professor of Geology Indiana University Northwest SEPTEMBER 29, 1971 PUBUCATION 1135 Library of Congress Catalog Card Number: 70-152389 PRINTED IN THE UNITED STATES OF AMERICA BY FIELD MUSEUM PRESS TABLE OF CONTENTS PAGE Introduction 1 Acknowledgements 2 Taxonomic history 3 Geographic and stratigraphic distribution 5 Preservation 8 Growth 14 Functional morphology 24 Internal Anatomy 32 Paleoecology 33 Systematic paleontology 44 References 156 Appendix Holocy stiles fauna localities 162 REVISION OF THE HOLOCYSTITES FAUNA (DIPLOPORITA) OF NORTH AMERICA The Holocystitidae, with the associated genus Triamara Tillman, represent a brief flourishing of diploporite cystoids during the Middle Silurian in North America which apparently had its origins in Europe during the Upper Ordovician. Since all the American species were originally described under Holocystites s.l. the term "Holocystites fauna" is used for lack of a better name. The Holocystites fauna includes almost all the known Silurian Diplo- porita and for this reason alone warrants study. The geographical distribution of the Holocystites fauna is not coextensive with the outcrop area of Niagaran rocks in North America (see fig. 10) which suggests contemporaneous ecological control. Depth of water and suitable substrate for attachment may have been determining factors. Some evidence of adverse environmental conditions, and the reaction of the cystoids to these conditions, is preserved. Several cystoids continued to live and grow after toppling over onto one side due to inadequate substrate for attachment. Others suffered numerous small borings, the activities of an unknown animal which never penetrated the thecal cavity. Quite apart from these interesting paleoecological and paleo- geographical aspects, the entire fauna is in need of systematic revi- sion largely due to S. A. Miller's use of thecal plate arrangement as a specific character. In this fauna plate arrangement is so variable that Miller was able to describe more than 40 "species" between 1878 and 1895. Further taxonomic difficulties arise from the manner in which James Hall originally erected the genus Holocystites (see Taxonomic History) and from the fact that in many cases the first found and earliest described species were based upon poorly preserved or incomplete specimens. This paper is an attempt to clarify and systematize the taxonomy of the Holocystites fauna, to trace its evolution from its European origins, and to investigate the paleoecology, functional morphology, and mode of life of these primitive Silurian echinoderms. ACKNOWLEDGEMENTS Grateful thanks are extended to the following for the loan of, or access to, specimens in their care: Dr. R. L. Batten, American Museum of Natural History, New York (AMNH); Dr. R. P. S. Jefferies and Dr. H. G. Owen, British Museum Natural History, London (BMNH) ; Dr. W. Beecher and Mr. A. Pivorunas, Chicago Academy of Science (CAS); Dr. T. E. Bolton, Canadian Geological Survey, Ottawa (CGS) ; Dr. E. S. Richardson Jr., Field Museum of Natural History, Chicago (FMNH); Dr. A. S. Horowitz, Indiana University, Department of Geology, Bloomington (lU); Dr. B. Kummel and Mr. J. Sprinkle, Museum of Comparative Zoology, Harvard (MCZ) ; Dr. J. K. Pope, Miami University, Oxford, Ohio (MU) ; Dr. J. Burke, Orton Museum, Ohio State University (OSU) ; Dr. H. Mutvei, Naturhistoriska Riksmuseum, Stockholm (RM); Dr. J. Monteith, Royal Ontario Museum, Toronto (ROM) ; Mr. A. G. Brighton, Sedgwick Museum, Cambridge (SM); Prof. K. E. Caster, University of Cincinnati (UC) ; Dr. D. B. Macurda, Jr., University of Michigan, Museum of Paleontology, Ann Arbor (UMMP); Dr. P. M. Kier and Mr. T. Phelan, United States Na- tional Museum, Washington (USNM) ; Dr. Katherine G. Nelson, University of Wisconsin, Milwaukee (UWM) ; and Dr. C. McClin- tock, Peabody Museum of Natural History, Yale (YPM). I am particularly grateful to Dr. Kier, Dr. Richardson, and Dr- Horowitz, for assistance with photography and to Prof. L. R. Laudon and Dr. F. D. Holland for information on cystoid localities. This work was commenced in 1967 at the University of Michigan, Museum of Paleontology under a National Science Foundation Fellowship in a program of systematic and evolutionary Biology (NSF GB-3366). Continued support came from an Indiana Uni- versity Foundation grant-in-aid. Both are gratefully acknowl- edged. Finally, I wish to thank the Director of Field Museum for publishing this study. TAXONOMIC HISTORY In 1864 (p. 7) James Hall proposed the genus Holocystites for six species of cystoids from the Niagaran Dolomite of Wisconsin, two of which he had previously described (Hall, 1861, p. 23) as Caryo- cystites cylindricus and C. alternatus. Lonsdale (1849, p. 296) had already proposed the name Holocystis for a coral genus from the Lower Greensand (Cretaceous) of England. Between the ap- pearance of the first 29 pages of Hall's paper, which were published privately in December, 1864, and the appearance of the complete work in 1868, Hall became aware of Lonsdale's name and suggested the alternative name Megacystites (1868, p. 380) should anyone object to the similarity between Holocystites Hall and Holocystis Lonsdale. On current interpretation of the ICZN rules both Holocystites and Holocystis are valid generic names. Therefore, Megacystites Hall, 1868 is an objective junior of synonym of Holo- cystites Hall, 1864. Hall did not select a type species for his genus. Angelin (1878, p. 29) introduced the contraction Megacystis when he described two Swedish species, "M. ovalis" (a rhombiferan cystoid) and "M. alternata Hall var." (described here as Brightoni- cystis sp., p. 147). F. A. Bather, who unilaterally changed generic names to suit his fancy, contracted Holocystites Hall, 1864 to ''Holocystis," regarded this as a junior synonyn of Holocystis Lonsdale, 1849 and therefore adopted Hall's alternative name in the form "Megacystis" (Bather, 1900. p. 47). Clearly Megacystis Angelin, 1878, Bather, 1900 is also an objective junior synonym of Holocystites Hall, 1864. Winchell and Marcy (1865), S. A. Miller (1878, 1879, 1880, 1882, 1889, 1891, 1892b), Miller and Faber (1892), Miller and Gurley (1894, 1895) and Foerste (1917) described new species of Holocystites (s.l.), raising the total to 51. Miller (1889, p. 253) selected Caryo- cystites cylindricus Hall, 1861 as type species of Holocystites and introduced the family Holocystitidae (1889, p. 215). Otto Jaekel (1899, p. 413) believed Hall's original species of Holocystites were aberrant crinoids. He therefore erected the genus Trematocystis, (type species Holocystites subglobosus Miller, 1889), for cystoids which have globular thecae and pore-structures with 4 FIELDIANA: GEOLOGY, VOLUME 24 omega-shaped tangential canals described by Miller from southern Indiana. Foerste (1917, pp. 234-235) suggested that most of Miller's species belonged to Trematocystis but later (1920b, p. 49) he ap- parently modified this view. Both Regn^ll (1945, p. 183) and Kesling (1963, p. 115) referred Holocystites and its relatives to the family Aristocystitidae. Regn^ll suggested that the most characteristic feature of Holocystites was the nature of the pore-structures which, in my opinion, is sufficiently distinct to warrant the recognition of a separate family, the Holo- cystitidae. Comparison of the oral areas of Holocystitidae, Sphaero- nitidae, and Aristocystitidae (fig. 1) shows that the former pair are more closely related to each other than either is to the Aristo- cystitidae, thus providing additional support for the separation of the Holocystitidae from the Aristocystitidae. All six species originally assigned to Holocystites by Hall were based on internal molds. All the material from the Osgood Forma- tion of southern Indiana used by Miller in his descriptions was well preserved and many specimens were complete. The characters of many of Miller's "species" can be ascertained easily but those of Hall's species cannot. Very few external molds have been collected from the Racine Dolomite and related formations which makes comparison of Racine and Osgood species very difficult. Regn^ll (1945, p. 183) suggested that the characteristic feature of Holo- cystites was the nature of the pore-structures, but this cannot be confirmed from Hall's original specimens since they are all internal molds. An attempt has been made to define the characters of the genus Holocystites on the basis of the original type material and then to recognize these characters in well-preserved specimens from the Osgood Formation. This has resulted in a restricted definition of Holocystites and a much clearer understanding of its characters. The type specimens of all but five described species have been examined; the remaining five types appear to have been lost. A reduction in the number of species recognized from 51 to 18 has resulted and two new European species are described. These are attributed to six genera: Holocystites Hall, Trematocystis Jaekel, Triamara Tillman, Pentacystis n.g., Pustulocystis n.g., and Bright- onicystis nov. Of these Triamara is assigned to the Aristocystitidae; the remaining five genera to the Holocystitidae. The holocystitid genera are distinguished on their pore-structures and the arrange- ment of plates in the oral area which, unlike that of the rest of the theca, is remarkably constant. GEOGRAPHIC AND STRATIGRAPHIC DISTRIBUTION All the cystoids described in this paper belong to two families, the Aristocystitidae and the Holocystitidae. Curiously all the known Silurian members of both families are found in North America and all North American specimens are exclusively Middle Silurian in age. Conversely all Eurasian and North African representatives of both families are from the Middle or Upper Ordovician. The Aristocystitidae achieved considerable diversification and wide geographic distribution during the Ordovician. Examples, be- longing to several genera, have been reported from China (Sun, 1936), Bohemia (Barrande, 1887), North Africa (Termier and Termier, 1950; Chauvel, 1966), and Germany (Sdzuy, 1955), and, from personal collecting, isolated plates undoubtedly occur in England. The American Silurian genus Triamara Tillman is the last representative of this once successful group. In contrast, the Holocystitidae is represented by just two species from two European Upper Ordovician localities, but flourished, albeit briefly, during the Middle Silurian in North America where the fauna contains four genera and 14 species. Despite this difference it is clear that both families arose in Europe, or possibly Asia in the case of the Aristocystitidae, and migrated to North America where for some unknown reason they were able to outlive their Eurasian relatives. The total absence of representatives of either family in the Lower Silurian may be due to the vagaries of preservation. However, Lower Silurian cystoids of any sort are extremely rare and this could reflect a genuine evolutionary crisis especially since several major groups of cystoids became extinct at the end of the Ordovician . Comparison of the oral areas of typical representatives shows clearly that the Holocystitidae is more closely related to the Sphaeronitidae than to the Aristocystitidae (fig. 1). In both the Sphaeronitidae and Holocystitidae the oval or circular mouth is surrounded by a quadrilateral or pentagonal peristome which is covered by six relatively large immovable plates. There are always four or five ambulacra and the food grooves pass beneath the cov- ering plates at the edge of the peristome, a fact first noted by Prokop (1964, p. 22) in the genus Eucystis Angelin. In specimens Fig. 1. Diagrammatic representations of the oral areas of Sphaeronitidae (A-C), Holocystitidae (D-F), and Aristocystitidae (G-I) showing mutual re- lationships. A, Sphaeronites n. sp., RM Ec25944; B, Eucystis angelini Regne II, author's coll.; C, Haplosphaeronis oblonga (Ang.), author's coll.; D, Holocystites cylindricus (Hall), USNM S3070; E, Trematocystis globosus (Miller), USNM S3063a; F, PentacysHs simplex n. sp., AMNH 20271b; G, Aristocystites bohemicus Barrande, USNM; H, SinocysHs loczyi Reed, SM A3208; I, Triamara ventricosa (Miller), FMNH Pel0954. Abbreviations: a — anus, f — ambulacral facet, g — gonopore, h — hydropore, m — mouth, p — oral pores, 1-6 palatal plates covering mouth. Outlines of peri- stome and ambulacral facets, and of periproct, shown by dashed line. Figures not to scale. (Continued on next page) 6 PAUL: NORTH AMERICAN HOLOCYSTITES FAUNA 7 with the covering plates missing the food grooves can be seen to continue from the edge of the peristome to the edge of the mouth (figs, 56d, 59c, 62j). The covering plates form a true tegmen, i.e., a covering to both the food grooves and the mouth, but since this tegmen differs markedly from typical crinoid tegmina or even the tegmen of the cystoid Caryocrinites Say, I propose to call this structure the palate. The plates forming the palate have previously been called "orals." However, there is no homology at all between the orals of the Glyptocystitida, which surround the peristome, and the "orals" of the Sphaeronitida, which cover the peristome, hence the new term palatals is proposed for these latter plates. The oral area of aristocystitids (fig. IG-I) is characterized by two broad (up to 7 mm. wide) food grooves which meet at the small oval mouth. Neither, one, or both of these food grooves may divide once to give two, three, or four ambulacra. The food grooves are covered by many small covering plates which in Triamara are ar- ranged in four series. Triamara differs from Aristocystites Barrande in having three, not two, ambulacra and in having a sieve-like hydropore instead of a slit. Jaekel (1899) reported rare examples of Aristocystites with three ambulacra so it is possible that a fore- runner of Triamara appeared in Europe during the Ordovician. In summary, the evidence of geographic and stratigraphic dis- tribution shows clearly that the Aristocystitidae arose and flourished in Eurasia during the Ordovician and migrated to North America where a single genus, Triamara occurs in the Middle Silurian. The Holocystitidae, on the other hand, evolved from the Sphaeronitidae late in the Ordovician in Western Europe and migrated to America where they flourished briefly during the Middle Silurian. A more detailed discussion of the distribution of Middle Silurian cystoids in North America is presented in the section on Paleoecology. Note in the Sphaeronitidae (A-C) and Holocystitidae (D-F) the large mouth is set within a pentagonal (B-D) or quadrilateral (E) peristome which is cov- ered by six large palatal plates in life. The food grooves are very narrow and end in small ambulacral facets in the Sphaeronitidae but in large ambulacral facets in the Holocystitidae. The Aristocystitidae (G-I) have a small mouth set within very broad ambulacral grooves which are lined with oral pores and end in large ambulacral facets. PRESERVATION Two distinct modes of preservation predominate. These are: 1) preservation of the original test, often with minute detail showing on suitably weathered specimens; 2) preservation as molds in dolomite, sometimes accompanied by partial replacement of the test. In the latter case internal molds far outnumber external molds in museum collections. These distinct modes of preservation cause difficulties in taxonomy. The first six species described, in- cluding the type species of Holocystites, were all based on internal molds. From such molds it is possible to deduce the plate arrange- ment, the positions of the fundamental orifices, the presence, but not the type, of pore-structures, and the overall thecal shape. These features are adequate to characterize the genus Holocystites when compared with complete thecae but are inadequate at specific level. It has proved necessary to restrict some of the earliest names to their type specimens on the grounds that the species are unrecog- nizable. A further difficulty has arisen in that generally the speci- mens described first by S. A. Miller were incomplete thecae whereas complete specimens found later illustrate the entire morphology of the species (or genus). Strict application of the law of priority would result in some of Miller's acceptable species being based upon inadequate type specimens. In an attempt to lessen future confu- sion all accepted specific names are based on clearly identifiable type specimens: where types are lost or have the oral surface missing, the names have been restricted to the type specimens. The European Ordovician species have the original test pre- served and so the nature of the pore-structures can be confirmed. Suitably weathered specimens from the Osgood of Southern Indiana show microscopic details of the plates and add considerably to knowledge of cystoid plate structure. Echinoderm plates consist of a meshwork of calcite rods and spaces which are filled with soft tissue in life. Another echinoderm characteristic is that each plate acts as a single optical unit in polarized light and has generally been considered to be a single calcite crystal. In both these respects cystoid plates were no different from those of other echinoderms. 8 PAUL: NORTH AMERICAN HOLOCYSTITES FAUNA 9 In the process of fossilization the soft tissue rots and the spaces between the organically secreted calcite are usually filled with secondary calcite which grows in optical continuity with the primary organic calcite. Thus plates of most fossil echinoderms are solid calcite crystals and exhibit excellent cleavage. The original struc- ture of the organic calcite can be detected in suitably preserved fossils. Some thin-sections show ghosts of the original structure. X-ray fluorescence and differential weathering sometimes reveal the original structure but the structure is best shown when some other mineral, generally pyrite, fills the soft tissue spaces ab initio. Examples of the latter preservation in specimens of Archegocystis Jaekel from Shole'shook, South Wales have provided the best available information on cystoid plate structure. Details revealed in these specimens can be confirmed in suitably weathered specimens from southern Indiana. A full description of the Shole'shook material will be published separately but a brief account is necessary to interpret evidence on plate structure and growth in the Holo- cystites fauna. Two specimens of Archegocystis from South Wales (SM A 57521, A57522) had the original soft tissue spaces filled with pyrite soon after death. Subsequent weathering has turned the pyrite into a rusty brown limonitic deposite and in places has leached away the calcite completely. Where the calcite remains, it is possible to distinguish original soft tissue spaces from organically secreted calcite. Where the calcite has completely dissolved, the limonitic deposit faithfully preserves the structure of the soft tissue. Cross- sections through plates (fig. 2e-f) show an inner region of coarser mesh and an outer region of finer mesh. By a fortunate accident one specimen (A57521) parted along the junction of the outer and inner regions thus revealing tangential surfaces of the layers in several plates around the mouth (fig. 2c-d). The inner region con- sists of sheets of calcite with small soft tissue connections passing through them alternating with sheets of soft tissue with small calcite rods passing through (fig. 3b) . This inner region thins toward the margins of the plates and the sheets of soft tissue and calcite increase in area towards the internal surface of the plate (fig. 4g). The outer region (figs. 2a, 3a) is composed of a network of finer calcite rods and soft tissue fibres of which those perpendicular to the plate sutures are by far the most obvious. At first sight the entire structure appears to consist of rods perpendicular to the plate sutures. The outer region thickens towards the sutures and ac- counts for the entire thickness of the plate at the suture. 10 PAUL: NORTH AMERICAN HOLOCYSTITES FAUNA 11 The structure of the plates is further modified at the junction of the inner and outer regions, towards the external surface of the theca and by the thecal canals of the pore-structures which pass right through both regions (fig. 2f). The sheet of soft tissue which lies between the inner and outer regions is similar to the sheets of the inner region but has many finer calcite rods passing through it (fig. 2b). It shows growth lines very well as does the inner surface of the outer region. If the structure of the outer region persisted unmodified to the external surface, it would have been reported before. Echinoids, among recent echinoderms, display marked modification of the skeletal network at the external surface dependant upon function. The spine bosses are almost solid calcite whereas the general test surface is highly porous and the areas for muscle attachment around the spine mamela are extremely porous and have larger pore spaces than the general test surface. Nevertheless, the outer region of rods and soft tissues fibres perpendicular to sutures is present in the plates of recent sea urchins and can be seen on abraded tests. The external surface of the test in Archegocystis was similarly modified probably for functional reasons. To summarize, the thecal plates of Archegocystis consist of two distinct regions: an inner region of alternating sheets of soft tissue and calcite, and an outer region largely made of soft tissue fibres and calcite rods set perpendicular to the plate sutures. This struc- ture is modified toward the outer surface of the test. Barrande (1887, p. 30) described three layers to the cystoid theca but he maintained the pore-structures were confined to the middle layer and sealed by the inner and outer layers. Barrande's three layers, which were described in error, do not correspond to the features described here, all of which would correspond to his middle layer since the pore-structures in Archegocystis pass right through the test. Fig. 2 (opposite). Structure of thecal plates in Archegocystis spp. 2A-D, SM A57521. a, Detail of outer region of one plate to show mesh work of fine soft- tissue strands (replaced by limonite, showing white in photograph) and calcite rods (leached away, showing black) set perpendicular to plate sutures. Note irreg- ular meshwork between casts of perpendicular canals of diplopores. b, Detail of median sheet of soft tissue which is pierced by many fine calcite rods (black). Both Xl2. c. General view of outer region of plates around the mouth. X2. d, General view of cystoid to show inner region of split plates around the mouth. X2. e-f. SM A57522. Cross-sections through two plates so show inner (lower) region of coarser meshwork in sheet-like structure (2e) and outer (upper) region of finer mesh. The white pillar to the right of Figure 2f is a cast of a perpendicular canal of a diplopore which clearly passes through both regions of the plate. 2e-f X14. Fig. 3. Plate structure in Archegocystis. 3a-b SM A57521. a. Detail of outer region of two plates. Note plate suture to left of center. X12. b, Detail of inner region of another plate showitig two separate soft tissue sheets with much coarser mesh . X 1 2 . 12 PAUL: NORTH AMERICAN HOLOCYSTITES FAUNA 13 Suitably weathered specimens of Trematocystis, Triamara, and Pentacystis from the Osgood of southeastern Indiana also reveal evidence of this test structure. Most specimens add to knowledge of surface modifications but some abraded specimens show the deeper layers. Triamara bears diplopores (fig. 6a) and AMNH 20267 reveals that the meshwork in the peripores is consistantly coarser than that between the diplopores. In addition, a thin line of clear calcite runs around the peripore probably marking the limit of the soft tissue papula that extended from the diplopore. Very similar clear calcite lines can be seen around the pore-pairs of recent echinoids. In Trematocystis which bears humatipores (fig. 6b) the meshwork over the tangential canals appears to be finer than that between them, which in turn is finer than that flooring the tangential canals. Some of the dolomitized specimens of Triamara and Holocystites show growth lines and have tumid plates (fig. 19b). Generally growth lines are absent on the internal and external surfaces of the plates but show clearly at the junction of the inner and outer regions of the plates. Close examination of these specimens shows that the plates are partially dolomitized and for some unexplained reason all or part of the inner region of each plate has been pre- served (by replacement) but none of the outer region. A further curious feature is that in some specimens where only part of the inner region of each plate has been replaced, topologically equivalent parts are replaced in most plates. GROWTH In other groups of cystoids it is convenient to discuss growth and morphology under three headings: stem, theca, and ambulacral or subvective system. None of the Holocystites fauna has a stem, attachment was direct, and no specimen has been found with the subvective appendanges (?brachioles) preserved. Hence, at present, only growth of the theca can be discussed. Although no fossil evidence exists it is assumed that cystoid larvae underwent metamorphosis as all living echinoderms do. Post-metamorphic growth of cystoid thecae was achieved by two distinct methods. The Superfamilies Glyptocystitida and Hemi- cosmitida have thecae with a fixed number of plates; growth of the theca was entirely by enlargement of these plates. In most other cystoids, including those in the Holocystites fauna, the number of thecal plates was not fixed and in many new plates were added as growth proceeded. Before discussing the manner in which these later plates were added, it is appropriate to consider how the plates themselves grew. a. Plate growth As stated previously, cystoid plates consist of two regions or layers: an inner region of alternating sheets of calcite and soft tissue, and an outer region of calcite rods and soft tissue fibres principally set perpendicular to plate sutures. The inner region thins toward the plate sutures and in Archegocystis the outer region accounts for the entire thickness of the plates at the suture (fig. 4a-c). Growth of the plate was achieved by lengthening of the outer region rods and fibres at the suture. New calcite rods and soft tissue fibres were added as the sutures lengthened and the plates thickened. If the outer region were the only layer, plates would develop which were very thin in the center and became progressively thicker toward the suture (fig. 4f). The inner region grows by the addition of sheets of calcite and soft tissue on the internal surface of the plate. Each successive sheet covers a greater area of the plate than its predecessor (fig. 4g). Thus the presence of an inner region allowed the plates to maintain a constant thick- 14 PAUL: NORTH AMERICAN HOLOCYSTITES FAUNA 15 A B C Fig. 4. Diagram to illustrate growth and structure of cystoid thecal plates. Sequence A-B-C illustrates growth as shown in diagrammatic cross-section. Growth at the suture is accomplished by the outer region (e) while sheets of the inner region (i) are added to maintain thickness. Sequence D-E-F illustrates growth of the outer region only and demonstrates the weakness of such a structure. G, Diagrammatic perspective view of a single plate with a section of the outer region (e) removed to show sheet structure of inner region (i). In reality the sheets are pierced by soft tissue strands and connected by calcite pillars neither of which is shown. ness as they grew. Indeed it is possible that as thecal growth slowed the plates continued to thicken by the addition of sheets of calcite and soft tissue to the inner region. As yet the evidence for this is inconclusive. b. Thecal growth No species is represented by an adequate growth series and most specimens are larger than 25 mm. in major dimension. The following statements about the growth of the theca are deduced from examination of presumably mature specimens. None of the 16 FIELDIANA: GEOLOGY, VOLUME 24 B V ^^ ^^ ^ C A i B "-* C 0 — > E Fig. 5. Diagram to illustrate thecal growth by the introduction of secondary and tertiary plates. In sequence A-B-C maximum growth occurs in an ad-aboral (vertical) direction which results in circlets of plates of distinct generations and is typical of cylindrical thecae. In sequence A-D-E growth is almost equal in all directions which results in complete separation of the primary plates and less dis- tinct generations of plates. This sequence is typical of spherical thecae. The small quadrilateral figures represent the area bounded by four homologous points in each plate arrangement, starting with square A. Primary plates white, second- aries light stipple, tertiaries heavy stipple. species which comprise the Holocystites fauna has a fixed number of thecal plates and in many species extra plates were added during growth. Even when new plates were added the original plates con- tinued to enlarge and there is no characteristic maximum plate size for any species as far as is known. Neither is there a char- acteristic maximum thecal size for any species: growth apparently proceeded until death. In discussing the addition of new plates it is important to con- sider when and where they were developed. New plates may develop throughout growth, in which case the complete range of plate sizes from largest to smallest will be present assuming the growth rate is constant for all plates. Alternatively, new plates may be added at definite intervals in the growth of the theca and, ideally, plate sizes will be clustered about certain norms. In this latter case distinct generations of plates can be recognized. These generations of plates are more easily recognized if new plates are PAUL: NORTH AMERICAN HOLOCYSTITES FAUNA 17 added systematically around the theca and not randomly (figs. 5c, e). The positions in which new plates develop depend largely on thecal shape which in turn depends upon growth gradients. The genus Holocystites illustrates this well (fig. 5). Holocystites cylindricus (Hall), the type species, has the simplest plate arrange- ment. In the main body of the cylindrical theca the plates are hexagonal and arranged in circlets, each of which contains eight plates. Since the theca is cylindrical, growth is greater parallel to the axis of the cylinder than radially. H. abnormis Hall is very similar to H. cylindricus but has a few additional plates low down on the theca. The new plates form a distinct second generation and are added between primary circlets and not between plates in a primary circlet (fig. 5b). This has the effect of lengthening the theca without increasing the diameter, thus maintaining the cylin- drical shape. H. alternatus (Hall) carries the process much further and large examples have three distinct plate generations. First, plates of the second generation are added between the primary circlets (fig. 5b) until complete secondary circlets are formed. Then plates of the third generation are added between the primary and secondary circlets until complete tertiary circlets develop (fig. 5c). This still has the effect of lengthening the theca without markedly in- creasing the diameter, hence maintaining a cylindrical theca. In the case of a cylindrical theca, such as that of H. alternatus, length- ening is achieved largely by the addition of new plates while the diameter increases entirely by the enlargement of the original plates. H. scutellatus Hall has a pyriform to globular theca the diameter of which increased considerably during growth. At least two generations of plates occur but the secondary plates do not form circlets and are intercalated between primary plates of any one circlet as well as between primary circlets (fig. 5d-e). Thus primary plates which originally formed a closed circlet become separated during growth. This, coupled with the fact that plate growth is greatest near the widest part of the theca and least near the narrowest part, makes recognition of distinct plate generations difficult. The following generalizations about thecal growth in the other five genera are offered; details will be discussed under individual species. Trematocystis: Plates of one generation arranged in circlets which frequently have eight plates but may have from seven to sixteen. Thecal shape pyriform to globular. 18 FIELDIANA: GEOLOGY, VOLUME 24 Pustulocystis: Plates of one generation arranged in circlets of eight to seventeen plates. Theca pyriform to oval. Pentacystis: Plates of one or more generations, primary circlets of variable number of plates, intercalation of new plates irregular. Theca pyriform or globular, frequently large (greater than 50 mm. diameter). Brightonicystis: Plates of more than one generation, primary circlets with many plates, intercalation of new plates apparently random. Theca globular or cylindrical. Triamara: Plates of at least two generations, primary plates apparently not always arranged in distinct circlets, intercalation of new plates random. Theca globular, pyriform or elongate pyriform, large (50-150 mm.) The taxonomic implications of thecal growth are considerable. It is generally unwise to chose as a specific character one which changes radically during growth since immature specimens will differ markedly from mature ones. Some species do have a fairly constant arrangement of plates (e.g., Holocystites cylindricus) but generally both the number of plates per circlet and the total num- ber of circlets varies in all species. In those species which added plates during growth both the number and arrangement of plates changes as growth proceeds. Indeed as recently as 1967 Tillman was mislead into describing a new species, Triamara cutleri, at least in part because his specimen had a much smaller number of thecal plates than the obviously closely related and much larger "Holo- cystites" gyrinus Miller and Gurley (Tillman, 1967, p. 223). Clearly the arrangement of thecal plates is a poor character to use in de- fining species of the Holocystites fauna however useful it may be in other groups of cystoids. Nor is it at all surprising that S. A. Miller was able to recognize 40 so-called species using this character. c. Growth of fore structures Before considering the growth of pore-structures it is necessary to describe the morphology of the two types present in the Holo- cystites fauna. Triamara possesses true diplopores (fig. 6a) whereas the Holocystitidae are characterized by humatipores (fig. 6b). In diplopores two canals pass more or less perpendicularly through a thecal plate and open on the external surface of the plate in a shallow depression called a peripore. A soft tissue papula arose from the peripore much as the tube-feet of sea urchins arise from the pore- pairs. In life the entire structure formed a single thecal canal Fig. 6. Diagrammatic representations of the structure of diplopores (A) and humatipores (B). Both types of pore structure consist of a pair of perpen- dicular canals (PC) which open internally in pores (P). In humatipores (B) the perpendicular canals are connected by several fully calcified tangential canals (TC) whereas in diplopores the perpendicular canals are connected by a soft tissue papula (Pa). The soft tissue papula is never preserved in fossils so that the per- pendicular canals of diplopores open externally in two pores (P) set within a depression called a peripore (Pp). In both A and B the internal surface is toward the bottom of the figure. 19 20 FIELDIANA: GEOLOGY, VOLUME 24 which connected two internal pores and the middle portion of which formed a papula. The whole structure formed a single functional unit: internal fluid entered one pore, passed through the plate into Fig, 7. Diagrammatic representation of the canal system in the pustular humatipores of Holocystites s.s. One perpendicular canal (PC) opens into several tangential canals (TC) subcentrally in the tubercle while the other is peripheral to the structure. P — pore. (cf. fig. 30b.) the papula where respiratory exchange took place, and returned to the thecal cavity, exiting via the other pore. The evidence for this interpretation is detailed in "Functional Morphology." Humatipores (defined in Paul, 1968) are entirely within a thecal plate and have no soft tissue papula (fig. 6b). Again two tubes pass perpendicularly through the thecal plate from two internal pores and almost reach the external surface. They are connected by two to ten or more canals which are tangential to, and which lie just beneath, the external surface of the plate. Again the whole structure forms a single thecal canal: internal fluids pass up one of the tubes, through the series of tangential canals where exchange probably took place and down the other tube back into the thecal cavity. The tubes perpendicular to the plate surface are called perpendicular canals (PC) and those tangential to the surface, tangential canals (TC). All thecal canals consist of two perpendic- ular canals connected by one or more tangential canals. In diplo- pores the tangential canal is a papula and is never preserved in fossils. Diplopores are simple thecal canals since there was ap- parently a single tangential canal (the papula) connecting the two perpendicular canals. Humatipores are compound thecal canals PAUL: NORTH AMERICAN HOLOCYSTITES FAUNA 21 as there is more than one tangential canal connecting the pair of perpendicular canals. The structure of both diplopores and humati- pores is basically the same. Modifications of the peripore allow recognition of different types of diplopore. However, only one type, which has simple but rather irregular peripores without rims, is found in Triamara. Two types of humatipore occur in the Holocystitidae. Trernatocystis, Penta- cystis, and Glohulocystis have smooth or finely granular plates; the humatipores are truly buried beneath the plate surface. Some species of Pustulocystis and Holocystites have plates covered with small pustules, each of which corresponds to a single humatipore (fig. 7, 30a-b, 57c). The tangential canals are developed beneath the surface of the pustules in this type. Thecal pores and canals may develop either by a primary lack of secretion of calcite at the plate suture as the plates grow, in which case the canals tend to be aligned perpendicular to plate sutures, or by a secondary resorbtion of calcite. The latter seems to have been the case with both diplopores and humatipores which are commonly randomly arranged with respect to plate sutures. Both humatipores and diplopores reached a maximum size very soon after they started development and as thecal plates grew new pore-structures developed, the existing ones did not grow larger. From the mode of growth of thecal plates it follows that thecal canals could only originate at the sutures if resorbtion did not take place. If this were so, partially developed thecal canals would be found along plate sutures in immature thecae. This has never been reported in the literature and I have never observed it. In Sphaeronites Hisinger the diplopores are independent of plate sutures. They are densely and evenly developed over the entire thecal surface and hence most sutures have some diplopores devel- oped across them. If no resorbtion took place, diplopores which cross plate sutures would have become elongate as the plates grew. Again there is no evidence of this. The best evidence for resorbtion by diplopores is found in the exceptionally preserved specimens of Archegocystis. The perpen- dicular canals were filled with sediment after death and can be de- tected in the plate structure. In all cases growth lines are truncated by the canals; they never pass around the canals (fig. 2b). The fine calcite rods and soft tissue fibres of the outer region of the plates are truncated by the canals; they also never pass around the canals (figs. 2a, 3a). The area between any pair of perpendicular 22 FIELDIANA: GEOLOGY, VOLUME 24 canals is filled with a random network of fine rods and fibres; not with the regular arrangement of rods and fibres perpendicular to plate sutures. The canals are randomly placed with respect to plate sutures and in one case a canal passes up the suture between Fig. 8. Sketch of the meshwork structure of the outer region of the test in Archegocystis where a perpendicular canal (PC) passes up through a plate suture (S). To prevent a slit forming as the plates grew, meshwork has been added as curved rods and fibres on either side of the perpendicular canals thus filling the slit. SM A5752L two plates. As the plates grew a slit would have formed unless the areas adjacent to this canal were filled in with more plate material. Little curved fibres are seen on either side of the canal in both plates indicating where and how this infilling took place (fig. 8). Several other canals have similar curved fibres on one side which suggests they adjusted position as the plates grew. A most interesting feature is evidence of the former existence of diplopores in addition to those present. Typical diplopores show up as an oval area of randomly arranged fibres with a perpendicular canal at either end of the oval. There are several oval areas, identical in size, shape, and arrangement of fibres but which entirely lack canals (figs. 2a-b, 3a). From the truncation of growth lines and the calcite meshwork it appears that diplopores arose by resorbtion and in some cases the diplopores degenerated leaving an area which was filled with a random network of rods and fibres. Presumably the number and position of diplopores changed during the lifetime of an Archegocystis. All available evidence suggests that humatipores were no different from diplopores in mode of growth. No humatipore has been ob- served across a suture although rarely the tangential canals of humatipores may infringe upon a suture. In Holocystites humati- pores nearest the sutures are frequently smaller than those near the center of a plate. This suggests that as the plate grew space became available within which a new humatipore could develop. PAUL: NORTH AMERICAN HOLOCYSTITES FAUNA 23 d. Enlargement of thecal orifices The thecal orifices (mouth, anus, gonopore, and hydropore) may enlarge either by resorbtion or by the differential growth of plates around them. The former must have been the case with the gonopore which, in the Holocystites fauna, typically opens entirely within a single plate (fig. Id-f). The mouth and anus are always surrounded by a number of plates. Enlargement of both these orifices was almost certainly achieved by growth of the plates around them without any resorbtion. The hydropore is developed as a slit across a plate suture in Holocystites, Pustulocystis, and Trematocystis (see fig. Id-e). Its slit-like form results from a primary lack a secretion of calcite in both plates. In Triamara (fig. II), the hydropore is a sieve-like structure frequently developed within a single plate. Here enlargement must have been by resorb- tion to form additional fine canals as the sieve grew. In Pentacystis no clear hydropore can be seen. However, Bather (1919c, p. 323, fig. 31) pointed out that occasionally the suture where one might expect to find a hydropore (P01:P06, see p. 71 below) is very irregular and may have functioned as a hydropore. If so, enlarge- ment would most probably have been by primary lack of secretion. FUNCTIONAL MORPHOLOGY Man is a purposive animal and tends to look for function or purpose in other things. We may design a new machine part specifically to improve the machine's functional efficiency but organic structures arise by chance and none ever evolved to serve a purpose. A new organic structure will have certain more or less well defined physical effects. Such effects may be beneficial, neutral, or positively lethal to the organism bearing the structure. Those effects which are beneficial constitute what we call the function (or functions) of the structure. Few, if any, organic structures have a single effect (function) and improvement in the beneficiality of one effect (i.e., increase in functional efficiency) is generally achieved at the expense of other effects. Thus most structures are compro- mises between the maximum benefit (maximum functional efficiency) of several effects (functions). Consider one basic function of the cystoid theca, for example. The cystoid theca is inflexible and its most obvious effect is to provide a rigid housing for the bulk of the soft anatomy. This protects the soft anatomy from mechanical damage and predators, which is surely beneficial, but at the same time it restricts access to the ambient sea water which provides the food and oxygen for life. Specialized entrances must be made for the ingestion of food and uptake of water into the water-vascular system, and exits for genital and waste products. These orifices must not defeat the "function" of the theca by structurally weakening it or by allowing predators access to the thecal cavity. The problem of respiration is overcome in cystoids by the production of pore-structures in the thecal wall. The thicker the thecal wall, the stronger it will be, but the thinner it is, the greater the amount of respiratory exchange that can take place. Pore-structures are a compromise between the structural requirements of strength (protection) and exchange (respiration). Probably all organic structures are compromises between several ideals. However, nature is not concerned with perfection, merely with the minimum efficiency necessary for survival. 24 PAUL: NORTH AMERICAN HOLOCYSTITES FAUNA 25 With this in mind, I shall discuss functional morphology under the headings: theca, orifices, and pore-structures, considering gen- eralities first and then the detailed morphology of members of the Holocystites fauna, a. The Theca The rigid theca provides protection and support for the internal organs: protection because it completely encloses these organs and support because it is rigid and the organs may be anchored to it by mesenteries. The theca is not a single structure but composed of discrete plates which grow as the theca grows. By analogy with recent echinoids the sutures between the plates of the cystoid theca are potential lines of weakness. An arrangement of plates that minimizes the weakness of the sutures and allows growth with the minimum of wasted energy in the form of calcite resorbtion is ideal. The arrangement of thecal plates in most cystoids conforms to the ideal case fairly well. One important factor controlling the weakness introduced by plate sutures is the maximum length of suture along an arc of a great or small circle (in the case of an ideal spherical theca). The total length of all sutures is minimum with the smallest number of plates but growth without resorbtion is impossible and the sutures between plates extend through large arcs. For example, a theca composed of one spherical plate has no sutures, which is ideal from the point of view of strength, but it can only grow by secretion on the external surface and resorbtion on the internal surface. Thecae with two, three, or four equal plates have sutures which extend through 360, 180, and 120° of arc respectively (fig. 9) . Such large plates cannot grow and maintain the same thecal shape without resorbtion due to changes in radius of curvature during growth. In a large number of plates the individual plate sutures are short and the effect of changes in radius of curvature during growth is minimal but the total length of all plate sutures is very large. A plate arrangement such that sutures of adjacent plates are never continuous and aligned is strongest since no sutures extend through more than a few degrees of arc. Maximum strength is achieved when no more nor less than three sutures meet at any point on the theca. This in turn controls the shapes of thecal plates. For efficient protection of internal organs the entire thecal surface must be covered without any gaps. The plates must be polygonal and fit together exactly. If all, or almost all, the plates are the same shape, then hexagonal, and to a lesser extent pentagonal. 26 FIELDIANA: GEOLOGY, VOLUME 24 Fig. 9. Diagrammatic representations of spheres bounded by 2 (A), 3 (B), and 4 (C) plates to show very extensive sutures which weaken such structures. plates are ideal. Plates with more than six sides cannot be fitted together without gaps unless additional small plates with three, four, or five sides are intercalated between them (fig. 5). A regular arrangement of hexagonal plates has an additional advantage in that the centers of any three adjacent plates lie at the corners of a triangle. The very common echinodern "ornament" of ridges con- necting plate centers forms a geodesic system which increases both strength and rigidity. The general case may be stated as follows: an arrangement of thecal plates with short sutures no more nor less than three of which meet at any point on the theca is ideal from the points of view of structural strength and simplicity of growth. Members of the Holocystites fauna conform to the general case stated above. Primary plates are initially hexagonal except where they border thecal orifices or form the base of attachment and no more than three sutures meet at any point on the theca. A feature peculiar to holocystitids is the frequent occurrence of circlets of eight primary plates. No special functional significance is attached to this beyond the general statement that the more plates there are in a circlet the more easily changes in radius of curvature during growth can be accomodated. There are, however, many specimens with circlets of more and less than eight plates. PAUL: NORTH AMERICAN HOLOCYSTITES FAUNA 27 Those species which add plates during growth also conform to the general case in plate arrangement. Small, initially triangular, secondary plates are intercalated between primary plates in such a way that no more than three sutures meet at a point. As the secondaries enlarge they become pentagonal, thus preserving the arrangement with no more than three sutures meeting at any point (fig. 5). The insertion of tertiary plates is essentially similar to that of secondaries. All plates are small and have relatively short sutures which extend through very small arcs. b. Thecal orifices The theca restricts access to the ambient sea water and special orifices are necessary for the continuation of vital activities. Gen- erally a maximum of five orifices is present in cystoids. These are: mouth, anus, gonopore, hydropore, and an aboral opening which connects the lumen of the stem to the thecal cavity. The latter orifice is not found in the Holocystites fauna, all members of which lack a true stem. There has been considerable disagreement not only as to what the orifices represent but more particularly as to which orifice is which. At present a consensus of opinion exists but as far as I am aware it was not until 1967 that any arguments were advanced to substantiate currently accepted interpretations (see Paul, 1967a, 1967b). The fact that there is a consensus of opinion as to the interpretation of cystoid thecal orifices is evidence that most authors recognize the same arguments to support their interpretations. Surprisingly no author before 1967 clearly stated these arguments which include comparison with homologous orifices in living echino- derms and a test of the mechanical fitness of the structures to perform their supposed functions. The arguments are as follows : One orifice is situated at the center of the ambulacral system. By analogy with living crinoids, the ambulacral system functioned for food gathering and the grooves leading to this orifice are inter- preted as food grooves. Clearly this orifice is the mouth and it functioned for the ingestion of food. Two of the remaining orifices are frequently covered by small pyramids of triangular plates. There is no evidence of any external musculature to open these pyramids but, in some cases, there are distinct muscle pits on the internal surfaces of the triangular plates. Such pyramidal structures are ideally contructed as one-way valves which allow egress from the theca and both orifices were therefore exits. By analogy with living sea urchins, the large orifice, which 28 FIELDIANA: GEOLOGY, VOLUME 24 in some cystoids is set within a plated flexible integument (peri- proctal membrane), is the anus and the smaller is the gonopore; both of which were functionally exits. The remaining orifice has been called the hydropore. In many cases this orifice is a narrow slit while in other specimens it is a sieve- like structure which strongly resembles the madreporite of recent starfish and sea urchins. Either slits or sieves are ideally suited to prevent passage of particles through an orifice. Since the water vascular system must take up water through the hydropore, the latter is functionally an entrance. The fourth orifice in cystoids is admirably constructed to prevent entry into the water vascular system of harmful particles and is most probably the hydropore. There remains a possibility that one or more of these orifices in cystoids may have been connected to internal organs not found in living echinoderms and whose function is unknown. The fact that all four can be interpreted in terms of known orifices in living echinoderms argues strongly against such a possibility. S. A. Miller (1878 et seq.), following Billings (1870), had most unorthodox views on the interpretation of cystoid thecal orifices. To facilitate comparison of Miller's descriptions with those given here, the following table is presented: S. A. Miller Current interpretation Ambulacral orifice Mouth, peristome Mouth Anus, periproct Anal aperture Gonopore in most cases but also hydro- pore when Miller spotted the latter. See Miller, 1892, p. 624 description of "H." commodus "supposed anal openings two." In holocystitids the mouth is oval or circular and about 5 mm. in diameter. It is set within a four or five-sided peristome which has a narrow border and carries four or five narrow food grooves from the ambulacral facets^ to the edge of the mouth (figs. Id, 24). In all species bar two of Pentacystis the peristome is composed of peri-oral plates and the peristome border is pitted with numerous "oral pores." The function of the "oral papulae" which arose from these pores is unknown but was presumably sensory. While taste is the most obvious suggestion the oral pores are not specially common near food grooves. * Since there is no evidence that the ambulacral appendages were in fact brachioles, I prefer the less specific term "ambulacral facets" to the more common term "brachiole facets." PAUL: NORTH AMERICAN HOLOCYSTITES FAUNA 29 The entire peristome was covered by six plates which form a rigid roof, here termed the "palate." The complete set of palatals has been observed in one specimen of Trematocystis and two of them in one specimen of Holocystites s.s. The palate is effectively a con- tinuation of the general thecal skeleton and has a similar protective effect. Despite the large relative size of the mouth the openings to the exterior for the food grooves are very small. The palate main- tains a complete protective covering to the soft anatomy. In Triamara (fig. li) the mouth is oval, 3-4 mm. across and set entirely within the wide ambulacral grooves. The left groove divides once to give three broad ambulacral grooves, each of which ends in two ambulacral facets. Central in an ambulacral groove is a narrow food groove on either side of which lies a single row of ambulacral pores. Like the oral pores of holocystitids these ambu- lacral pores open beneath covering plates and were presumably sensory in function. The covering plates are arranged in four series and were immovable. Again the only openings to the exterior are small and lie at the base of the ambulacral appendages. The anus in Triamara and holocystitids is rounded and 5-9 mm. in diameter. As far as is known it was always covered with a simple anal pyramid only. There were no auxiliary plates nor a plated periproctal membrane. Anal plates have only been seen in one specimen each of Trematocystis and Holocystites. The gonopore is a circular pore, usually 0.5 mm. in diameter. In the Holocystites fauna no gonopore has been seen with a covering pyramid, although one may have existed in life. The gonopore may be set in a raised tubercle or open flush with the thecal surface. It may or may not be developed across a plate suture. The external form of the hydropore varies. It may be slit-like or sieve-like. In the former case it is not known whether the slit passes right through the thecal plates or whether it is a narrow groove with porous sides and base. Sieve-like hydropores may have two or three separate clusters of fine pores. Slit-like hydropores are invariably developed across a plate suture; P01:P06 (see p. 71) in holocystitids (fig. Id). A surprisingly large number of well preserved specimens of Trematocystis either lack any evidence of the gonopore or of the hydropore. It is conceivable that asexual cystoids could survive and be preserved so absence of a gonopore in life seems possible. More puzzling is the absence of a hydropore. Surely no cystoid could survive without a hydropore? Some other connection between 30 FIELDIANA: GEOLOGY, VOLUME 24 the water vascular system and the sea water may have existed in some members of the Holocystites fauna. Bather (1919c) suggested that the irregular suture P01:P06 could represent a functional hydropore in Pentacystis which characteristically lacks a hydropore. Jaekel (1899) and others have suggested that where a distinct hydropore is absent the gonopore (Jaekel's parietalporus) may have functioned as a combined gonopore and hydropore. Although not impossible, this seems very unlikely since the gonopore is funda- mentally an exit and the hydropore is basically an entrance. More likely is the possibility that a "hidden hydropore" was developed in soft tissue somewhere within the peristome in cystoids which lack evidence of an external hydropore. It might be pointed out, however, that slit-like hydropores are sometimes exceptionally diffi- cult to detect. c. Pore-structures Almost certainly all thecal and calycinal pore-structures in primitive echinoderms were respiratory and facilitated gaseous ex- change between the body fluids and sea water. In such exchange, only the gases are exchanged; an exchange surface prevents direct mixing of body fluids and sea water. The amount of exchange is controlled by the area of the exchange surface, its resistance to exchange, and, lastly, the difference in concentrations of the gases in the two fluids on either side of the exchange surface (concentra- tion gradient). Where a concentration gradient exists diffusion of gases from higher to lower concentrations will take place. To maintain a concentration gradient a current system must continual- ly replenish fluid depleted of its gases. The most efficient current system is a counter current system. All pore-structures are either endothecal or exothecal (i.e., the exchange surfaces are either within the theca or outside it) and are composed of pores and canals (see Paul, 1968). All the pore- structures of the Holocystites fauna are exothecal and only this type will be considered. In exothecal pore-structures the pores open in the internal surface of the thecal plates and the fluids which flowed through the thecal canals were internal body fluids. The exchange surfaces were external (by definition) and therefore liable to me- chanical damage. They tend to be restricted in area and thicker than endothecal exchange surfaces. In holocystitids the pore-structures are humatipores (Paul, 1968) which have thecal canals entirely buried within the thecal wall and PAUL: NORTH AMERICAN HOLOCYSTITES FAUNA 31 calcified exchange surfaces. A canal arises in an internal pore and passes perpendicularly almost completely through the plate. Near the external surface of the theca the perpendicular canal divides into several canals which lie tangential to and just beneath the ex- ternal surface. The tangential canals reunite to form a second perpendicular canal which passes back through the thecal plate and opens in a second thecal pore internally (fig. 6b). The total area of the exchange surface in a theca is less than the total thecal surface area but the exchange surface is very much thinner than the thecal plates. In some species the humatipores are raised in tu- bercles. This increases the area of the exchange surface but renders mechanical damage more likely. Triamara, unlike holocystitids, has diplopores. In diplopores both the perpendicular canals pass right through the plate and open externally in a shallow depression called the peripore. The peripores of diplopores share so many common features with the peripodia of echinoid pore-pairs that it is almost impossible to escape the con- clusion that in life they gave rise to papulae similar to the tube-feet of sea urchins. Cystoid papulae have never been preserved and hence nothing can be stated with certainty about their shape, thickness, or extensibility. That they could form excellent ex- change surfaces is demonstrated by the respiratory tube-feet of recent sea urchins. It is supposed that currents entered the pore-structures through one of the two internal pores, passed through the canals where ex- change took place, and exited through the other pore. Quite pos- sibly the external surface also had ciliary currents but the directions of both internal and external currents are unknown. In some cases exothecal pore-structures show a regular asymmetry which is prob- ably associated with the current direction. The humatipores of Holocystites s.s. are an example of this (fig. 7). Possibly fluids came up the central perpendicular canal and descended the peripheral canal but the opposite direction seems equally plausible. The orientation of the humatipores in Holocystites is apparently random with respect to the orientation of the theca or individual thecal plates. Therefore no thecal current pattern is evident. INTERNAL ANATOMY Very little is known about internal anatomy in cystoids. The Holocystites fauna is not exceptional in this respect. Internal molds do reveal some details of internal organs, however. Apparently the gut passed straight down into the thecal cavity from the mouth. Presumably the path of the gut within the theca formed a loop but the precise arrangement is unknown. Internal molds of Holocystites spp. show two tubular structures which pass to the right between the mouth and anus from the positions of the gonopore and hydro- pore. One of the best examples of H. scutellatus Hall is illustrated in Figure 37b. The more posterior of the two ducts is less clearly defined, narrower, and is connected to the gonopore. It is inter- preted as the gonoduct and presumably the single gonad lay to the right in the thecal cavity. The more anterior duct is broader, much more obvious, and connects to the hydropore. Most likely it represents part of the water vascular system but it is rather thick and, in some examples, flares out to the right. It is unlikely that it represents part of the ring canal, but it could be either the stone canal or a madreporic caecum. This more obvious duct is readily apparent in speciments of H. cylindricus, H. alternatus, and H. scutellatus. A similar duct, usually referred to as the gonoduct, has been reported in other cystoids (Regn^ll, 1951). One internal mold of H. alternatus (FMNH 10967a) shows narrow grooves on the left side of the theca. These could represent the positions of mesenteries since they correspond to ridges on the internal surface of the theca. Unfortunately, nothing is known of the nervous systems nor of the internal connections of the pore- structures. 32 PALAEOECOLOGY A glance at Figure 10 shows that the Holocystites fauna in Eastern North America is very restricted in occurrence compared with the distribution of Middle Silurian rocks. A few specimens have been collected from Western Tennessee, Ohio, and possibly the Bruce Peninsular in Ontario, but the vast majority come from the Western sides of the Michigan Basin and the Cincinnati Arch (localities 2, 4-5, fig. 10). This occurrence must reflect ecological factors in the Middle Silurian seas of North America. As far as is known almost all Silurian cystoids inhabited shallow seas with a stable substrate. It is not surprising therefore that no cystoids have been reported from the geosynclinal Silurian of the Appa- lachians. However, other cystoids, notably Caryocrinites and several genera of the Callocystitidae, had a much wider distribution within the shallow seas of the stable interior of North America than holocystitids. The Holocystitidae and Callocystitidae seem to be mutually exclusive in occurrence but are possibly found together in Illinois, Wisconsin, and Tennessee. The distribution of cystoids in the Osgood beds of Indiana is apparently related to the "Ripley Island" positive area (fig. 11). Foerste (1904, p. 325) coined the term "Ripley Island" for an area in southeastern Indiana where the Lower Silurian Brassfield Lime- stone is missing and the succeeding Osgood rests directly on the Ordovician. The presence of Ordovician pebbles and cross bedding in the basal Brassfield adjacent to the "Ripley Island" suggested to Foerste that the Ordovician was undergoing erosion during lower Silurian times. He therefore explained the absence of Brassfield in the "Ripley Island" as being due to original non-deposition rather than to subsequent erosion. Only very rarely do pebbles occur in the basal Osgood which is a medium- to fine-grained dolomitic rock showing no evidence of current action. Clearly shallow seas, possibly with emergent shoals, existed in Southeastern Indiana during the Lower Silurian. The Osgood, which formed during earliest Middle Silurian times yields cystoids only in the area adjacent to and in- cluding the "Ripley Island." Apparently this Lower Silurian posi- tive area still exercised some control over the distribution of cystoids 38 34 FIELDIANA: GEOLOGY, VOLUME 24 M. SILURIAN M. & L. SIUreiAN SILURIAM ORDOVICIAN, SILUKIAN & DEVONIAN Fig. 10. Map to show outcrop of Silurian rocks in the Eastern United States and adjacent Ontario and localities for the Holocystites fauna. 1, west central Tennessee; 2, southeastern Indiana and adjacent Kentucky; 3, west central Ohio; 4, northeastern Illinois; 5, southeastern Wisconsin; 6, Wiarton, Ontario. Silurian of Quebec and New Brunswick not shown. and other fauna in the Early Middle Silurian. The Osgood is only locally richly fossiliferous but fossils are much more common in the area shown in Figure 11. North of Decatur County, Indiana and south of Bullitt County, Kentucky the Osgood is virtually barren. A similar and equally puzzling distribution of fossils oc- curs in the Waldron Shale in which there are several abundantly fossiliferous localities in Southeastern Indiana but which is generally barren north and south of the area shown in Figure 11 (Kindle and Barnett, 1909). Precisely what control the "Ripley Island" exer- cised is unclear. No cystoids (and few other fossils) occur in the lower half of the Osgood (fig. 12) and the formation as a whole is not typical of a sediment accumulated in very shallow waters with a stable substrate. It is predominantly fine-grained, silty, and shaley, and shows no evidence of wave or current action nor of shallow water conditions other than the presence of fossils which normally Fig. 11. Map of southeastern Indiana to show "Ripley Island" (diagonal shading) and cystoid localities (1-12, see Appendix). Isopachs of Brassfield at two-foot intervals. Symbols indicate localities where thickness of Brassfield was measured, as follows: o Brassfield absent, +l''-2'0", X 2'l"-4'0*, * 4'l''-6'0". Largely based on Foerste, 1897, plus the author's fieldwork in 1968-9. 35 36 FIELDIANA: GEOLOGY, VOLUME 24 PAUL, 1968 FOEBSTE, 1897 H. scutellatus T. ventricosus Main cystoid layer Few cystoids CRYSTALLINE LIMESTONE RUBBLY LIMESTONE CLAY, SHAI£ Fig. 12. Stratigraphic section from Upper Ordovician to base of Laurel Dolo- mite in Jefferson County, Indiana. Section measured just south of locality 7 (fig. 11) but lithologies and thicknesses vary little within the area covered by Figure 11. Records of cystoid occurrences taken from all localities. inhabit shallow seas. A combination of sedimentary and paleonto- logic evidence suggests that the seas were shallow but also very quiet. There is no evidence that the bottom sediments were dis- turbed by current action despite their fineness of grain. Priddy (1939) concluded on the basis of insoluble residues (particularly the presence of glauconite) and arenaceous foraminifera that the Niagaran rocks of Southeastern Indiana and West Central Ohio were deposited in 150-500 ft. of water and suggested the former PAUL: NORTH AMERICAN HOLOCYSTITES FAUNA 37 figure was more likely for sediments with a high proportion of elastics such as the Osgood. Further detailed fieldwork is necessary before the relationship between the "Ripley Island" and the distribution of fossils in the Osgood can be settled. It is doubtful that any of the cystoids in the Osgood are in place, although most cannot have traveled far since they are fairly complete and well preserved. However, only two of the several hundred specimens examined have any oral or anal covering plates preserved and none has the brachioles or arms preserved. Several specimens collected by the author were incomplete and this condi- tion is probably common. Although most specimens show definite evidence of attachment to a hard substrate, at least in the early stages of growth, only one specimen is known still attached to its substrate: a specimen of Trematocystis with the base of another specimen attached to it. At least in the case of Trematocystis there is very little associated fauna which could have acted as substrate. There is no detailed evidence available on the occurrence of these cystoids elsewhere. Brief searches in Northeastern Illinois and Ontario produced no specimens. In both these areas (and in Wisconsin and Ohio) the rocks are rather pure dolomites. The relationship, if any, of the occurrence of cystoids to the well known bioherms in these areas is regrettably unknown. Sediment types in Western Tennessee apparently resemble those in Southeastern Indiana but again no detailed information is available about the precise horizons and lithologies in which the cystoids occur. A rich macrofauna is associated with the Holocystites fauna in Southeastern Indiana. It consists primarily of shallow water invertebrates and includes crinoids, brachiopods, nautiloids, trilo- bites, gastropods, solitary and colonial corals, bryozoa, and worm tubes. Boyce (1956) records 44 species of invertebrates other than cystoids, almost all of which also occur in the Waldron Shale. However, none of the cystoids occur in the Waldron Shale. Browne and Schott (1963) record abundant arenaceous foraminifera and NicoU & Rexroad (1968) a rather sparse conodont fauna. This general fauna is no more evenly distributed (stratigraphically or geographically) than the Holocystites fauna. I have not found any fossils below the Trematocystis bed but neither have I searched particularly diligently for them. The Trematocystis bed (fig. 12) itself is generally barren but contains rare bryozoan colonies and complete specimens of the brachiopod Whitfieldella nitida (Hall) which also occurs rarely in the beds immediately above and below. 38 FIELDIANA: GEOLOGY, VOLUME 24 The rubbly lower part of the lower limestone band seems to be barren except for occasional cystoids (Triamara tumida and cylin- drical Holocystites spp.) and rare brachiopods. The shale between the two limestone bands is often barren but is locally abundantly fossiliferous and also more lime rich. Much of the fauna described by Boyce can be collected in the middle shale and succeeding upper limestone band. Atrypa cf. reticularis, Pisocrinus gemmiformis, Stephanocrinus gemmiformis, and the solitary coral Enterolasma caliculum (Hall) are particularly prominent members of the fauna. The Upper Osgood Shales are rarely well exposed. In several places, particularly in Southern Jefferson County, they are dark grey and totally unfossiliferous. However, most of the fauna found at Napoleon and Osgood comes from the Upper Shales which are richly fossiliferous in these two localities. Foerste (1897) records many fossils from the Upper Shales at Big Creek where several thin limey lenses occur. Rich faunas are also recorded from other Holocystites fauna localities but no details are available. If anything, the associated faunas indicate similar environments in both dolomitic and non- dolomitic areas. The principal difference seems to lie in the ratio of lime to elastics in the sediments. Correlation of cystoid-bearing strata in the principal localities for the Holocystites fauna has proved very difficult. Recent studies of conodonts promise that this complicated task is about to be re- solved. Zonal conodonts from the Osgood of Southeastern Indiana indicate a basal Wenlock (Middle Silurian of the standard British succession) age (Rexroad, 1967). The Osgood is apparently equiv- alent to the middle part of the Joliet Dolomite of northeastern Illinois which may well be the source for the cystoids from Romeo and Lemont, Illinois. Cystoids from these sediments are thus slightly older than most of the Holocystites fauna and Triamara is confined to this horizon. The exact age of the "Osgood" in west central Tennessee is unknown but is presumed to be approximately the same as that of the Osgood in southeastern Indiana. Tremato- cystis and Pustulocystis are not definitely known from localities other than the former two. Pentacystis is unknown outside south- eastern Indiana. Undoubted Holocystites s.s. are known from somewhat younger horizons in Illinois, Wisconsin, and Ohio (fig. 13). The detailed correlation of the dolomites in Ohio is still a subject for discussion. The genera of the Holocystites fauna seem to have restricted vertical distribution and are potentially useful PAUL: NORTH AMERICAN HOLOCYSTITES FAUNA 39 in correlation. However, Holocystites cylindricus Hall is known from Newsom, Tennessee and Grafton, Wisconsin. The precise horizon is unknown at both localities but, if the Newsom locality is in the Osgood and the Grafton locality is in the Racine, this species would seem to range from the oldest to the youngest hori- zons, all of which are still Middle Silurian. Teller (1911, p. 202) doubts the accuracy of the Grafton locality but confirms the Racine age of H. cylindricus in Wisconsin. A most interesting paleoecological feature of the Holocystites fauna is the frequent occurrence of small hemispherical pits in cystoid thecae from the Osgood of southeastern Indiana. These pits are apparently the result of boring or solution by some other organism which in most cases attacked the cystoids while they were alive. The evidence for this interpretation is as follows: clearly plate material has been dissolved or scraped away from the sites of the pits. The circular outline and smooth sides of the pits suggests boring by rotary action possibly aided by chemical solution. For Fig, 13. Correlation diagram to show the stratigraphic horizons within the Middle Silurian at which the Holocystites fauna (H) occurs in Ohio, Indiana, Illinois, and Wisconsin. Precise horizons are unknown in Illinois and Wisconsin but it seems likely that all Illinois localities are either in the Upper Joliet or the Waukesha, whereas all Wisconsin localities are probably in the Waukesha or Racine. Based on Mound, 1968; Pinsak and Shaver, 1963; Rexroad and Orr, 1967. 40 FIELDIANA: GEOLOGY, VOLUME 24 the purposes of this discussion the pits are regarded as borings. Although they vary in size the pits have a fairly constant maximum diameter of 3.5 mm, and maximum depth of 2 mm. Many of the pits have raised rims which were secreted by the cystoids in re- sponse to the borings. In the holotype of Pentacystis simplex n. sp. (AMNH 20271a) tangential canals of humatipores occur parallel to the walls and floor of three pits (fig. 42c) clearly indicating that the cystoid modified its pore-structures after the pits were formed. Both lines of evidence indicate that the cystoids were alive when the borings were made. Boring activity on the part of modern organisms is related to three vital needs: predation, parasitism, and protection. Calcium uptake is a fourth possible explanation of boring activity but is unnecessary for marine organisms since most sea water is saturated with respect to calcium carbonate. In attempting to interpret the borings two salient facts emerge: 1) not one of several hundred pits examined penetrates the thecal cavity, and 2) although other echinoderms and other phyla occur in the same rocks only holo- cystitids bear these pits; not even Triamara was attacked. The first point would seem to rule out predation as a possible explana- tion of the borings. For successful predation boring organisms must gain more energy from the nutrients they obtain by boring than they expend in the boring activity. No holocystitid theca was penetrated and hence the boring organism could not have gained access to the majority of the soft tissue. Surely the small amount of soft tissue within the plates could not have compensated for the energy expended in getting it? Despite this, several cystoids have 50 or more pits, indicating repeated attacks. Very similar but usually smaller pits are known from other horizons. This author has collected echinoderms with them from the Middle Ordovician, Benbolt Formation of Tennessee and Virginia, the Middle Silurian, Waldron Shale of Indiana, and from the Lower Mississippian, Burlington Limestone of Missouri. In these cases other echino- derms were attacked, including crinoid stems. Again, there is little chance of adequate nourishment having been obtained if predation was the object. However, the organisms which were responsible for the borings survived at least from the Middle Ordovician to the Mississippian and were presumably well adapted to their mode of life. Protection is possible but in recent examples the borings tend to be tubular rather than shallow pits. No trace of organisms oc- PAUL: NORTH AMERICAN HOLOCYSTITES FAUNA 41 cupying the pits is known. Parasitism seems the most likely ex- planation. The cystoids were attacked but not killed by the boring organism. They responded by modifying the plates and pore- structures at the sites of the borings. Modern parasites are often extremely host-specific which may explain why only holocystitids were attacked in the Osgood and principally Calceocrinus in the Waldron. Even if parasitism is accepted as the probable cause of the pits it does not help identify the parasite. Recent echinoderms do not have any parasites which produce similar pits. A variety of recent echinoids and crinoids produce galls in response to the activities of parasitic myzostome annelids, stiliferan gastropods, and copepod crustaceans. Possibly the pits in Holocystites et al. represent a similar effect. Recent crinoids often play host to large numbers of such parasites as did holocystitids. Finally, ectopara- sites such as stiliferan gastropods bore into the skeletal material and derive nutriment from the soft tissues enclosed in the skeleton. Such "leech-like" activity can support the parasite for extended periods of time whereas simple predation (i.e., taking a bite) would not. The conclusion that holocystitids suffered from parasites similar, but not identical, to those of recent echinoderms seems inevitable. The distribution of the pits on any one theca can be used to interpret the life attitude of the cystoid. It is likely that the boring organisms were epifaunal and lived on the exposed surface of the cystoid since many specimens have pits all over the oral surface and around the ambitus. It is assumed that these cystoids lived upright. Cylindrical species of Holocystites s.s. are an exception to the upright position. Two examples have pits all over one side of the theca only. Most cylindrical Holocystites specimens have the mouth offset to one side of the oral pole. Apparently cylindrical Holocystites fell over onto one side early in life but continued to grow and modified the oral area so that the mouth pointed up- wards. The holotype of "H." canneus Miller has borings along one side of the theca and the mouth is offset toward the side with the borings (fig. 33b). For further discussion of the pits see the individual holocystitid genera. Members of the Holocystites fauna were epifaunal, at least early in their lifetimes, and required a solid substrate. Yet, as stated earlier, only one example of a cystoid still attached to its substrate is known among Silurian species. This specimen is also the only known example of a Silurian cystoid epifaunal on another cystoid. 42 FIELDIANA: GEOLOGY, VOLUME 254 In Brightonicystis gregarius, the British Ordovician species, two of three complete thecae have others attached to them. Haplo- sphaeronis epifaunal on other cystoids are common in the Upper Ordovician Boda Limestone of the Siljan District in Sweden. Most cystoids in the Holocystites fauna of North America have indeter- minate attachment areas. However the following organisms have definitely been recognized from their impressions on the base of the theca: straight and cyrtoconic nautiloids (Trematocystis) , trilobite cephala, and twig-like Bryozoa (Pustulocystis) . One can generalize about the ecology of individual genera from the size of the attach- ment areas. Trematocystis and Pentacystis almost always have large, smooth, flat attachment areas. Holocystites and Pustulocystis almost always have small attachment areas and may have stem- like projections to the theca aborally. Triamara is variable — some specimens have large attachment areas but most do not. The type of "Holocystites" gyrinus is about 140 mm. long and has an attach- ment area only 5 mm. at its widest. Trematocystis and Pentacystis apparently required a large flat surface to live. There is no clear evidence in either genus that any specimens fell over or were dis- turbed during life. The other three Silurian genera apparently survived where only small surfaces (usually fragmentary shell material) existed for attachment. In Triamara ventricosa and cylin- drical species of Holocystites most specimens fell over and lived on their sides for considerable portions of their lives. The small British species of Brightonicystis has a relatively large attachment area whereas the large Swedish species had a much smaller area. No evidence as to life orientation of the latter species is available. The cystoids themselves acted as substrate for other epizoic animals. Brightonicystis gregarius supported Bryozoa and worm tubes as well as other B. gregarius. Among Silurian genera a rich epifauna is found containing the following groups in decreasing order of abundance: Bryozoa (probably three species), crinoid stem bases (five types), worm tubes (three types), sponges, corals, the brachiopod Petrocrania, and one Trematocystis. From available specimens it seems that Triamara, which on average has the largest theca, supported the least epifauna. At least two large and sub- stantially complete thecae are known without any epifauna at all. Trematocystis apparently supported the richest epifauna. From the general distribution of the epifauna over the theca and specific instances of epifauna growing over ambulacral facets, across peri- PAUL: NORTH AMERICAN HOLOCYSTITES FAUNA 43 stomes and periprocts which lack covering plates (e.g., fig. 38b) and across attachment areas, it is clear that most, if not all, epifauna attached after the death of the cystoids. USNM S3058a, an example of Trematocystis, has a bryozoan colony curved around an ambulacral facet as though the "brachiole" was present when the colony grew. This is the only example of epifauna which may have been growing during a cystoid's lifetime. In contrast to the epifauna, the majority of the bored pits seem to have formed during the lifetime of the cystoids. Field evidence of the orientation of epifauna on cystoids is almost entirely lacking: most of the specimens collected by the author lack epifauna. Further information on epifauna is given under each genus. SYSTEMATIC PALEONTOLOGY All members of the Holocystites fauna bear dipores (Paul, 1968), a type of pore-structure characteristic of the Class Diploporita. The epithecal portions of the ambulacral grooves are short with the ambulacral facets close to the mouth. This, together with the plate arrangement, indicates affinities of the Holocystitidae with the superfamily Sphaeronitida. Despite Miller's creation of a third family, the Holocystitidae, recent classifications (Bassler and Moodey, 1943; Regn^ll 1945; Kesling, 1963, 1968) accept only two families within the Sphaeronitida, viz: the Sphaeronitidae and Aristo- cystitidae. Holocystites s.l. has generally been included within the latter family yet the structure of the oral area indicates closer afRnites with the former. All genera except Triamara, also possess some unique features, notably the presence of humatipores, which leads me to resurrect Miller's family Holocystitidae. More drastic is the realization that the oral area of aristocystitids differs fundamentally from that of all other diploporites and warrants the recognition of a separate super- family. Accordingly, the superfamily Sphaeronitida is restricted to include only the Sphaeronitidae and Holocystitidae, and the family Aristocystitidae is elevated to superfamily rank. Recently Chauvel (1966) has erected several subfamilies within the Aristocystitidae which are probably valid but further study is needed to determine their precise limits. Such a study is outside the scope of the present work and hence Triamara is not assigned to a subfamily within the Aristocystitidae. The following key is offered in the hope that it will aid identifica- tion of members of the Holocystites fauna which has hitherto been next to impossible: Key to Families and Genera of the Holocystites Fauna 1. Pore-structures diplopores (fig. 6a), wide ambulacral grooves ending in large facets Aristocystitidae 2. Pore-structure humatipores (fig. 6b), four or five narrow ambulacral grooves at margins of large pentagonal or quadrilateral peristome. . .Holocystitidae 3. 44 PAUL: NORTH AMERICAN HOLOCYSTITES FAUNA 45 2. Only one Silurian genus, with three ambulacral grooves each ending in two facets Triamara, 3. Peri-oral plates (p. 71) very small or absent, thecal plates always smooth. Pentacystis. Peri-oral plates present and well developed 4. 4. Ten peri-oral plates, Ordovician Brightonicystis. Six peri-oral plates, Silurian 5. 5. Six f acetal plates Ptistulocystis. Eight facetal plates ... \ ......... 6. 6. Four ambulacral facets, thecal plates always smooth and of one genera- tion Trematocystis. Five ambulacral facets, thecal plates pustular or smooth, and of one to three generations Holocystites s.s. Keys to species are given under individual genera. Class Diploporita Miiller, 1854 Superfamily Aristocystitidae Neumayr, 1889 (Nom. transl.) Definition. — Stemless diploporites with variable shaped theca, all plates pierced by diplopores; relatively small mouth at junction of two wide ambulacral grooves, neither, one, or both of which may divide once, ambulacral grooves covered by series of cover plates which were immovable; periproct with simple anal pyramid of triangular plates; circular gonopore and slit-like or madreporite-like hydropore between mouth and anus, latter nearer the mouth some- times with multiple openings. The oro-anal region is the most characteristic feature of the Aristocystitidae. All genera in which the oral surface is known have two broad ambulacral grooves which unite at a small oval mouth. The mouth is developed entirely within these grooves. In typical Aristocystites the ambulacral grooves do not divide and each terminates in a single ambulacral facet (fig. IG). In Calix, Hippocystitis, and Sinocystis both ambulacral grooves divide once and each branch terminates in one facet to give a total of four (fig. IH). Triamara differs from all other aristocystitid genera in that (1) only one (the left) of the two ambulacral grooves divides and (2) all three resulting branches end in a pair of facets giving a total of six (fig. II). The nature of the appendages which articulated on the ambulacral facets is unknown in all genera and species ex- cept one. Chauvel (1966, p. 41) described the following in Glapho- cystis globulus: "Latex molds have revealed in several examples remains of brachioles, the ossicles which form them are indistinct but [they] appear [to be] alternate, the ventral gi'oove is clearly 46 FIELDIANA: GEOLOGY. VOLUME 24 visible. In aj one of the brachioles has a base 1.5 mm. wide but the diameter decreases rapidly to about % mm. The remains observed are scarcely more than 2 mm. long." Clearly the structures are unbranched but whether they are biserial is not established beyond doubt. I know of only two other examples (one published, one not) of diploporite cystoids with ambulacral appendages preserved : both have uniserial pinnules. The ambulacral grooves of Aristocystites and Triamara are lined with small pores which opened into the grooves and were covered by the covering plates in life. The pores connect to canals which pass right through the thecal plates and are identical in all respects to the perpendicular canals of normal diplopores. In several examples of Triamara these oral pores form lines near the margins of the ambulacral grooves and are arranged in pairs. Presumably these pores gave rise to papulae which had a tactile or chemo-receptive function. Central in the ambulacral grooves is a narrow food groove much like those of other diploporite cystoids. As pointed out by Paul and Kesling (1968) in their review of Chauvel's (1966) paper, the Aristocystitidae contains a number of species of very variable cystoids which are exceptionally difficult to characterize. Difficulties caused by variability are compounded by preservation. Most species have large thecae which are rarely preserved complete and uncrushed. Undoubtedly, the family Aristocystitidae can be subdivided as Chauvel (1966) has already attempted. Triamara does not fit into any of the three subfamilies erected by Chauvel and it is not assigned to any of them; neither is a new subfamily proposed for it. Triamara Tillman, 1967 Type species. — Triamara cutleri Tillman, 1967 = Holocystites ventricosus Miller, 1879. Fig. 14. Camera lucida drawings of the diplopores of Triamara spp. A, C, Triamara tumida (Miller), 1879. A, holotype of "Holocystites" adipatus Miller, FMNH 6020; C, holotype of "H." tumidus Miller, FMNH 8763. B, D, Triamara ventricosa (Miller), 1879. B, holotype of "H." gyrinus Miller and Gurley, FMNH 5997; D, holotype of "H." amplus Miller, FMNH 6021; E, Triamara multiporata n. sp. holotype, FMNH 15217. F, Triamara laevis n. sp. holotype, USNM S5446. G, Triamara sp. USNM 166580. S-plate suture, peripores stippled, positions of perpendicular canals (where visible) black. All figures approx. X9. A ♦) «*^ C T3 -^^ ^ 0) — a (-1 en IV 0 ^ /^^ ^ w^ / JMV II -,-fHr ^^^J/^\r^ ""X/'l/ 3 \ 3 ^_, ..^ *il _>\/ ^^»^^^X l^^*V»^ / /X^ ' 1*N^^>8 ' / 1* i^~ y ~y \ y 1 ""v*? V / 1 ' 2 11' •g 1 '^ \ Fig. 46. Pentacystis wykoffi (Miller), 1891. Camera lucida drawing of the oral area of holotype, FMNH 6022 to show small peri-oral plates. G — gonopore, m — mouth, small numerals = peri-orals, large numerals = facetals, roman num- erals = ambulacral facets. Specimen is crushed and distorted, FA 8 has been pushed over FA 1 which shows as a very narrow plate. The specimen is unusual in that ambulacral facet I is developed on two facetal plates, not one. facets 4,1^ mm. X 4 mm. with raised rims and slight median ridge. In holotype ambulacral facet I shared by FA2 and FAS, other four entirely within a single plate. In USNM S3063a all five are each within a single plate so that FAl, FAS, and FAT do not bear facets. Periproct: Crushed, separated from peristome by POl, P06, and FAl, the latter is very narrow in the holotype. Anals unknown. Gonopore: Pentagonal, 0.5 mm. across, flush with surface of FA2, about 1.5 mm, from edge of ambulacral facet I in holotype- Not detectable in USNM SS06Sa. Hydropore: Slit-like, across suture P01:P06 in USNM S3063a (within peristome border). Not detected in other specimens. Attachment: Direct, attachment area large. Remarks. — P. wykoffi is immediately distinguishable from other species of Pentacystis by the presence of small peri-oral plates al- though these are sometimes so small that they are entirely within the peristome. P. wykoffi has oral pores on the peri-orals as do all other holocystitids. P. simplex and P. sphaeroidalis lack both oral pores and peri-orals. It seems therefore that oral pores are only de- veloped when peri-orals are present. PAUL: NORTH AMERICAN HOLOCYSTITES FAUNA 115 Pentacystis was probably derived from Holocystites, one species of which has smooth plates and another a similarly shaped theca. All species of Holocystites have five ambulacral facets and presumably Pentacystis evolved by the reduction of the peri-orals until they disappeared completely. If so P. wykoffi, is more closely related to Holocystites than other species of Pentacystis since it still retains traces of the periorals. The humatipores of Pentacystis most closely resemble those of Trematocystis and do not have the radiate pattern of tangential canals typical of the humatipores of Holocystites. Since Pentacystis and Holocystites appear at the same stratigi'aphic hori- zon, it is also possible that Holocystites evolved from Pentacystis, although this seems less likely. Trematocystis Jaekel, 1899. Type species. — Holocystites subglobosus Miller, 1889 = Holocys- tites globosus Miller, 1878. Trematocystis Jaekel, 1899, Stammesgeschichte der Pelmatozoen., 1, p. 413; Kesling, 1963, Contr, Mus. Paleontol. Univ. Mich., 18, p. 115; Kesling, 1968, Treatise on Invertebrate Paleontology, part S, p. S258, fig. 148, 2a-b. Diagnosis. — A genus of Holocystitidae with globular theca, six peri-orals, eight facetals, four ambulacral facets, smooth plates of one generation. Description. — The theca in Trematocystis is globular to inverted pyriform and usually rather small for the fauna (25-35 mm. in diameter, only rarely reaching 50 mm.). The peristome is quad- rilateral and always has four ambulacral facets. The peristome border is always pitted with oral pores. The mouth is surrounded by six peri-orals and then eight facetals of which FAl, FA3, FA5, and FAT do not bear ambulacral facets. The palate (roof over the mouth) is composed of six plates (figs, le, 50c). The ambulacral facets are raised but not as much as those of Pustulocystis. They are planar and have a median ridge. Each facet is developed on a single facetal plate. The gonopore opens either in FAl or in FA2 (fig. 47) and is in a tubercle in well-preserved specimens. The hydropore is slit-like or oval and in a tubercle across suture P01:P06. The periproct is pentagonal or hexagonal and surrounds a rounded anal opening. The total number of thecal plates is small for the fauna (usually 50-80) but reaches 105 in T. wetherbyi. The plates are always of 116 FIELDIANA: GEOLOGY, VOLUME 24 Fig. 47. Diagrammatic representation of the oro-anal regions of Tremato- cystis globosus (Miller) (A) and T. rotundus (Miller) (B). The principal difference lies in the position of the gonopore (g). A — anus, f — facet, g — gonopore, h — hydro- pore, m — mouth, small numerals indicate peri-oral plates, large numerals indicate facetals, roman numerals indicate ambulacra. Not to scale. one generation and smooth or finely granular. They are arranged in circlets of seven to 16 plates. The humatipores are buried beneath the smooth plate surfaces and have two to seven tangential canals. Attachment was direct by a small to moderate attachment area. There is never a truly stem-like projection aborally. Paleoecology. — Trematocystis was attached directly by a moderate attachment area at the base of the theca which varies from a fifth to a third of the ambital diameter. Trematocystis frequently at- tached to smooth flat surfaces but two USNM specimens were attached to nautiloids, one cyrtoconic, the other orthoconic. USNM S3068B attached to a smooth surface to which a bryozoan colony was already attached and USNM S3075a has the base of another Trematocystis attached to it. This is the only known occurrence of a Silurian cystoid epifaunal on another cystoid, although this was apparently not unusual in the Ordovician genus Brightonicystis. In many specimens the size and shape of the attachment area was at least partially controlled by the substrate to which it was attached. In other cases the attachment area is almost perfectly circular and flat. In such cases the attachment area rarely exceeds a third the ambital diameter. It seems that Trematocystis attached to a variety of objects which subsequently proved to be more or less PAUL: NORTH AMERICAN HOLOCYSTITES FAUNA 117 suitable as the cystoid grew. Despite this there is no clear evidence that any Trematocystis specimens fell over or became detached during life. Trematocystis supported an abundant and varied epifauna which seems to have become attached after the death of the cystoids in most cases. Forty specimens in the USNM collections display three types of Bryozoa, three types of worm tube, five types of crinoid stem bases, the brachiopod Petrocrania, and the coral Favosites as epifauna. In many cases, Bryozoa and worm tubes are found growing over attachment areas, ambulacral facets, and the peristome and periproct which lack covering plates and, in these cases, epi- faunal attachment obviously took place after the death of the cystoids. In only one case (USNM S3058a) a bryozoan colony has an embayment by an ambulacral facet suggesting that it grew around the "brachiole" and it is possible this bryozoan was growing while the cystoid was alive. Among these same 40 specimens only three have bored pits in them. All three have pits in the oral surface and one specimen has pits with distinct rims. This might suggest that Trematocystis escaped the attentions of the boring organism but the holotype of "H." turbinatus has 41 borings in the oral surface which are so closely packed together that the details of this surface are un- recognizable. This specimen poses a problem. Almost all the pits have very distinct raised rims which suggests the cystoid was bored into while alive. On the other hand, the ambulacral facets are almost unrecognizable, and while there are no pits actually in the areas where the facets should be, it is difficult to imagine how the "brachioles" could have been present and functioning when the pits were bored. This in turn suggests the cystoid was dead. Perhaps the boring activity so upset the physiology of the cystoid that it secreted excess calcite all over the oral surface indiscrimi- nately. The holotype of "H." rotundus Miller is similarly damaged and has 18 pits on the oral surface. Almost all of the pits have rims and are smaller than usual. These are also exclusively on the oral surface. Trematocystis seems to have suffered less frequently from the attacks of the boring organism but the few attacks that took place were unusually severe. The fact that borings occur exclusively in the oral surface strongly suggests that these cystoids lived upright. Taxonomy. — Trematocystis was erected by Otto Jaekel because he thought that all the internal molds James Hall originally de- 118 FIELDIANA: GEOLOGY, VOLUME 24 scribed as Holocystites represented primitive "cladocrinoids," Jaekel designated "H." subglobosus Miller as type species and stated that the pore-structures with w-shaped canals (i.e., humatipores) were characteristic of the genus. "H." subglobosus is one of about a dozen of Miller's "species" which have six peri-orals, eight facetals, four ambulacral facets, and smooth plates of one generation. They differ in the latter two characters from Holocystites as here restricted, and in the former pair of characters they differ from Pustulocystis and Penlacystis. On these grounds Trematocystis Jaekel is retained as a valid genus but with a restricted definition. Trematocystis is undoubtedly the most common genus of cystoids in Southeastern Indiana and hence Miller was able to describe at least 11 "species" which are here reduced to three, one of which may not be truly distinct. The names "H." globosus, "H." wetherbyi, and "H." rotundus are the oldest available accompanied by reasonably well- preserved type specimens. Where the gonopore can be detected it is either in FAl or in FA2 without exception. This feature is used to separate two species (fig. 47). The holotype of H. wetherbyi differs from all other specimens in having more thecal plates (there are nine facetals and four circlets of 15-16 plates: no other specimens have more than one circlet with 15 odd plates in it). For this reason H. wetherbyi is provisionally accepted as a distinct species. Key to Species of Trematocystis 1 . Gonopore in FA2 T. globosus. Gonopore in FAl 2 2. Plates in circlets of up to 12 plates T. rotundus. Three or four circlets of 15-16 plates T. wetherbyi. Geographic distribution. — Trematocystis is recorded from Big Creek, West Fork of Indian Kentuck Creek, 12 miles North of Madison and North Madison, all in Jefferson County, Indiana, and from Osgood and Vestal Branch of Indian Kentuck Creek in Ripley County, Indiana, localities 2-5, 9, 11, and possibly 1 (fig. 11, ap- pendix, pp. 163-164). Fig. 48 (opposite). Trematocystis globosus (Miller), 1878. a-b. Stereophotos of lectotype of Holocystites globosus (Miller), FMNH 8765a. a, oral view, b, pos- terior view to show thecal outline, c-d, Stereophotos of holotype of "H." hammelli Miller, FMNH 5999. c, oral view, d, anterior lateral view. e-h. Lectotype of "H." subglobosus Miller, FMNH 8766a. e, stereophotos of oral view. Note gonopore beside ambulacral facet L h, lateral view, f-g, Holotype of "H." scitu- lus Miller, FMNH 6012. f, stereophotos of oral view. Note position of gonopore. g, lateral view. All figures natural size, all whitened with ammonium chloride sublimate. 119 120 FIELDIANA: GEOLOGY, VOLUME 24 Stratigraphic range. — All specimens of Trematocystis collected by the author come from a grey-green silty rock band 9 in. to 1 ft. 3 in. below the base of the Osgood Limestone. This is the "main cystidean bed" of Foerste (1897, p. 257) which is apparently slightly thicker (8 in.) in Big Creek. Foerste also records rare cystoids in the beds immediately above and below. These could also be Trematocystis. Trematocystis globosus (Miller), 1878. Figures 47a, 48-51. Holocystites globosus Miller, 1878, Jour. Cincinnati Soc. Nat, Hist., 1, p. 133, pi. 5, figs. 5, 5a; ? Parks, 1913, Canad. Geol. Surv. Guidebook, 4, p. 132; Bassler and Moodey, 1943, Geol. Soc. Amer. Spec. Pap., 45, pp. 44, 168. Holocystites hammelli Miller, 1889, North American Geology and Palaeon- tology, p. 254, figs. 336-337; Bassler and Moodey, 1943, Geol. Soc. Amer. Spec. Pap., 45, pp. 44, 169. Holocystites subglobosus Miller, 1889, North American Geology and Palaeon- tology, p. 255, fig. 338; Bassler and Moodey, 1943, Geol. Soc. Amer. Spec. Pap., 45, pp. 44, 170. Holocystites scitulus Miller, 1891, Rept. Indiana Dept. Geol. Nat. Resources, 17, (adv. pub.), p. 14, pi. 2, figs. 5-6; Miller 1892a, op. cit., p. 624, pi. 2, figs. 5-6; Miller, 1897, North American Geology and Palaeontology, 2nd app., p. 747, fig. 1359; Bassler and Moodey, 1943, Geol. Soc. Amer. Spec. Pap. 45, pp. 44, 170. Trematocystis hammelli Jaekel, 1899, Stammesgeschichte der Pelmatozoen, 1, p. 409. Trematocystis subglobosus Jaekel, 1899, Stammesgeschichte der Pelmatozoen, 1, 414, pi. 4, figs. 2, 2a; Kesling, 1968, Treatise on Invertebrate Paleon- tology, part S, fig. 148, 2a-b. Megacystis subglobosus Miiller, 1963, Lehrbuch der Palaozoologie, Band 2, Teil, 3, p. 287, fig. 393. Diagnosis. — A species of Trematocystis with gonopore in FA2. Types.— "H." globosus Miller, FMNH 8765a here selected as lectotype, paralectotypes FMNH 8765b, c. "H." hammelli Miller, holotype (monotype), FMNH 5999. "H." subglobosus Miller, Fig. 49 (opposite). Trematocystis globosus (Miller, 1878. a, Stereophotos of lectotype of "Holocystites" subglobosus Miller, FMNH 8766a to show details of humatipores. Note the calcite meshwork in the floors of the tangential canals is coarser than that between them. X12. b, Stereophotos of AMNH 20971, lateral view to show thecal outline and large attachment area. Natural size, c, USNM S 5449, aboral view to show attachment to orthoconic nautiloid. X3. d, AMNH 2156a, lateral view, natural size, e, USNM S3063a, oral view to show preserved oral and anal covering plates (palatals and anals). X2.6 (cf. fig. 50c). f, USNM 3075c. Aboral view to show attachment to ?cyrtoconic nautiloid. X2.3. All whitened with ammonium chloride sublimate. 121 122 FIELDIANA: GEOLOGY, VOLUME 24 Fig. 50. Tremaiccystis globosus (Miller), 1S78. Cair.era lucida drawings of the oro-anal areas of three specimens. A, UMMP 45862, normal example with gono- pore (g) in FA2. B, BMNH E7630 abnormal example with double gonopore. C, USNM S3063a. Example with palatals and anals (fine stipple) preserved, a — anus, f — ambulacral facet, g — gonopore, h — hydropore, m — mouth, small numerals indicate peri-orals, large numerals indicate facetals, roman numerals indicate ambulacra. FMNH 8766a, here selected as lectotype, paralectotype FMNH 8766b. "H." scitulus Miller, holotype (monotype), FMNH 6012. Horizon and locality. — Four types are from Jefferson County, Indiana. Most specimens with a precise locality come from Big Creek and this is designated type locality. The "main cystidean bed" is designated type stratum. Specimens I collected come from the Trimatocystis bed at locality 7 (fig. 11). PAUL: NORTH AMERICAN HOLOCYSTITES FAUNA 123 Material.— At least 46 complete thecae. AMNH, BMNH, FMNH, UC, UMMP, USNM. Description. — Theca: Irregularly inverted pyriform to globular, without aboral stem-like projection, and with moderate, generally flat attachment area. Plates not pustular and without impressed sutures. Plates: Of one generation, up to 12 mm., smooth or finely granular, all bearing numerous humatipores. Arranged in circlets, maximum number of plates per circlet 8-14. Humatipores: Buried, not developed as pustules, with three to seven tangential canals, generally closely packed together, 1.1- 1.4 mm. long (fig. 51). Mouth: Rounded oblong, 4-5 mm. across, set within quadri- lateral peristome. Peristome border pitted with oral pores, 1-1 3^ mm. wide and bearing four straight food grooves from ambulacral facets to mouth border. Six palatals in USNM S3063a (fig. 50c) which cover a peristome 7 x 5.5 mm. Palatals ornamented with fine granulation like that of other thecal plates. Mouth surrounded by six peri-orals and then eight facetals of which FAl, FAS, FA5, and FAT do not bear ambulacral facets. Four oval ambulacral facets, 4x3 mm. to 5 x 3.7 mm., raised up on facetals, with median ridge, slight rim, and concave lateral areas. Fig. 51. Trematocystis globosus (Miller), 1878. Camera lucida drawings of humatipores in two typical examples. A, AMNH 2156b. B, AMNH 2156a. Both approx. Xl4. 124 FIELDIANA: GEOLOGY, VOLUME 24 Anus: Rounded, about 5 mm. in diameter, with narrow periproct border defining generally pentagonal periproct. In USNM S3065a, periproct pentagonal, 6 mm. in diameter, and covered with low pyramid of five triangular plates. GONOPORE: Small, circular pore, 0.5 mm. in diameter and one- half mm. from edge of ambulacral facet I in FA2. In well-preserved specimens gonopore is in distinct tubercle about twice the diameter of the pore. BMNH E7630 has two gonopores (fig. 50b). Hydropore: Slit-like, across P01:P06 suture, often raised in tubercle. Slit 0.8-1.0 mm. long. Attachment: Direct, by a generally flat attachment area be- tween one-third and one-fifth ambital diameter. Remarks. — Generally T. globosus has slightly more thecal plates per theca than T. rotundus Miller but still considerably less than T. wetherbyi. There is very little to distinguish T. globosus and T. rotundus from each other apart from the position of the gonopore. Apparently, the hydropore is more definitely slit-like in T. globosus but this is not a constant feature. Both T. globosus and T. rotundus are common in the Osgood beds of southeastern Indiana. Trematocystis rotundus (Miller), 1879. Figures 47b, 52-54. Holocystites rotundus Miller, 1879, Jour. Cincinnati Soc. Nat. Hist., 2, p. 107, pi. 9, figs. 3, 3a-b; Bassler and Moodey, 1943, Geol. Soc. Amer. Spec. Pap., 45, 44, 170. Holocystites turbinatus Miller, 1880, Jour. Cincinnati Soc. Nat. Hist., 2, p. 259, pi. 15, figs. 7, 7a; Miller, 1889, North American Geology and Palaeontology, p. 255, fig. 339; Bassler and Moodey, 1943, Geol. Soc. Amer. Spec. Pap., 45, pp. 44, 171. Holocystites gorbyi Miller, 1891, Rept. Indiana Dept. Geol. Nat. Resources, 17, (adv. pub.), p. 14, pi. 2, figs. 3-4; Miller, 1892a, op. cit., p. 624, pi. 2, figs. 3-4; Bassler and Moodey, 1943, Geol. Soc. Amer. Spec. Pap., 45, pp. 44, 168. Holocystites commodus Miller, 1891, Rept. Indiana Dept. Geol. Nat. Resources, 17, (adv. pub.), p. 14, pi. 3, figs. 1-2, 5-6; Miller, 1892a, op. cit., p. 624, Fig. 52. Trematocystis rotundus (Miller), 1878. a, c-d, holotype of "Holo- cystites" rotundus Miller,' FMNH 6004. a. Lateral view to show thecal outline, c. Oral view to show peristome, periproct, gonopore, hydropore and many bored pits on oral surface, d, Aboral view to show attachment area. b. Stereophotos of lectotype of "H." commodus Miller, FMNH 6018. Lateral view. e. Stereophotos of holotype of "H." parvus Miller, FMNH 6010. Lateral view. f-g. Holotype of " H." turbinatus Miller, MCZ 79. f, lateral view to show thecal outline, g, oral view to show abundant bored pits. All figures natural size, all whitened with ammonium chloride sublimate. 125 126 PAUL: NORTH AMERICAN HOLOCYSTITES FAUNA 127 pi. 3, figs. 1-2, 5-6; Bassler and Moodey, 1943, Geol. Soc. Amer. Spec. Pap., 45, pp. 44, 168. Holocystites indianensis Miller, 1891, Rept. Indiana Dept. Geol. Nat. Re- sources, 17, (adv. pub.), p. 15, pi. 3, fig. 7; Miller 1892a, op. cit., p. 625, pi. 3, fig. 7; Bassler and Moodey, 1943, Geol. Soc. Amer. Spec. Pap., 45, pp. 44, 169. Holocystites parvus Miller, 1891, Rept. Indiana Dept. Geol. Nat. Resources, 17, (adv. pub.), p. 16, pi. 4, figs. 4-5; Miller, 1892a, op. cit., p. 626, pi. 4, figs. 4-5; Bassler and Moodey, 1943, Geol. Soc. Amer. Spec. Pap., 45, pp. 44, 169. Holocystites papulosiis Rowley, 1903, (non Miller, 1891), Cont. Ind. Palae- ontol., 16, p. 166, pi. 48, figs. 16-18. Diagnosis. — A species of Trematocystis with gonopore in FAl and no more than 12 plates in any one circlet. Types. — "H." rotundus Miller, holotype (monotype), FMNH 6004. "H." turbinatus Miller, holotype (monotype), MCZ 79. "H." gorbyi Miller, holotype (monotype), FMNH 6014. "H." commodus Miller, original of Miller, 1891, pi. 3, figs. 1-2, FMNH 6018a here selected as lectotype; original of Miller, 1891, pi. 3, figs. 5-6, FMNH 6018b paralectotype. "H." indianensis Miller, holotype (monotype), FMNH 6016. "H." parvus Miller, holotype (monotype), FMNH 6010. Horizon and locality. — Holotype of "H." rotundus is from Osgood, Ripley County, Indiana. Many other specimens come from Big Creek. The type stratum is assumed to be the same as for T. globosus. Material. — At least 46 complete thecae. AMNH, BMNH, FMNH, MCZ, ROM, UC, USNM. Description. — Theca: Irregularly inverted pyriform to globular, without aboral stem-like projection, with moderate attachment area. Plates with- out pustules, sometimes with slightly impressed sutures. Plates: One generation, up to 19 mm. in type of "H." indi- anensis but generally much less. All smooth or finely granular, with many humatipores. Arranged in circlets of 8-11 plates. Fig. 53. Trematocystis rotundtis (Miller), 1879. a-b, Stereophotos of lecto- type of "H." commodus Miller, FMNH 6018a. a, oral view to show gonopore and hydropore. b, aboral view to show attachment area, c, Stereophotos of paralecto- type of " H." comm4>dus Miller, FMNH 6018b. Detail of humatipores. d. Lateral view of holotype of "H." indianensis Miller, FMNH 6016. e. Oral view of holo- type of "H." gorbyi Miller, FMNH 6014, Figures a-b, d-e, natural size; figure c Xl4. All whitened with ammonium chloride sublimate. 128 FIELDIANA: GEOLOGY, VOLUME 24 A / \. 4 Fig. 54. Trematocystis roturtdus (Miller), 1879. Camera lucida drawings of the oro-anal areas of three examples. A, Typical example with gonopore in FAl. AMNH 23977. B-C, Two unusual examples with FAl developed as two plates. Bi-BMNH E7639 with ;gonopore in FAlb; C, BMNH E7633 with gonopore in FAla. A— anus, g7— gonopore, m — mouth, small numerals indicate peri-oral plates, large numerals indicate facetals, roman numerals indicate ambulacral facets. HUMATIPORES: Buried, not raised into pustules, with 2-7 tan- gential Canals, generally closely packed together, 1.1-1.4 mm. long. Identical to those of T. globosus (fig. 51). Mouth: Rounded oblong, 4-6 mm. across, set within a quadri- lateral peristome. Peristome border pitted with oral pores, 1-13^2 mm. wide, with four straight food grooves from ambulacral facets to mouth border. Palatals unknown, presumably as in T. globosus. IVlouth surrounded by six peri-orals and then eight facetals of which FAl, FA3, FAS, and FAT do not bear facets. Ambulacral facets PAUL: NORTH AMERICAN HOLOCYSTITES FAUNA 129 oval, 5 X 3.5 mm. in large specimens and raised up on facetal plates, with median ridge, slightly raised rim and concave lateral areas. Anus: Rounded, about 5 mm. in diameter, within usually penta- gonal periproct which has ledges for insertion of anals. Anals unknown. GONOPORE: Small, circular, 0.4 mm. in diameter in lectotype of "H." commodus, set within tubercle twice that diameter. In FAl, 3-5 mm. from ambulacral facet I. USNM S3059a, BMNH E7639, and E7633 have FAl represented by two plates. The gonopore is in FAla in the former pair and in FAlb in the other (fig. 54.). Hydropore: Frequently rounded, in tubercle across suture P01:P06. Occasionally slit-like. Attachment: Direct, by moderate attachment area one-fifth to one-third the ambital diameter. Remarks. — USNM S3079b is a small specimen provisionally ac- cepted as T. rotundus. It is atypical in having only seven facetals, very few plates (total about 40), never more than seven plates per circlet and small sparsely packed humatipores with only two to four tangential canals. The humatipores are very similar to those of Pentacystis simplex, and quite dissimilar from those of any other specimen examined. It is quite possible that this speci- men represents a fourth species of Trematocystis. Trematocystis wetherbyi Miller, 1878. Figure 55. Holocystites wetherbyi Miller, 1878, Jour. Cincinnati Soc. Nat. Hist., 1 , p. 131, pi. 5, figs. 2, 2a-b; Bassler and Moodey, 1943, Geol. Soc. Amer. Spec. Pap., 45, pp. 44, 171. Diagnosis.^A species of Trematocystis with gonopore in FAl, with up to 16 plates per circlet. Type. — Holotype (monotype), FMNH 6005. From the Osgood beds, Osgood, Ripley County, Indiana. Material. — The holotype is the only specimen known. Description. — Theca: Inverted pyriform, without aboral stem-like projection, attachment area small. Plates smooth without impressed sutures, numerous. 130 FIELDIANA: GEOLOGY, VOLUME 24 Fig. 55. Trematocystis wetherbyi (Miller), 1878. a-b. Stereophotos of holo- type, FMNH 6005. a. Lateral view to show thecal outline and circjets of small plates, b, Oral view to show small bored pits. Both figures natural size, both whitened with ammonium chloride sublimate. Plates: Of one generation, regularly hexagonal, up to 10 mm. long, arranged in regular circlets, 6P00, 9FAA, 14, 15, 16, 16, 14, 8, plus about six plates which form an incomplete circlet between the last two listed above. All plates bear numerous humatipores. HUMATIPORES: tangential canals. Buried, not developed as pustules, with 3-6 0.8-1.1 mm. long, frequently broader than long. Mouth: Oval, 5 x 3.5 mm., set within quadrilateral peristome. Peristome border up to 1 mm. wide, marked by oral pores and bearing food grooves, only one of which shows clearly. Palatals unknown. Mouth surrounded by six peri-orals and then nine facetals, the extra plate is anterior not posterior as in T. rotundus. Four ambulacral facets on FA2, FA4, FA6, and FA8 oval and poorly preserved, 4 X 23^ mm., approximately. Anus: Rounded, poorly preserved and partly covered by a bryozoan. Approximately 4 mm. in diameter. PAUL: NORTH AMERICAN HOLOCYSTITES FAUNA 131 Gonopore: Small, circular, 0.5 mm. in diameter, in FAl and nearly 4 mm. from edge of ambulacral facet 1. Hydropore: Short slit 0.65 mm. long, set in oval tubercle 1.3 mm. long across suture P01:P06. Attachment: Direct, by a moderate attachment area nearly one-third the ambital diameter. Remarks. — Most specimens of T. rotundus have between 40 and 60 thecal plates, which average 7-8 per circlet and reach a maximum of 11 plates in one circlet. Most specimens of T. glohosus have between 50 and 75 plates which average 8-10 per circlet and reach a maximum of 14 plates in one circlet. The holotype of T. wetherbyi has a total of 104 or possibly 105 plates arranged in one incomplete and eight complete circlets. The average number of plates per circlet is almost 12 and there are five circlets with 14-16 plates each. This specimen appears to be distinct from all the others and so it is accepted as a third species. Sufficient specimens of the other two species are known to indicate that this difference is quite significant. Pustulocystis new genus Type species. — Holocystites ornatissimus Miller, 1891. Diagnosis. — A genus of Holocystitidae with six peri-orals and six facetals, four or five ambulacral facets, plates of one generation, humatipores pustular with up to four tangential canals. Description. — The theca of Pustulocystis is generally pyriform to fusiform with a relatively narrow base of attachment. The theca is small to moderate in size for the fauna, reaching 40 mm. in diam- eter and 50 mm. in length. The peristome is usually subcentral in the prominent oral area which frequently protrudes from the main body of the theca to give the theca a spindle shape. The peristome is quadrate, except in P. pentax where it is pentagonal, and is surrounded by six peri-orals and then six facetals. By analogy with Trematocystis the ambulacral facets are developed on FA2, FA4, FA6, and FA8. FAl and FA5 are also present and hence FA3 and FA7 are the two missing facetals (fig. 58) . The peristome border is generally pitted with oral pores. The ambulacral facets are prominent, usually planar, oval, and have a slight median ridge. There are four ambulacral facets, except in P. pentax n. sp. where there are five. Each facet is developed in a single facetal plate. 132 FIELDIANA: GEOLOGY, VOLUME 24 The periproct may be four or five sided and has a narrow border. Within it is the rounded anus. The gonopore is very variable in position but usually is developed across a suture and associated with POl, FAl, or FA2. The hydropore is usually slit-like and in a tubercle across suture P01:P06. The total number of thecal plates, which are always of one generation, is moderate for the fauna and varies from 70-100. The plates are arranged in definite circlets which may contain up to 20 plates. Usually the fourth or fifth circlet down from the peristome has the largest number of plates. All the plates are covered with prominent tubercles which are more strongly developed than those of Holocystites, except in P. sp., which has finely granular plate ornament. Each tubercle corresponds to one humatipore. The humatipores rarely, if ever, have more than four tangential canals which are not radially arranged like the tangential canals of Holocystites. The humatipore tubercles of Pustulocystis strongly resemble the tubercles of Sinocystis Reed. Attachment was direct by a moderately small attachment area. Some specimens have a rather thick aboral stem-like projection. Paleoecology. — Pustulocystis was attached directly by a small attachment area at the base of the theca and sometimes has a stem- like thecal projection aborally. USNM S3067b, 39520 both at- tached themselves to trilobite cephala over one of the eyes. MU 884 attached to a cylindrical bryozoan colony. In none of these speci- mens is any of the original substrate preserved. The substrate can be recognized because the base of the cystoid has an accurate im- pression of it. The small attachment area suggests that Pustulo- cystis may have been unstable but there is no evidence that these cystoids fell over during life. The mouth is subcentral in the oral surface and neither borings nor epifauna are consistently on one side of the theca. Pustulocystis supported epifauna, like other members of the Holocystites fauna, of which bryozoa, crinoid stem bases, and worm tubes were the most common in that order. Generally, there is less epifauna on species of Pustulocystis than on species of other genera, perhaps because of the extremely uneven thecal surface. There is no evidence that any of the epifauna attached during the life of the cystoids. One bryozoan colony obviously grew after the death of the cystoid to which it is attached because it covers the periproct and the anal plates are missing. PAUL: NORTH AMERICAN HOLOCYSTITES FAUNA 133 The boring organism attacked Pustulocystis as well as other genera. In at least two specimens the hemispherical pits have strongly raised rims which clearly indicates that the cystoids were alive at the time of boring. In the holotype of P. pentax a small pit occurs in ambulacral facet II equally clearly indicating post mortem attack. One specimen has 50 pits, some with rims, dis- tributed all over the theca. Taxonomy. — Pustulocystis is erected for a group of holocystitids with only six facetal plates (fig. 24d), not the usual eight. Five of Miller's "species" can be assigned definitely to Pustulocystis and all five are conspecific as suggested by Bather (1919b, p. 257). They are: "H." ornatissimus Miller, 1891, "H." subovatus Miller, 1891, "H." papulosus Miller, 1891, "H." affinis Miller and Faber, 1892, and "H." asper Miller and Gurley, 1895. A sixth "species," "H." benedicti Miller, 1891, probably also belongs to the same species but as the oral area is missing it cannot be assigned to any one species with certainty. "H." ornatissimus is the oldest name by page priority and is selected as type species since it is typical and common. In addition to P. ornatissimus, two possible new species are described. One differs in having five, not four, ambu- lacral facets; the other in having plates which lack the characteristic pustules. Both are apparently very rare. Pustular species of Pustulocystis might be mistaken for Holocystites scutellatus Hall but they differ in having only one generation of plates and only six facetals. Although typically pustular the humatipores of Pustulo- cystis differ from those of Holocystites s.s. in having fewer tangential canals which are not radially arranged. Key to the species of Ptistulocystis 1. Peristome pentagonal, five ambulacral facets P, pentax. Peristome quadrilateral with four ambulacral facets 2. 2. Plates pustular and finely granular P. ornatissimus. Plates finely granular only P. sp Regional distribution and stratigraphic range. — Southeastern Indiana, West Central Tennessee. Osgood beds. Middle Silurian. Pustulocystis ornatissimus (Miller), 1891. Figures 24d, 56-58. Holocystites ornatissimus Miller, 1891, Rept. Indiana Dept. Geol. Nat. Re- sources, 17, (adv. pub.), p. 17, pi. 5, figs. 1-2; Miller, 1892a, op. cit., p. 627, pi. 5, figs. 1-2; Bather, 1919b, Geol. Mag., 56, p. 257, pi. 6, fig. 25; Bassler and Moodey, 1943, Geol. Soc. Amer. Spec. Pap., 45, pp. 44, 169. Holocystites subovatus Miller, 1891, Rept. Indiana Dept. Geol. Nat. Resources, 17, (adv. pub.), p. 17, pi. 5, figs. 4-5; Miller, 1892a, op. cit., p. 627, pi. 5, 134 PAUL: NORTH AMERICAN HOLOCYSTITES FAUNA 135 figs. 4-5; Miller, 1892c, North American Geology and Palaeontology, 1st app., p. 680, fig. 1237; Foerste, 1931, Ky. Geol. Surv., ser. 6, 35, p. 206, pi. 23, fig. 2a-b; Bassler and Moodey, 1943, Geol. Soc. Amer., Spec. Pap., 45, pp. 44, 170. Holocystites papulosus Miller, 1891, Rept. Indiana Dept. Geol. Nat. Re- sources, 17, (adv. pub.), p. 18, pi. 5, figs. 7-8; Miller, 1892a, op. cit., p. 628, pi. 5, figs. 7-8; Bassler and Moodey, 1943, Geol. Soc. Amer., Spec. Pap., 45, pp. 44, 169. Holocystites affinis Miller and Faber, 1892, Jour. Cincinnati Soc. Nat. Hist., 15, p. 87, pi. 1, figs. 16-18; Bassler and Moodey, 1943, Geol. Soc. Amer. Spec. Pap., 34, pp. 43, 167. Holocystites asper, Miller and Gurley, 1895, Bull. 111. State Mus. Nat. Hist., 7, p. 84, pi. 5, figs. 1-2; Bassler and Moodey, 1943, Geol. Soc. Amer., Spec. Pap., 45, pp. 43, 168. Diagnosis. — A species of Pustulocystis with quadrilateral peri- stome and four ambulacra! facets, pustular plates. Types. — H. ornatissimus Miller, holotype (monotype), FMNH 8760. H. subovatus Miller, holotype (monotype), FMNH 6013. H. papulosus Miller, holotype (monotype), FMNH 6011. H. affinis Miller and Faber, holotype (monotype), FMNH 8761. H. asper Miller and Gurley, holotype (monotype), FMNH 6001. Horizon and locality. — Osgood Formation. Latter two types listed above are from Madison (i.e., North Madison), former three are from "Jefferson County," Indiana. Near North Madison is therefore selected as the type locality. This species is also known from Big Creek and Locality 7 (fig. 11) in Jefferson County, Indiana and from 1 mile East of Peagram, Tennessee (USNM). Material— Fifteen complete thecae. AMNH, FMNH, USNM. Description. — Theca: Irregularly inverted pyriform or fusiform, composed of 50-80 pustular plates arranged in circlets of up to 17. Moderately impressed sutures and small attachment area. Fig. 56. Pustulocystis ornatissimus (Miller), 1891. a-b, Stereophotos of holotype of "Holocystites" ornatissimus Miller, FMNH 8760. a. Lateral view to show thecal outline and pustular plates, b. Oral view, c-d, Stereophotos of holotype of "H." papulosus Miller, FMNH 6011. c, Lateral view, d. Oral view to show prominent oral area with four planar ambulacral facets, e-f, Holotype of "H." subonatus Miller, FMNH 6015. e. Oral view, f, Lateral view, g, Holo- type of "H." affinis Miller and Faber, FMNH 8761. Lateral view to show thecal outline and several bored pits with strongly raised rims. All figures natural size, all whitened with ammonium chloride sublimate. 136 PAUL: NORTH AMERICAN HOLOCYSTITES FAUNA 137 ^ \ \ 5 / B 1 1 Pe \ \ ^ " ~ > V-J^_^ V. ( " 1/ 3 _ _ 4 _ 'f ! ^ 7 jij 6 ^g^^'^"^"^^!^ \r^^^ ' / (■ r)G J 1 -J/?( [ M ] h\* 2 1 1 2 \J mi '.V^P^ . ' >/ -i 1 _^ "^. y^"- ^ \ i.'^ \ ' 1 M 1 ' ; V, 1 ■"V '; V ', w "^ ^ 7 >^^v \ ' T ^- G(| V'/ . FiG^ 58. Pustulocystis ornatissimus (Miller), 1891. Camera lucida drawings of the oral (A) and anal (B) areas of AMNH 20273. A — anus, F — ambulacral facet, G — gonopore, H — hydropore, M — mouth, Pe — peristome, small numerals indicate peri-oral plates, large numerals indicate facetals, roman numerals indicate am- bulacra. Plates: With prominent pustules and fine granulations, of one generation, up to 16 mm. in major dimension but usually smaller. HUMATIFORES: Each developed as a prominent tubercle up to 1.1 mm. high and 0.9 mm. wide at base. Usually with four tan- gential canals but USNM 39520 has seven to eight. Mouth: Round or oval, 4-5 mm. in diameter, set within quadri- lateral peristome with distinct border. Peristome border with few oral pores and narrow food grooves from ambulacral facets to edge of mouth. Palatals unknown. Mouth surrounded by six peri-orals and then six facetals. Four ambulacral facets, each developed entirely within a single facetal plate. FAl and FA5 lack facets. Facets oval or almost circular (even in same individual), 43^ X 3 mm. to 4 X 4 mm., with median ridge. Anus: Rounded pentagonal, 4-5 mm. in diameter, narrow border for insertion of anals defining pentagonal periproct. Anals unknown. Fig. 57 {opposite). Pustulocystis ornatissimus (Miller), 1891. a-b, Stereo- photos of holotype of "Holocystites" asper Miller, FMNH 6001. a, lateral view; b, oral view. Both natural size, c, Stereophotos of holotype of "H." papulosus Miller, FMNH 6011 to show details of humatipores. Note the fine granular oTnament on plate surface. Approx. X 10. All figures whitened with ammonium chloride sublimate. 138 FIELDIANA: GEOLOGY, VOLUME 24 Gdnopore: Small circular 3^2-1 rnm. in diameter, usually de- veloped across a suture and close to ambulacral facet I but position variable. (On suture FA1:FA2, AMNH 20273, FMNH 6013; FA2:P02, FMNH 6011; FA1:P01, FMNH 6001, at junction of FAl, FA2, and POl, USNM; in POl, USNM 35920; and in FA2, USNM). Hydropore: Slit-like, 1 X 0.05 mm., across suture P01:P06. Attachment: Direct, usually by small attachment area one- third to a quarter ambital diameter. Remarks. — This is the commonest species of Pustulocystis. It closely resembles Holocystites scutellatus Hall in thecal shape and ornament but differs in having one generation of plates and only six facetals. The humatipores are developed into large tubercles which are more widely spaced and generally have fewer tangential canals. I have collected P. ornatissimus from grey-green silty rock 11 in. below the base of the Osgood Limestone at locality 7 (fig. 11). It occurs with Trematocystis spp. which are much more common. Pustulocystis pentax new species. Figures 59, 60. Diagnosis. — A species of Pustulocystis with pentagonal peristome and five ambulacral facets; pustular plates. Type. — Holotype, MU 884. From quarry dump, Newsom Sta- tion, Tennessee. Either Osgood or Laurel beds, former more likely. Material. — The holotype is the only specimen known. Description. — Theca: Fusiform with aboral stem-like projection about one- half ambital diameter, (35.5 X 20 X 17 mm.), with pustular plates. Plates: Of one generation, arranged in ten circlets, largest with about 15 humatipores each. Humatipores: Developed as prominent pustules which project up to 0.6 mm. and are 0.6 mm. in diameter at base, with up to four tangential canals. Fig. 59. Pustulocystis pentax n. g. and sp. a-d, Stereophotos of holotype MU 884. a, Posterior lateral view to show thecal outline and humatipores. b, Aboral view to show attachment to cylindrical bryozoan colony, c, Oral view to show five ambulacral facets, pentagonal peristome, hydropore, gonopore, and periproct (cf. fig. 60). d, Left lateral view to show bored pits. All figures natural size, all whitened with ammonium chloride sublimate. 189 140 FIELDIANA: GEOLOGY, VOLUME 24 Mouth: Oval, 3.5 X 2.8 mm., set within pentagonal peristome. Peristome border pitted with oral pores, with five straight food grooves from ambulacral facets to mouth border. Palatals un- known. Mouth surrounded by six peri-orals and then six facetals. Fig. 60. Pustulocystis pentax n. g. and sp. Camera lucida drawing of the oral region of holotype MU 884. A — anus, G — gonopore, H — hydropore, M — mouth, roman numerals indicate ambulacra. Note in this species FAl is the only facetal plate without an ambulacral facet. Shaded areas indicate bored pits. Five ambulacral facets each entirely within a single plate. FAl does not bear a facet. Facets oval, 3x2 mm. with slight rim and central ridge, lateral areas concave. Anus: Rounded, set within an almost perfectly square periproct with narrow border. Anals unknown. Gonopore: Small, circular pore across suture FA1:FA2 close to junction with POl. Hydropore: Slit-like, across suture P01:P06. Attachment: Direct, to a cylindrical bryozoan colony. Remarks. — The only characters setting P. pentax apart from P. ornatissimus are the five ambulacral facets and the markedly fusi- form thecal shape. Apparently, the distinction is valid. None of the specimens of P. ornatissimus even closely approaches the elongate fusiform shape of P. pentax and all are constant in the details of the oral area. This, coupled with the fact that the two forms occur about 200 miles apart, seems to warrant the distinction of a separate species. PAUL: NORTH AMERICAN HOLOCYSTITES FAUNA 141 Fig. 61. PustulocysHs sp. a-b, Stereophotos of USNM S3067b. a, pos- terior lateral view to show thecal outline and smooth thecal plates. Traces of buried humatipores can be seen to the left aborally. b, oral view. Both figures natural size, both whitened with ammonium chloride sublimate. Pustulocystis sp. Figure 61. Description. — Theca: Inflated fusiform, composed of about 50 plates which are finely granular not pustular and do not have impressed sutures. Aboral stem-like projection present. Plates: Of one generation, arranged in seven circlets (6P00, 7FAA, ?8, ?8, ?8, 7 or 8, 5). All with humatipores. Humatipores: Buried within plates not pustular, with up to four tangential canals. Mouth: Oval, set within a quadrilateral peristome 6X4 mm. Palatals unknown. Mouth surrounded by six peri-orals, then seven facetals. Four oval ambulacral facets, 2.5 X 3.5 mm., each within a single plate, FAl, FA3, and FA5 do not bear facets. Anus: Rounded, 5 mm. in diameter, within pentagonal periproct which originally had two anals in place but they have been scraped off by a previous worker. 142 FIELDIANA: GEOLOGY, VOLUME 24 Gonopore: Small, circular in FA2. Hydropore: Unusually long, 2 mm,, slit-like, across suture P01:P06. Attachment: Direct, to a trilobite cephalon over one eye. Remarks. — USNM S3067b resembles a Pustulocystis in having a protruding peristome with four prominent, flat ambulacral facets but differs in lacking the characteristic pustular humatipores. The humatipores are buried and typical of Trematocystis. The specimen is poorly preserved but appears to have seven facetal plates (FA7 is definitely not developed), which precludes definite assignment to either genus. One other specimen has been seen which has seven facetal plates but this specimen can clearly be assigned to Tremato- cystis on the basis of humatipores and details of the oral area. USNM S3067b resembles a Pustulocystis in general shape and so is tentatively described as a species of that genus, but in the absence of any other specimens, it is not named. Brightonicystis new genus Type species. — Brightonicystis gregarius n.g. and sp. Diagnosis. — A genus of Holocystitidae with globular or oval theca, smooth, tumid, irregular plates of more than one generation, ten peri-oral plates surrounding the mouth. Description: See under B. gregarius sp. Paleoecology. — B. gregarius was attached directly to a solid substrate by a substantial attachment area between one-quarter and one-third the ambital diameter. SM A32813 and A32811 both have bases of other Brightonicystis specimens attached to them, hence the specific name gregarius. Other substrata are unknown. Presumably the Swedish specimens were also attached directly but no example has the base of the theca preserved. If attachment was direct, the attachment area was considerably smaller than in B. gregarius. B. gregarius acted as substrate for crinoids, worms, and bryozoa, as well as other B. gregarius. Evidence as to whether epifauna was attached during life or not is inconclusive. In at least one case epi- faunal B. gregarius became attached after the death of the specimen it grew on as it partially covers the periproct and the anal plates are missing. No specimens of either species have any borings in them. PAUL: NORTH AMERICAN HOLOCYSTITES FAUNA 143 Geographic distribution and stratigraphic range. — Northern Eng- land, Vastergotland, Sweden. Upper Ordovician. Brightonicystis gregarius new species. Figures 62-64. Trematocystis sp. Ingham, 1966, Proc. Yorks Geol. Soc, 35, p. 502; Paul, 1967b, Proc. Geol. Soc. London, (1967), p. 147. Diagnosis. — A species of Brightonicystis with globular theca, two generations of tumid plates, up to nine or ten tangential canals per humatipore. jTi/pe.— Holotype, SM A32813. From the Ashgill Shales in Swindale Beck just upstream of the junction with Great Rundale Beck, near Knock, Westmoreland, England (G.R. NY 688274). Horizon and locality. — Ashgill Shales (Ashgill, zone 8), Knock, Westmoreland (Paul, 1967b), Cystoid Limestone (Ashgill, zone 8 Murthwaite, Westerdale, and Taithes inliers, Cautley District, Yorks, England (Ingham, 1966). Material. — Holotype, three paratypes and several fragments. SM. Description. — Theca: Small (20-35 mm.), globular or oval, composed of 100- 150 smooth slightly tumid plates with slightly impressed sutures. Mouth large, opposite large attachment area. Plates: Of two generations, primaries arranged in rough circlets, secondaries small triangular plates. Plates of holotype up to 7 mm., all plates with many humatipores. HUMATIPORES: Well developed, with up to 9 or 10 tangential canals (fig. 63). Mouth: Oval, 5 mm. in major diameter, set within pentagonal peristome. Peristome border about 1.5 mm. wide and pitted by oral pores. Narrow food grooves pass from facets to edge of mouth. Palatals unknown. Mouth surrounded by ten peri-orals; facetals not recognized. Five oval ambulacral facets, one per radius except that in SM A32812 there are two facets in radius I (fig. 62h) and the two food grooves unite before reaching the edge of the mouth. Facets 2 X 1.7 mm., developed on two or three thecal plates, each facet with two shallow depressions on either side of median ridge. Anus: Rounded, 4 mm. in diameter in holotype, surrounded by rim (periproct border) for insertion of anals which are unknown. 144 PAUL: NORTH AMERICAN HOLOCYSTITES FAUNA 145 Fig. 63. Brightonicystis gregarius n. g. and sp. Camera lucida drawing of humatipores of paratype, SM A32814a. Xl5. Gonopore: Round, 0.4 mm. in diameter, very close to periproct, almost an embayment on adoral side of latter (fig. 64) . Hydropore: Not certainly detected. Suture between posterior pair of peri-orals irregular, possibly it functioned as a hydropore. Attachment: Direct, large attachment area, one- third ambital diameter in holotype. Fig. 62. Brightonicystis gregarius n. g. and sp. a-d, Stereophotos of holo- type, SM A32813. a, right lateral view to show thecal outline, b, anterior lateral view, c, aboral viewt o show attachment area pierced by thecal canals which have been weathered, d, oral view to show peristome, periproct, gonopore, ambulacral facets, and base of another specimen attached to oral surface, e-g, Stereophotos of paratype, SM A32811. e, right lateral view to show thecal out- line and attached base of another specimen, f, oral view to show peristome with oral pores, g, anterior lateral view with small epifaunal bryozoan colony and worm tube, h, Paratype, SM A32812. Stereophotos of oral area to show pair of ambulacral facets developed in ambulacrum I. i, Paratype, SM A32814a. Detail of humatipores to show large number of tangential canals per humatipore. j, Paratype SM A32811. Detail of oral area to show oral pores, a-c, e-h, natural size; d, j, X2; i, X5. All whitened with ammonium chloride sublimate except i, which was photographed under water. 146 FIELDIANA: GEOLOGY, VOLUME 24 Remarks. — Brightonicystis gregarius is quite certainly a holo- cystitid since it has well developed humatipores. However, it differs from all other holocystids in which the details of the oral area are known by having ten plates directly surrounding the Fig. 64. Brightonicystis gregarius n. g. and sp. Camera lucida drawing of the anal region of holotype, SM A32813. a — anus, f — ambulacral facet, g — gono- pore, m — mouth, stippled area is covered by base of another cystoid. mouth. These plates have oral pores developed in the peristome border and on this evidence are thought to be homologous with the peri-oral plates of other holocystitids. The facetal circlet cannot be recognized as such. In Brightonicystis the arrangement of adoral plates is much less regular than in later holocystitids. Brightoni- cystis gregarius, and the Swedish specimens, are the oldest known holocystitids. Nevertheless, both forms have well-developed hu- matipores. The holotype of B. gregarius has the aboral portion of the theca quite deeply weathered and shows that even plates forming the attachment area are penetrated by many thecal canals. This can be confirmed from the bases of the two specimens which grew epizoically on the holotype and SM A32811. In both these speci- mens the internal openings of the canals can be seen in the plates which form the attachment area. In some other specimens there is evidence that plate material beneath the ambulacral facets is also pierced by thecal canals. It seems that all plates were pierced by many thecal canals but functional pore-structures could only exist in areas exposed to the ambient sea water. Aristocystites and Sphaeronites also have all thecal plates pierced by canals and again diplopores were only developed where the thecal surface was not permanently in contact with any other surface. PAUL: NORTH AMERICAN HOLOCYSTITES FAUNA 147 Brightonicystis sp. Figure 65. Megacystis alternata Hall var. Angelin, 1878, Iconographia crinoideorum in stratis sueciae siluricis fossilium, p. 30, pi. 19, figs. 19, 19a, 20. Holocystitesi sp. Regnell, 1945, Meddn Lunds geol-miner. Instn., 108, p, 183, pi. 15, figs. 2-4; Regnell, 1948, Ark. Kemi Miner. Geol., 26, no. 13, p. 38. Material. — Three partially complete thecae and three fragments, all in Naturhistoriska Riksmuseum. From the grey limestone in the Dalmanites series, Kinnekulle, Vastergotland, Sweden. Description. — All three thecae are apparently inverted pyriform although RM Ec2738 is somewhat irregular. All are composed of irregular, smooth slightly tumid plates which are of at least three generations in RM Ec2738. Incomplete circlets of primary plates are detectable and the secondaries and tertiaries are intercalated in an irregular manner. The theca reached at least 46 mm. in height and 30 mm. in diameter in RM Ec3077 which is incomplete. None of the specimens shows the oral area and all four thecal orifices are unknown. The plates are smooth or occasionally have smooth surface irregularities. They reach a maximum size of 9-10 mm. in RM Ec3077. Humatipores can be detected on RM Ec2738, Ec3077, and Ec2740. They reach 1 mm. in length and generally have 2-4 tangential canals. Remarks. — As pointed out by Regnell (1945, p. 183), the presence of humatipores confirms beyond doubt that these incomplete speci- mens are holocystitids. They are here tentatively assigned to Brightonicystis primarily on the grounds that this is the only known Ordovician genus. Until the oral area becomes known, this assign- ment must remain tentative. It is not certain whether all these forms are conspecific. In particular RM Ec2738 is somewhat different from the other frag- ments. This and the lack of an adequate type specimen precludes description of a species at this point. If these specimens do belong to Brightonicystis, it is clear that they represent a second species. They differ from B. gregarius in thecal shape and size and in the number of tangential canals in the humatipores. Unrecognized Species The status of the following species are not recognizable with certainty either because the type specimens have been lost or be- cause types preserve insufficient information to make specific or 148 PAUL: NORTH AMERICAN HOLOCYSTITES FAUNA 149 even generic assignments. In the latter case the specific name is restricted to the type specimen. A brief description and inter- pretation of each species follows; those with lost type specimens are dealt with first. Holocystites ovatus Hall, 1864. Holocystites ovatus Hall, 1864, Rept. N. Y. State Cab. Nat. Hist., 18, (adv. pub.), p. 9; Hall, 1868, op. cit., 20, p. 313, pi. 12, fig. 2; Hall, 1870, op. cit., 20, (2nd rev. ed.), p. 356, pi. 12, fig. 2; Chamberlin, 1883 ^Geol. Wise, p. 191, fig. 51e; Bassler and Moodey, 1943, Geol. Soc. Amer., Spec. Pap., 45, pp. 50, 169. Type. — Hall's type was originally deposited in the American Museum of Natural History along with his other "Holocystites" types. Two separate searches failed to locate this type which must be presumed lost — hopefully not permanently. Description and interpretation. — Hall's original description and figure indicate an ovoid, perhaps originally pyriform, cystoid with more than ten circlets of plates. The circlets contain more than eight plates, perhaps as many as 14, which are apparently of one generation. Little can be said about the specimen without seeing it, but if it does have one generation of plates it is probably not conspecific with H. scutellatus Hall. Since the arrangement of plates around the mouth and the nature of the pore-structures are unknown, even generic assignment cannot be made. "In the limestone of the Niagara Group at Waukesha, Wiscon- sin." (Hall, 1864, p. 10.) "Holocystites" brauni Miller, 1878. Holocystites brauni Miller, 1878, Jour. Cincinnati Soc. Nat. Hist., 1, p. 130, pi. 5, fig. 1; Bassler and Moodey, 1943, Geol. Soc. Amer., Spec. Pap., 45, pp. 43, 168. Type. — In his later years Miller is said to have traded away much of his personal collection in a piecemeal fashion. Charles Fig. 65. Brightonicystis sp. a-b, Stereophotos of RM Ec3077. Two lateral views to show thecal outline and large thecal plates. Both figures natural size, both whitened with ammonium chloride sublimate. c-d, Stereophotos of RM Ec278i. c, details of humatipores, X 7 photographed under water, d, lat- eral view to show thecal shape and small thecal plates. Natural size, whitened with ammonium .sublimate. 150 FIELDIANA: GEOLOGY, VOLUME 24 Faber is credited with sufficient business acumen to have acquired type specimens in preference to others. Thus most of Miller's types found their way into Faber's collection which Faber sub- sequently sold to the Walker Museum, University of Chicago, and which is now housed in Field Museum. The types of "H." brauni and the following three species are not in Faber's collection and all inquiries as to their whereabouts have drawn blanks. Pos- sibly someone other than Faber bought these specimens which are effectively lost. Description and interpretation. — From the size of this specimen alone (3 in. long and 1}4 in. wide) it must belong to Triamara. This is clearly the oldest name for one species of Triamara, however, specific assignment cannot be made in the absence of information on pore-densities and plate ornament. Holocystites ornatus Miller, 1878. Holocyslites ornatus Miller, 1878, Jour. Cincinnati Soc. Nat. Hist., 1, p. 132, pi. 5, figs. 3, 3a-b; Bassler and Moodey, 1943, Geol. Soc. Amer., Spec. Pap., 45, pp. 44, 169. Type. — See "H." brauni. Description and interpretation. — Miller's description and figures indicate a pyriform theca composed of pustular plates of more than one generation. There are said to be four or possibly five ambulacral facets and the gonopore (Miller's "anal opening") is in FAl. This description best fits Holocystites scutellatus Hall which has two plate generations, five ambulacral facets, and a pyriform theca. The specimen cannot be assigned to Pustulocystis, all known specimens of which have one generation of plates. Holocystites plenus Miller, 1878. Holocystites plenus Miller, 1878, Jour. Cincinnati Soc. Nat. Hist., 1, p. 135, pi. 6, figs. 2, 2a; Bassler and Moodey, 1943, Geol. Soc. Amer., Spec. Pap., 45, pp. 44, 170. Type. — See "H." brauni. Description and interpretation.— Th.\^ specimen was apparently the upper portion of a fat cylindrical cystoid with the mouth offset to one side. There are five ambulacral facets and apparently all the other features of Holocystites s.s. are present. Miller states the plate surface is "slightly granulose" which would suggest H. spangleri. It is equally possible that the specimen is a worn H. alternatus. PAUL: NORTH AMERICAN HOLOCYSTITES FAUNA 151 "Holocystites" elegans Miller, 1878. Holocystites elegans Miller, 1878, Jour. Cincinnati Soc. Nat. Hist., 1, p. 136, pi. 6, figs. 3, 3a; Bassler and Moodey, 1943, Geol. Soc. Amer., Spec. Pap., 45, pp. 44, 168. Type.—^ee "H." brauni. Description and interpretation. — This specimen is clearly a species of Pentacystis and is the oldest available name for that species. It is ovate-pyriform, with eight plates directly surrounding the mouth, and with five ambulacral facets. The characters of P. elegans cannot be determined in the absence of the holotype and hence the name cannot be used. This and the preceding three species were all collected by Fred- erick Braun in "the lower part of the Niagara Group, in Jefferson County, Indiana." (Miller, 1878, p. 136.) "Holocystites" winchelli Hall, 1864. Figure 66. Holocystites winchelli Hail, 1864, Kept. N. Y. State Cab. Nat Hist., 18, (adv. pub.), p. 9; Hall, 1868, op. cit., 20, p. 313, pi. 12, fig. 3; Hall, 1870, op. cit., 20, (2nd rev. ed.), p. 356, pi. 12, fig. 3; Bassler and Moodey, 1943, Geol. Soc. Amer., Spec. Pap., 45, pp. 50, 171. Type.— Uolotyve, AMNH 2017. . . Fig. 66. "Holocystites" mnchelli Hall, 1864. a-b, Holotype, AMNH 2017. a, stereophotos of oral view, b, lateral view to show thecal outline. Both fig- ures natural size, both whitened with ammonium chloride sublimate. 152 FIELDIANA: GEOLOGY, VOLUME 24 Fig. 67. " Holocystites" sphaericus Winchell and Marcy, 1879. Gutta percha impression of holotype, UMMP 23707. Natural size, whitened with ammonium chloride sublimate. Description and interpretation. — The holotype is an internal mold of a pyriform theca with many plates, all of which bear pore-struc- tures. Vague circlets of plates can be distinguished with many more than eight plates per circlet. The arrangement of plates is irregular and more precise details cannot be made out. There are no cystoids from Southeastern Indiana which even vaguely resemble this specimen. For the present it is wisest to restrict the name to the holotype. "In the limestone (Niagara Group) at Waukesha, Wisconsin." Hall, 1864, p. 9. "Holocystites" sphaericus Winchell and Marcy, 1865. Figure 67, Holocystites sphaericus Winchell and Marcy, 1865, Mem. Boston Soc. Nat. Hist., 1, p. Ill; Bassler and Moodey, 1943, Geol. Soc. Amer., Spec. Pap., 45, pp. 50, 170. Type. — Holotype was originally deposited in the Winchell col- lection (No. 2480) at Alma, College, Alma, Michigan. A gutta percha impression of the holotype was made and deposited at the University of Michigan, Museum of Paleontology (UMMP 23707). All the Winchell collection was subsequently transferred to the University of Michigan but apparently this specimen did not get to its destination. As can be seen from Figure 67 the gutta percha impression has been broken and glued together again. PAUL: NORTH AMERICAN HOLOCYSTITES FAUNA 153 Description and interpretation. — Either the original external mold was very poorly preserved or the gutta percha impression was badly made but the net result is that preserved details are just not sufficient to say anything about the species. It was apparently roughly globular and covered with irregularly arranged plates which have impressed sutures and are smooth. As none of the oral area or at- tachment area is preserved, it is impossible to orient the specimen. Even if the holotype should be located it seems likely that this species will ever be recognizable. It is an echinodern but beyond that anybody's guess is valid. From the Niagara Dolomite of the Bridgeport quarry, Chicago. "Holocystites" subrotundus Miller, 1879. Figure 68. Holocystites subrotundus Miller, 1879, Jour. Cincinnati Soc. Nat. Hist., 2, p. 107, pi. 9, figs. 2, 2a; Bassler and Moodey, 1943, Geol. Soc. Amer., Spec. Pap., 45, pp. 44, 170. Type.— FMNR 6003. Description and interpretation. — The theca is globular, 43 mm. in diameter, and has smooth plates of one generation arranged in circlets of up to 13 plates each. The attachment area is preserved but the oral surface is missing. As stated previously (pp. 104, 111), this specimen is almost certainly a Pentacystis, but without the oral surface this cannot be confirmed. Fig. 68. "Holocystites" subrotundus Miller, 1879. Aboral view of holotype, FMNH 6003, natural size, whitened with ammonium chloride sublimate. 154 FIELDIANA: GEOLOGY, VOLUME 24 Fig. 69. " Holocystites" madisonensis Miller, 1891. Aboral view of holotype FMNH 6017. Natural size, whitened with ammonium chloride sublimate. From "the lower part of the Niagara group, in Ripley County, Indiana," Miller 1879, p. 108. "Holocystites" madisonensis Miller, 1891. Figure 69. Holocystites madisonensis Miller, 1891, Rept. Indiana Dept. Geol. Nat. Resources, 17, (adv. pub.), p. 15, pi. 3, figs. 3-4; Miller, 1892a, op. cit., p. 625, pi. 3, figs. 3-4; Bassler and Moodey, 1943, Geol. Soc. Amer., Spec. Pap., 45, pp. 44, 169. Type.— FM^B. 6017. Description and interpretation. — The theca is globular, 53 mm. in diameter, and has smooth plates apparently of one generation, circlets of up to 18 plates each. The attachment area is preserved and is large (22 mm. in diameter) and smooth. The oral area is missing and again, although it is almost certain that H. madisonensis is a specimen of Pentacystis, this cannot be confirmed, nor can the species be identified, "... in the lower part of the Niagara Group, in Jefferson County, Indiana," Miller, 1891, p. 15. "Holocystites" benedicti Miller, 1891. Figure 70. Holocystites benedicti Miller, 1891, Rept. Indiana Dept. Geol. Nat. Resources, 17, radv. pub.), p. 17, pi. 5, fig. 3; Miller, 1892a, op. cit. p. 627, pi. 5. fig. 3; Bassler and Moodey, 1943, Geol. Soc. Amer., Spec. Pap., 45, pp, 43, 168. PAUL: NORTH AMERICAN HOLOCYSTITES FAUNA 155 Fig. 70. " Holocystites" benedicti Miller, 1891. Lateral view of holotype, FMNH 6015. Natural size, whitened with ammonium chloride sublimate. Type.— FMNR 6015. Description and interpretation. — The theca is pyriform, composed of small pustular plates of one generation arranged in circlets of up to 16 plates. The oral surface is missing so that the precise nature of the specimen cannot be determined. However, from the plate arrangement and pustular humatipores it is certain that this is a Pustulocystis and most probably P. ornatissimus. There is the small- est chance that it might be P. pentax. "Found in the Niagara Group, of Jefferson County, Indiana . . ." Miller, 1891, p. 17. REFERENCES Angelin, N. p. 1878. Iconographia crinoideorium in stratis sueciae siluricia fossilium. Holmiae. IV + 32 pp., 29 pis. Barrande, J. 1887. Systeme Silurien du centre de la Boheme. 1® partie: Recherches paleon- tologiques. Vol. 7. Classe des Echinodermes. Ordre des Cystides. Leipzig and Prague. XVII + 233 pp., 39 pis, Bassler R. S. and M. W. Moodey 1943. Bibliographic and faunal index of Paleozoic pelmatozoan echinoderms. Geol. Soc. Amer., Spec. Pap., 45, VI + 733 pp. Bather, F. A. 1900. The echinoderms. In Lankester, E. R., A Treatise on Zoology, 3, VI + 344 pp., illust. London. 1919a. Notes on Yunnan Cystidea. 3 Sinocystis compared with similar genera. Geol. Mag., 56, pp. 110-115, pi. 3, figs. 14-21. 1919b. Notes on Yunnan Cystidea. 3 Sinocystis compared with similar genera, Geol. Mag., 56, pp. 255-262, pi. 6, figs. 22-30. 1919c. Notes on Yunnan Cystidea. 3 Sinocystis compared with similar genera' Geol. Mag., 56, pp. 318-325, fig. 31. Billings, E. 1870. Notes on the structure of the Crinoidea, Cystoidea and Blastoidea. Ann. Mag. Nat. Hist., ser. 4, 5, pp. 251-266, 15 figs., 409-416, 7 figs. Bolton, T. E. 1957. Silurian stratigraphy and palaeontology of the Niagara Escarpment in Ontario. Mem. Geol. Surv. Branch Canad., 289, 145 pp., 13 pis., 2 figs. Boyce, M. W. 1956. The Macropaleontology of the Osgood Formation (Niagaran) in Ripley County, Indiana. Unpublished Ph.D. thesis, Indiana University. Browne, R. G. and Schott, V. J. 1963. Arenaceous Foraminifera from the Osgood Formation at Osgood, Indiana. Bull. Amer. Paleontol., 46, pp. 189-242, pis. 48-52, 1 fig. Browne, W. R. (Brown, I. A.) 1964. A new cystoid (Pelmatozoa, Echinodermata) from the Silurian of New South Wales. Proc. Linn. Soc. N. S. W., 88, pp. 386-391, pi. 21, 2 figs. BUCH, L. VON 1846. tlber Cystideen eingeleitet durch die Entwicklung der Eigenthumlich- keiten von Caryocrinus ornatus Say. Abhandl K. preuss. Akad. Wiss., (1844), pp. 89-116, pis. 1-2. (see also English translation Quart Jour. Geol. Soc. London, 2, pp. 20-42, pis. 3-4.) 156 PAUL: NORTH AMERICAN HOLOCYSTITES FAUNA 157 BuscH, D. A. 1943. Some unusual cystoids and crinoids from the Niagaran (Silurian) of West-central Ohio. Jour. Paleontol., 17, pp. 105-109, pi. 20, 1 fig. Butts, C. 1915. Geology and mineral resources of Jefferson County, Kentucky. Ky. Geol. Surv., Ser. 4, 3, XIV + 270 pp. 65 pis. Carpenter, P. H. 1891. On certain points in the morphology of the Cystidea. Jour. Linn. Soc. (Zool.), 24, pp. 1-52, pi. 1, 3 figs. Chamberlin, T. C. 1883. General geology; I chemical geology, II lithological geology, III historical geology. In Geology of Wisconsin, part 1, pp. 1-300, figs. 1-102. Madison, Wise. Chauvel, J. 1941. Recherches sur les Cystoides et les Carpoides Armoricaines. M6m. Soc. G6ol. Min6r. Bretagne, 5, 286 pp., 7 pis., 101 figs. 1966. Echinodermes de I'Ordovicien du Maroc. Cah. Paleontol., Paris. 120 pp., 16 pis. Fisher, D. J. 1925. Geology and mineral resources of the Joliet Quadrangle. Bull. 111. State, Geol. Surv., 51, 160 pp., 6 pis., 47 figs. Foerste, a. F. 1897. A report on the geology of the Middle and Upper Silurian rocks of Clark, Jefferson, Ripley, Jennings and Southern Decatur Counties, Indiana. Rept. Indiana Dept. Geol. Nat. Resources, 21, pp. 213-288, pis. 14-17, illust. 1904. The Ordovician-Silurian contact in the Ripley Island area of Southern Indiana, with notes on the age of the Cincinnati geanticline. Amer. Jour. Sci., ser. 4, 18, pp. 321-342, pi. 17, illust. 1917. Notes on Silurian fossils from Ohio and other central states. Ohio Jour. Sci., 17, pp. 187-204, 233-267, pis. 8-12, figs. 1-3. 1920a. Intercalation of thecal plates in Holocystites in connection with the criteria upon which species can be distinguished. Bull. Geol. Soc. Amer., 31, pp. 207-208 (Abstr.). 1920b. Racine and Cedarville cystoids and blastoids with notes on other echinoderms. Ohio Jour. Sci., 21, pp. 33-78, pis. 1-4, 4 figs. 1931. Silurian fauna. In Jillson, W. R. (Compiler), The Palaeontology of Kentucky. Ky. Geol. Surv., ser. 6, 36, pp. 169-213, pis. 17-26. Grabau, a. W. and H. W. Shimer 1910. North American index fossils. Invertebrates. II Conularida-Echinoidea, appendices. New York. XIV + 909 pp., figs. 1,211-1,937. Haeckel, E. 1896. Amphorideen und Cystoideen. Beitrage zur Morphologic und Phylo- genie. In Festschrift zum Siebenzigstengeburtstage von Carl Gegenbaur am 21 August 1896. 1, pp. 1-180, pis. 1-5, 25 figs. Hall, J. 1861. Descriptions of new species of fossils; from the investigations of the survey. In Report of the superintendent of the geological survey [of Wiscon- sin] exhibiting the progress of the work, Madison, Wise, pp. 9-52. 1862. Palaeontology of Wisconsin. In Hall, J. and J. D. Whitney, Report on the geological survey of the State of Wisconsin, pp. 425-442, illus. 158 FIELDIANA: GEOLOGY, VOLUME 24 1864. Account of some new or little known species of fossils from rocks of the age of the Niagara Group. Albany, New York. 16 pp., 1 pi. (Advanced publication of the 18th Rept. N.Y. State Cab. Nat. Hist.) 1868. Account of some new or little known species of fossils from rocks of the age of the Niagara Group. Rept. N. Y. State Cab. Nat. Hist., 20, pp. 305- 401, pis. 10-23, 16 figs. 1870. Descriptions of new or little known species of fossils from rocks of the age of the Niagara Group. 2nd (rev.) edition. Rept. N. Y. State Cab. Nat. Hist., 20, pp. 347-438, pis. 10-25, 16 figs. Hubbard, G. C. 1892. The cystideans of Jefferson County, Ind. Proc. Ind. Acad. Sci. (1891), p. 67. Huxley, T. H. (translator) 1854. Bj/ Prof . J. Miiller. On the structure of the echinoderms. Ann. Mag. Nat. Hist., ser. 2, 13, pp. 1-24, 112-123, 241-256. (see Miiller, J. below.) Ingham, J. K. 1966. The Ordovician rocks in the Cautley and Dent Districts of Westmore- land and Yorkshire. Proc. Yorks. Geol. Soc, 35, pp. 455-505, pis. 25-28, 8 figs. Jaekel, O. 1899. Stammesgeschichte der Pelmatozoen. 1. Thecoidea und Cystoidea. Berlin. X + 422 pp., 18 pis., 88 figs. Kesling, R. V. 1963. Key for the classification of cystoids. Contr. Mus. Paleontol. Univ. Mich., 18, pp. 101-116. 1968. Cystoids. In Moore, R. C, ed., Treatise on Invertebrate Paleontology. S. Echinodermata 1. pp. S85-S267, figs. 32-157. Kindle, E. M. and V. H. Barnett 1909. The stratigraphic and faunal relations of the Waldron Fauna in Southern Indiana. Rept. Ind. Dept. Geol. Nat, Resources, 33, pp. 395-416. La Rocque, a. and M. F. Marple 1955. Ohio Fossils. Bull. Geol. Surv. Ohio, 54, 152 pp., 413 figs. Lonsdale, W. 1849. Notes on fossil Zoophytes found in the deposits described by Dr. Fitton in his memoir entitled "A stratigraphical account of the section Atherfield to Rocken End." Quart. Jour. Geol. Soc, London, 5, pp. 55-103, pis. 4-5. Miller, S. A. 1878. Descriptions of eight new species of Holocystites from the Niagara group. Jour. Cincinnati Soc. Nat. Hist., 1, pp. 129-136, pis. 5-6. 1879. Description of twelve new fossil species, and remarks upon others. Jour. Cincinnati Soc. Nat. Hist., 2, pp. 104-118, pis. 9-10. 1880. Description of two new species from the Niagara group, and five from the Keokuk group. Jour. Cincinnati Soc. Nat. Hist., 2, pp. 254-259, pi. 15. 1882. Description of three new orders and four new families, in the class Echinodermata, and eight new species from the Silurian and Devonian formations. Jour. Cincinnati Soc. Nat. Hist., 5, pp. 221-231, pi. 9. 1889. North American geology and palaeontology. Cincinnati, Ohio. 664 pp., 1,194 figs. 1891. Advance sheets from the seventeenth report of the geological survey of Indiana. Palaeontology. Indianapolis. 94 pp., 20 pis. (This publication not seen by author.) PAUL: NORTH AMERICAN HOLOCYSTITES FAUNA 159 1892a. Palaeontology. Rept. Ind. Dept. Geol. Nat. Resources, 17, pp. 611-705, 20 pis. 1892b. Advance sheets from the eighteenth report of the geological survey of Indiana. Palaeontology. Indianapolis. 79 pp. 12 pis. 1892c. First Appendix (to North American geology and palaeontology), pp. 665-718, figs. 1,195-1,265. 1894. Palaeontology. Rept. Ind. Dept. Geol. Nat. Resources, 18, pp. 257- 356, 12 pis. 1897. Second Appendix (to North American geology and palaeontology), pp. 719-793, figs. 1,266-1,458. Miller, S. A. and C. Faber 1892. Some new species and new structural parts of fossils. Jour. Cincinnati Soc. Nat. Hist., 15, pp. 79-87, 1 pis. Miller, S. A. and W. F. E. Gurley 1894. New genera and species of Echinodermata. Bull. 111. St. Mus. Nat. Hist., 5, 53 pp., 5 pis. 1895. New and interesting species of Palaeozoic fossils. Bull. 111. St. Mus. Nat. Hist., 7, 89 pp., 5 pis. Mound, M. C. 1968. Arenaceous Foraminiferida and zonation of the Silurian rocks of Northern Indiana. Bull. Ind. Dept. Nat. Resources Geol. Surv., 38, 126 pp., 7 pis., 6 figs. Muller, a. H. 1963. Lehrbuch der Palaozoologie. Band II Invertebraten. Teil 3 Arthropoda 2— Stomochorda. Jena. XVIII + 698 pp., 854 figs. Muller, J. 1854. Uber den Bau der Echinodermen. Abhandl. K. preuss. Akad. Wiss. (1854), pp. 123-219, pis. 1-9. (see Huxley, T. H. above). Neumayr, M. 1889. Die Stamme des Thierreiches. Wirbellose Thiere. Wien & Prag. VI + 603 pp., 192 figs. NiCOLL, R. S. and C. B. Rexroad 1968. Stratigraphy and conodont paleontology of the Salomonie Dolomite Lee Creek Member of the Brassfield Limestone (Silurian) in Southeastern Indiana and adjacent Kentucky. Bull. Ind. Dept. Nat. Resources Geol. Surv., 40, 73 pp., 7 pis., 4 figs. Nicholson, H. A. and R. Lydekker 1889. A manual of palaeontology for the use of students with a general intro- duction on the principles of palaeontology, vol. 1, 3rd (rev.) ed. London and Edinburgh. VII + 885 pp., 812 figs. Parks, W. A. 1913. Excursion B3. The Palaeozoic section at Hamilton, Ontario. Canad. Geol. Surv. Guidebook, 4, pp. 125-140, illust. Paul, C. R. C. 1967a. The British Silurian Cystoids. Bull. Brit. Mus. Nat. Hist. (Geol.), 13, pp. 297-356, 10 pis., 44 figs. 1967b. New records of cystoids from Britain. Proc. Geol. Soc. London, (1967), p. 147. 1967c. The functional morphology and mode of life of the cystoid Pleurocystites, E. Billings, 1854. Symp. Zool. Soc. London, 20, pp. 105-121, 22 figs. 160 FIELDIANA: GEOLOGY, VOLUME 24 1968. The morphology and function of dichoporite pore-structures in cystoids. Palaeontology, 11, pp. 697-730, pis. 134-140, 37 figs. Paul, C. R. C. and R. V. Kesling 1968. Review of J. Chauvel, "Echinodermes de I'Ordovicien du Maroc." Jour. Paleontol., 42, pp. 245-247. PiNSAK, A. P. and Shaver, R. H. 1964. The Silurian formations of Northern Indiana. Bull. Ind. Dept. Conserv. Geol. Surv., 32, 87 pp., 2 pis., 6 figs. Priddy, R. R. 1939. A petrographic study of the Niagaran rocks of Southwestern Ohio and Southeastern Indiana. Jour. Geol., 47, pp. 489-502, 3 figs. Prokop, R. J. 1964. Sphaeronitoidea Neumayr of the Lower Paleozoic of Bohemia (Cystoidea, Diploporita). Shorn. Geol. Paleontol., 15, pp. 7-37, pis. 1-8, 14 figs. REGNfiLL, G. 1945. Non-crinoid Pelmatozoa from the Paleozoic of Sweden. A taxonomic study. Meddn Lunds geol-miner. Instn., 108, 255 pp., 15 pis., 30 figs. 1948. An outline of the succession and migration of noncrinoid pelmatozoan faunas in the Lower Palaeozoic of Scandinavia. Ark. Kemi Miner. Geol., 26 (13), 55 pp., 4 figs. 1951. Revision of the Caradocian-Ashgillian cystoid fauna of Belgium with notes on isolated pelmatozoan stem fragments. Mem. Inst. Roy. Sci. Nat. Belg., 120, pp. 1-47, 6 pis. 1960. Donn^es concernant le development ontogenetique des pelmatozoaires du Paliozoique (Echinodermes). Bull. Soc. Geol. Fr., s6r. 7, 1, pp. 773-783, 6 figs. Rexroad, C. B. 1967. Stratigraphy and conodont paleontology of the Brassfield (Silurian) in the Cincinnati Arch area. Bull. Ind. Dept. Nat. Resources Geol. Surv., 36, 64 pp., 4 pis., 4 figs. Rexroad, C. B. and R. W. Orr 1967. Silurian and Devonian stratigraphy of Southeastern Indiana. Guide- book to fieldtrip 2, 1st annual meeting. North Central section, Geol. Soc. Amer., pp. 41-74. Rowley, R. R. 1903. Descriptions of fossils. In Greene, G. K., Contributions to Indiana Palaeontology, part 16, pp. 156-157, pis. 46-48. Sdzuy, K. 1955. Cystoideen aus dem Leimitz-Schiefern (Tremadoc). Senckenb. Leth., 35, pp. 269-276, 1 pi., 5 figs. Shimer, H. W. and Shrock, R. R. 1944. Index fossils of North America. New York and London. IX + 837 pp., 303 pis. Sun, Y. C. 1936. On the occurrence of Aristocystis [sic] faunas in China. Bull. Geol. Soc. China, 15, pp. 477-488, 2 pis. Teller, E. E. 1911. A synopsis of the type specimens of fossils from the Palaeozoic forma- tions of Wisconsin. Bull. Wis. Nat. Hist. Soc, 9, pp. 170-271. PAUL: NORTH AMERICAN HOLOCYSTTTES FAUNA 161 Termier, G. and Termier, H. 1950. Contribution a I'^tude des faunes Pal^zoiques de I'Alg^rie. Bull. Serv. Carte G^ol. Alg6r, ser. 1, 11, 83 pp., 17 pis. Tillman, C. G. 1967. Triamara cutleri, a new cystoid from the Osgood Formation (Silurian) of Indiana. Jour. Paleontol., 41, pp. 222-226, pi. 25, 1 fig. WiNCHELL, A. and O. Marcy 1865. Enumeration of fossils collected in the Niagara Limestone of Chicago, 111., with descriptions of several new species. Mem. Boston Soc. Nat. Hist., 1, pp. 81-113, pis. 2-3. ZiTTEL, K. A. VON 1880. Handbuch der Palaeontologie, 1 Protozoa, Coelenterata, Echinodermata und Molluscoidea. Miinchen & Leipzig. VIII + 765 pp., 558 figs. APPENDIX: HOLOCYSTITES FAUNA LOCALITIES The precise locations of most of the old localities for the Holo- cystites fauna are unknown as accurate maps were not available, and localities such as "Grafton, Wisconsin" and "Jefferson County, Indiana" were considered quite adequate in the nineteenth century. I have attempted to trace as many old localities as possible and have met with some success in Indiana. Insufficient time has been available for fieldwork to trace Wisconsin, Tennessee, and Ohio localities and brief searches in Illinois and Ontario were unsuccess- ful. The following list of localities gives the most precise informa- tion available at present. I would be grateful to hear from anyone who has further information about these or other localities. Illinois 1). Joliet. There are many quarries in the Niagaran Dolomites around Joliet. The precise horizons and localities from which cystoids have been collected are unknown. 2). Romeo (Romeoville). Fisher (1925, p. 48) states that most of the fossils collected by Slocom in 1905-1906 came from two old quarries just east of the Illinois and Michigan Canal. The more northerly is in SWK, NEH, Section 35, T. 37N, R. lOE; the southern quarry extends from NE3^, NW34 into NW34> NE34, Section 2, T. 36N, R. lOE, Will County, Illinois; Romeoville 7}4' quadrangle. Both quarries had been abandoned and were already filled with water in 1925. 3). Lemont. Spoil banks on the west side of the Sanitary and Ship Canal extending from Romeo (SWM, SEi^, SW^, Section 35, T. 37N, R. lOE) north to SWJ^, NWH, NEM, Section 25, T. 37N, R. lOE. These spoil banks are still accessible but are now very unfossiliferous. 4). Chicago. Most, if not all, specimens recorded from Chicago probably came from the Bridgeport Quarry, 2 blocks west of Halsted at 27th St. Sec. 29, T. 39N, R. lOE, Englewood ly/ quadrangle. There are many other quarries on the southwest side of Chicago which could also be sources for cystoids. The cystoid horizon at 162 PAUL: NORTH AMERICAN HOLOCYSTITES FAUNA 163 Bridgeport is probably Waukesha Dolomite. This quarry is now largely filled with refuse. Indiana. (See fig. 11) 1). Marble Hill. Locality No. 18 of Foerste (1897) who records collecting one cystoid from the top of the Lower Osgood Shales. Not confirmed but from Foerste's description this probably lies on the section line between sections 20 and 29, T. 2N, R. lOE, Jefferson County, Indiana; Bethlehem 73^' quadrangle. 2). North Madison (Madison). Precise location unknown but the late W. H. Shideler gave the following information to Dr. F. D. Holland to whom I am indebted for it: "Banks of small tributary of Clifty Creek from the point where the road North of North Madison crosses this creek East to the point where the Madison- Versailles road crosses the creek. The tributary is the first creek North of North Madison." This information does not fit the geology of the area very well since all the tributaries of Clifty Creek in the area north of North Madison should be flowing on the Laurel Dolomite. They are about 100 ft. higher in elevation than the nearest outcrop of Middle Osgood Limestone. Either Dean's Branch or Little Clifty Creek best fits Shideler's information and presumably the location is further west (i.e., downstream) than Shideler states. Not confirmed. Shideler's information refers to Sections 15 and 22, T. 4N, R. lOE., Jefferson County, Indiana; Clifty Falls 73^2' quadrangle. 3-5). Big Creek. Localities 74-76 of Foerste (1897). Two to three miles of the banks of Big Creek, beginning midway between West Recovery Road and Center Recovery Road (NWlii, NW3<4, Section 16, T. 5N, R. lOE) upstream to a point where an unnamed tributary joins Big Creek on the north side, just east of East Re- covery Road (NWM, SEM, Section 3, T. 5N, R. lOE), Jefferson County, Indiana; San Jacinto 1}/^' quadrangle. This entire stretch of Big Creek is now within the Jefferson Proving Grounds. 6). Unnamed tributary to west fork of Indian Kentuck Creek where it crosses under county road. S}/2> SE34, SE34, Section 7, T. 5N, R. HE, Jefferson County, Indiana; Rexville 73^2' quadrangle. 7). Locality 58 of Foerste (1897), unnamed tributary to west fork of Indian Kentuck Creek where it parallels county road and upstream for 100 yds. SW^, NEM, SW^, Section 6, T. 5N, R. HE, Jefferson County, Indiana; Rexville 73^^' quadrangle. 164 FIELDIANA: GEOLOGY, VOLUME 24 8). Locality 77 of Foerste (1897), banks of Big Creek where it flows under the old Michigan Road and upstream for 200 yds. SWM, NEK, SWM, Section 25, T. 6N, R. lOE, Ripley County, Indiana; Rexville 73^' quadrangle. 9). Locality 81 of Foerste (1897), banks of Vestal Branch just upstream of the point where it crosses county road 225 W. S>W]4> NE}4, Section 21, T. 6N, R. HE, Ripley County, Indiana, Rexville, 73^2' quadrangle. 10). Locality 131 of Foerste (1897), road cut on Michigan Road where it descends to cross Big Graham Creek 34 mile north of cross- roads in New Marion. SW3^, NE34, Section 36, T. 7N, R. lOE, Ripley County, Indiana; Versailles l}/^' quadrangle. 11). Strippings from sump-pit in Ripley County Construction Co. quarry on southwest edge of Osgood. NW34, SW3<4, Section 22, T. 8N, R. lOE, Ripley County, Indiana; Versailles 73^' quad- rangle. 12). Strippings from ditches on the north side of the west quarry of the Napoleon Limestone Co. 0.8 miles west of the inter- section of U. S. Highway 421 and Indiana Highway 229 in Napoleon. NW34, NE34, Section 29, T. 9N, R. lOE, Ripley County, Indiana; Osgood 73^' quadrangle. Localities 1-12 are all in the upper part of the Osgood (see fig. 12) . 13). Dr. F. D. Holland (personal communication) records col- lecting Holocystites s.l. from the Lower Waldron in a quarry 3^ mile east of Sandusky, Decatur County, Indiana. SE3 Section 12, T. UN, R. 9E; Milroy 73/^' quadrangle. 14-16). Among old collections specimens are recorded from "Laurel," "Waldron," and "Near St. Paul" Indiana. Precise locations for these three are unknown and the cystoids may have come from the top of the Laurel rather than the Osgood. Quite probably other localities in the Osgood will be discovered in southeastern Indiana. The Middle Osgood Limestone is rela- tively resistant and forms waterfalls in all the creeks of the area. As Hubbard (1892) described the best method of looking for the cystoids is to walk up the creeks examining the underside of the Middle Osgood Limestone and the beds immediately beneath. Kentucky 1). Butts (1915, pp. 79-80) records cystoids from the top of the Lower Osgood Shales in Jefferson County, Kentucky. No more precise data are known. PAUL: NORTH AMERICAN HOLOCYSTITES FAUNA 165 Ohio 1). Foerste (1917, p. 260) records cystoids from the banks of Greenville Creek near Brierly Quarry, 4.5 miles east of Greenville. SWi^' quadrangle. According to Pinsak and Shaver (1963, pp. 33) the strata in this quarry are equivalent to the Louisville Limestone of Indiana and are thus younger than the Greenville Creek exposures. Tennessee 1) . Newsom : The holotype of Pustulocystis pentax is said to have come from a quarry dump at Newsom Station, Davidson County, Tennessee; Belleview 73^' quadrangle. Other specimens are just labelled "Newsom." Presumably the horizon is Osgood but the Waldron is also exposed at Newsom. 2). Pegram: One uncatalogued specimen in the USNM col- lections is labelled "1 mile East of Pegram" Cheatham County, Tennessee; Kingston Springs 73^2' quadrangle. No further informa- tion is available. 3). Clifton: One specimen of Holocystites alternatus in the USNM collections is labelled as coming from Clifton, Wayne County, Tennessee; Clifton 73^' quadrangle. Wisconsin. 1). Grafton: The holotype of Holocystites cylindricus is recorded from Grafton, Ozaukee County, Wisconsin; Cedarburg 73^^' quad- rangle. Since Grafton is practically on the Silurian-Devonian boundary the horizon is probably high up in the Racine. Teller (1911) doubts that this locality is correct. 2) Racine: Many specimens in old collections are from Racine. There is a very extensive quarry which extends over almost all of the SEM, Section 29, T. 4N, R. 23E, Racine County, Wisconsin, a smaller quarry is located in the center of the north half of Section 32, T. 4N, R. 23E. Whether these are the source of cystoids or not is unknown. Racine North 73^' quadrangle. Presumably type Racine. 166 FIELDIANA: GEOLOGY, VOLUME 24 3). Waukesha: Many specimens in old collections are from Waukesha. There is an extensive quarry on the northeast side of the city which extends from the NW'^^j', section 35 into the SWK, section 26, T. 7N, R. 19E, Waukesha County, Wisconsin; Waukesha 73 2' quadrangle. This could be the source of these cystoids. Many other large quarries occur 7-8 miles north of Waukesha however. Presumably type Waukesha. Ontario 1). Parks (1913, p. 132) records "Holocystites globosus Miller" from the lower, more shaley beds of the Rochester Shale at Grimsby, Ontario. I have not seen, nor do I know the whereabouts of, the specimens on which this record is based. It seems most unlikely that the specimens were incorrectly identified since no other record of holocystitids from these fossiliferous strata is known. 2). Wiarton. Bolton (1957, p. 61) records "Megacystites (?) sp. indet." from the Wiarton Member of the Amabel Formation at Wiarton and north of Colpoy Village, Bruce Peninsular, Ontario. The specimens are small and indeterminate so this record is also doubtful. Finally, mention might be made of Browne's tentative identifica- tion of an echinoderm fragment from the Upper Silurian of Bowning Village, near Yass, New South Wales, Australia as Holocystites (Browne, 1964, pp. 286-287, pi. 21, fig. d). Mrs. Browne very kindly sent me a photographic negative of this specimen (Australian Museum F. 28030) from which prints were made. The specimen is incomplete and unidentifiable in my opinion.